After the Ninth CIIE: Record Numbers, Three Destinations for an Exhibit, and a Customs Route That Lets It Stay

  • The ninth China International Import Expo closed on 10 November 2026 after six days, with 4,142 exhibitors and an intended deal value of USD 85.45 billion, up 2.3 per cent on the previous edition.
    第九届中国国际进口博览会于 2026 年 11 月 10 日闭幕,会期 6 天,参展企业 4142 家,按一年计意向成交额 854.5 亿美元,较上届增长 2.3%。
  • Total exhibition space passed 440,000 square metres for the first time, of which the enterprise exhibition alone accounted for 370,000 square metres.
    展览展示总面积首次突破 44 万平方米,其中企业展面积首次达到 37 万平方米。
  • The five record figures do not share a denominator: they count companies, floor space, corporate size, products and money, and a record in one says nothing about the others.
    五项创新高的数据并不同用一个分母:它们分别数的是企业、面积、企业体量、展品和金额,其中一项破纪录,并不说明其余四项同样如此。
  • An exhibit has three destinations after the show: return, retention in a bonded site, or conversion into a cross-border e-commerce product where the positive list allows it.
    展品闭幕后有三种去向:退运出境、留存于保税场所,或在正面清单允许时转为跨境电商商品。
  • Exhibit retention has a defined route: an exhibitor that does not ship its goods home may move them into a customs special supervision area or a bonded supervision site, at which point the goods count as remaining in China.
    展品留存有一条明确通道:展商不将展品运回,可先将货物移入海关特殊监管区或保税监管场所,货物即算留在境内,无需原路退回。
  • Reasonable quantities of exhibits sold during the show are exempt from customs duty, import value-added tax and consumption tax, while other exhibits attract tax-favoured sales of no more than USD 20,000 per exhibitor.
    展期内销售的合理数量进口展品免关税、进口环节增值税和消费税,其他展品每家展商享受税收优惠的销售限额不超过 2 万美元。
  • The USD 20,000 figure governs sales made under the favourable treatment, not the retention of goods, and it is counted per exhibitor rather than per product.
    2 万美元这个数字管的是适用优惠的销售,而不是货物留存,而且按展商计、不按单品计。
  • A year-round layer sits beside the venue, and the tenth edition already has 143 companies and institutions signed up with more than 100,000 square metres committed.
    场馆之外还有一层常年安排;第十届已有 143 家企业及机构签约,展览面积超过 10 万平方米。

After the Ninth CIIE: Record Numbers, Three Destinations for an Exhibit, and a Customs Route That Lets It Stay | 第九届进博会闭幕之后:五项数据、展品的三种去向与留在境内的通道

The numbers that closed the ninth CIIE

The ninth China International Import Expo closed on 10 November 2026, six days after it opened. At the closing briefing the organisers put five headline figures at record highs: 4,142 exhibitors; total exhibition space above 440,000 square metres for the first time, with the enterprise exhibition alone at 370,000 square metres, also a first; more than 300 Fortune Global 500 companies and industry leaders; 485 representative debuts, more than 200 of them global premiers; and intended deal value of USD 85.45 billion, 2.3 per cent above the previous edition.

Around those five sit the supporting figures. The country exhibition drew 162 countries, regions and international organisations, of which 69 mounted comprehensive displays, 56 were BRI partner countries and 12 were developed economies including Sweden and Canada; five countries and one international organisation appeared for the first time. On the buying side there were 51 trading delegations and more than 700 sub-delegations, over 470,000 registered visitors, and 80 side events, among them a first China-EU medical device matchmaking session that produced more than 100 preliminary cooperation intentions on site. The innovation incubation zone hosted 633 science and technology projects from nearly 40 countries and regions, up nearly 20 per cent, with nearly 50 debut products at the exhibit end. The cultural exchange area covered more than 40,000 square metres with over 1,300 exhibiting institutions, 296 China Time-honoured Brands and 374 intangible cultural heritage items.

A closing briefing is a genre with its own conventions, and the convention is to lead with records. The useful question for a company that did not attend — or attended and is now deciding whether to return — is not whether the numbers are large, but what each of them is counting.

Five records, five different denominators

The five figures do not measure the same thing, and they do not move together.

Exhibitor count measures breadth: how many organisations decided the trip was worth making. Floor space measures capacity: how much of the venue the organisers could fill, and this year the enterprise exhibition reached 370,000 square metres inside a total above 440,000. The count of Fortune Global 500 companies and industry leaders measures the top of the distribution rather than its middle, and it is a subset of the 4,142. The 485 debuts measure novelty — how many products the exhibitors chose to launch here rather than elsewhere. The USD 85.45 billion measures money, and it is the only one of the five for which the briefing gave a growth rate.

That asymmetry is worth pausing on. A record exhibitor count with flat deal value would describe a show that is popular but not productive. A record deal value with a falling exhibitor count would describe a show that is concentrating. This year the direction is the same across all five, which is a stronger signal than any one of them, but the strength comes from the combination rather than from the size of any single figure.

For an exhibitor planning next year, the practical use of the five is diagnostic. A company whose objective is lead generation should watch visitor and delegation numbers; one whose objective is launch visibility should watch the debut count; one whose objective is contract signature should watch the deal value and, more importantly, its own conversion rate from the previous edition.

The country exhibition is not a sales channel

The country exhibition drew 162 countries, regions and international organisations. Sixty-nine of them mounted comprehensive displays, 56 were BRI partner countries, and 12 were developed economies, Sweden and Canada among them. Five countries and one international organisation appeared for the first time.

A national pavilion does a different job from an enterprise stand. It promotes a country — its exports, its tourism, its investment climate, sometimes its culture — and the companies inside it appear under a national umbrella rather than on their own account. Deals are rarely signed on a pavilion floor; the meetings that matter happen in the enterprise halls or in the hotels around them.

That does not make the pavilion irrelevant to a company, for two reasons.

The first is informational. The composition of the country exhibition is a reading of which governments are currently investing in access to the Chinese market. Twelve developed economies taking part, Sweden and Canada among them, is a statement that the European and North American presence has not thinned; 56 BRI partner countries is a statement about where the growth in participation is coming from; five countries and one international organisation arriving for the first time is a pipeline, because a first pavilion is usually followed by a first company stand.

The second is practical, and it is the reason trade officials are worth meeting at a fair. A regulatory question — a standard that does not fit, a registration that has stalled, a certification nobody will accept — is rarely solved by a commercial counterpart. It is solved, when it is solved, through the channel that begins with the trade authority of the exhibitor’s own country. A week in which that authority is in the same building as the regulator is a week worth using.

The buyer side: delegations, visitors and eighty side events

The buying side of the expo is organised in a way that a first-time exhibitor often underestimates.

There were 51 trading delegations and more than 700 sub-delegations. The delegation is the unit through which Chinese buyers are brought to the show: provinces, municipalities, state-owned groups and industry associations assemble their purchasers, give them a schedule, and route them through the halls. For an exhibitor, being on a delegation’s route is worth more than a good position by the main entrance, because a routed visitor arrives with a mandate.

Registered visitors exceeded 470,000, and that figure should be read with the same care as any footfall number. A registered visitor is not a qualified buyer; the qualification happens in the 80 side events, in the meeting rooms and on the stands where a conversation lasts longer than ninety seconds. The side events are also where the format is being tested — this year’s new China-EU medical device session is an example — and a sector session is usually a better use of an afternoon than a general hall.

The cultural exchange area sits at the consumer end of the same system: more than 40,000 square metres, over 1,300 exhibiting institutions, 296 China Time-honoured Brands and 374 intangible cultural heritage items. It is the part of the programme that brings a domestic public through the gates, and it is why the retail tie-ins around the venue — the consumption map, the ticket-stub offers at more than 100 merchants — have an audience to work with. A consumer brand launching in China reads this area as a distribution channel for attention, not as a trade channel for orders.

What “intended deal value” actually counts

An intended deal is a signed statement of intent. It is not an invoice, not a purchase order, and not revenue.

The measure is annualised — “counted on a one-year basis” — which tells you something about how the figure is assembled. A framework agreement covering several years is expressed as a one-year equivalent; a one-off order is counted at its face value. The number is therefore a common denominator imposed on commitments of different shapes, which makes it useful for comparison across editions and imprecise as a forecast of anything in particular.

Three gaps sit between the figure and cash. The first is performance: an intention signed at a trade fair is followed by due diligence, internal approval, and sometimes a change of mind. The second is timing: an annualised figure says nothing about which quarter the money lands in. The third is scope: an intention may cover distribution rights, service agreements or cooperation frameworks rather than goods alone.

None of this makes the number meaningless; it makes it a demand indicator rather than a revenue line. Read alongside the customs backdrop — goods trade of RMB 44.81 trillion in the first ten months of the year, up 20 per cent — the USD 85.45 billion describes an appetite rather than a delivery. The report issued around the closing makes the same point from the other side: China’s share of global exports absorbed rose from 3.2 per cent in 2000 to 9.2 per cent in 2025, and the share of BRI partner countries’ exports going to China reached 59.4 per cent.

For a foreign company, the operational conclusion is narrow and useful. If your own intention is one of the USD 85.45 billion, the number that matters to you is not the aggregate but the conversion rate on your own pipeline, measured at the next edition.

The debuts, and what a buyer reads in them

Of the 485 representative debuts, more than 200 were global premiers — products shown nowhere before. The rest were Asian or China firsts.

A buyer reads a debut for two things: whether the product exists anywhere else, and whether the China launch is simultaneous with the global one. A global premier answers both questions in the affirmative, and that is why the classification is watched. A product launched in China at the same time as everywhere else tells a distributor that the supplier treats this market as a primary one, with the supply commitment that implies. A product that arrives two years after its European launch tells the opposite story.

There is a limit worth stating plainly. A debut is a marketing classification, not a regulatory one. A device shown as a global premier at the expo still needs whatever approvals its category requires before it can be sold in China, and the expo does not shorten that path. What the expo does is compress the period in which a buyer can compare competing claims in one hall.

That is also why the count of 485 matters less than its composition. A hall of incremental variants produces a large number without producing much information. A hall in which more than 200 products are being shown for the first time anywhere produces a genuine reference point, because a buyer can see, in a single week, where the technology frontier currently sits in each category.

The innovation pipeline behind the 633 projects

The innovation incubation zone hosted 633 science and technology projects from nearly 40 countries and regions, up nearly 20 per cent on the previous edition, with nearly 50 debut products launched at that end of the hall.

The zone is the pipeline, and it is worth understanding why it is structured as a separate area rather than as a cheap corner of the main halls. An incubator booth is a small commitment: less space, less build cost, a shorter presence. For a project that has a prototype and a patent but no Chinese entity, no distributor and no registration, that is the right size of commitment — enough to test whether anyone in the market cares, without the cost of a full stand.

The organisers’ interest is straightforward. A project small enough to need an incubator booth this year is a candidate for a full stand in two or three years, and the growth rate of the zone is a leading indicator of the growth rate of the enterprise exhibition. That is why the count of projects is reported alongside the count of deals.

For the project itself, the zone is a low-cost test of three questions: whether Chinese buyers understand the product without a long explanation, whether a distributor or a strategic partner is interested, and whether the regulatory path is worth starting. None of those questions is answered by the expo, but all three can be scoped in a week. The zone is an exhibition, not an approval channel; nothing about a booth changes what a product must do to be sold.

The first China-EU medical device session

Among the 80 side events, one was new: a China-EU medical device matchmaking session, held for the first time, which produced more than 100 preliminary cooperation intentions on site.

Sector-specific matchmaking is a different instrument from a general fair. A general fair puts a product in front of whoever walks past; a matchmaking session puts a defined group of sellers in a room with a defined group of buyers and gives them a timetable. For medical devices the format matters more than in most sectors, because the route to market runs through registration, and the party that holds a registration or a distribution network is often not the manufacturer.

The figure to note is not the 100-plus intentions but the fact that the session was held at all. A first session is a test of format: if the sector responds, the model tends to be repeated and extended to adjacent categories. For a device maker that did not attend this year, the practical question for next year is whether a sector session exists in its category, because a matchmaking timetable is a more efficient use of four days than a stand.

Three destinations for an exhibit

For an exhibitor, the last day of a trade fair has always posed one question: ship it home, sell it, or write it off. The measures the customs administration applied to this edition answer it with a set of defined destinations.

The first is return. The goods leave as they came, and the temporary entry is closed. This is the default for equipment that has a home market commitment, and it is the most expensive option when the goods are heavy.

The second is retention. An exhibitor that does not ship its goods home may move them into a customs special supervision area or a bonded supervision site. Once there, the goods count as having remained in China, and there is no requirement to send them back the way they came. The machine that was demonstrated on the stand can stay in the country as a bonded asset — available for a later demonstration, a spare-parts pool, or a sale that has not yet closed.

The third is conversion. Where an exhibit appears on the cross-border e-commerce retail import positive list, it can be converted inside the bonded zone into a cross-border e-commerce product and sold online. The goods reach individual buyers through a channel with its own rules rather than entering the domestic market in the ordinary sense.

Alongside those three sits the sale channel that operates during the show: reasonable quantities of exhibits sold at the expo are exempt from customs duty, import value-added tax and consumption tax. That is not a destination but an exit, and it is why a consumer-goods exhibitor can sell samples off the stand.

The decision is not symmetrical, and it is not reversible in the way exhibitors assume. Return closes the file. Retention keeps the file open under supervision. Conversion moves the goods into a different regime with different documentation. Choosing late is the expensive mistake; the paperwork for retention has to be arranged while the goods are still in the hall.

Timing is what decides the outcome. The commercial negotiation usually runs to the last hour of the last day; the customs arrangement has to be in place before that, because a site has to agree to take the goods and a broker has to file. An exhibitor that starts both conversations on the same morning will finish only one of them.

What retention does not do

Retention is a status, not a clearance. Moving goods into a bonded site changes where they sit and which set of rules governs them; it does not settle the duty that a domestic sale would attract.

That distinction is where most misunderstandings start. An exhibitor that moves a machine into a bonded warehouse has solved a logistics problem, not a tax problem. If the machine is later sold to a buyer in China, the sale is an import like any other, and the fact that the goods arrived as exhibits does not change the treatment. What retention buys is time and optionality: the goods are in the right country, under supervision, while the commercial question is settled.

The same logic applies to conversion. Moving a product into the cross-border e-commerce channel does not make it a domestically traded good; it places it in a channel designed for retail sales to individuals, with its own list of eligible products and its own documentation. The channel is a route to consumers, not a route around product rules.

For a company deciding between the three, the useful test is what the goods are for. If they are for demonstration and after-sales support, retention is the natural fit. If they are consumer products with a retail price point, conversion is worth examining. If they are committed to another market, return is cheaper than it looks once storage and supervision costs are counted.

The USD 20,000 cap: it governs sales, not retention

The favourable treatment has a second tier. Beyond the reasonable quantities sold during the show, other exhibits attract tax-favoured sales of no more than USD 20,000 per exhibitor.

Two readings of that sentence cause most of the confusion, and both are avoidable.

The first is that the cap limits what an exhibitor may sell. It does not; it limits what is sold under the favourable treatment. A sale above the threshold is not prohibited, it is simply handled as an ordinary import, with the duty and taxes that implies.

The second is that the cap applies to goods the exhibitor keeps. It does not. Retention under the bonded route is a separate mechanism with its own conditions; the USD 20,000 figure belongs to the sales tier.

The unit of measurement is also worth noting: the cap is set per exhibitor, not per product or per stand. A company showing six product lines from one booth has one limit, not six. That is a planning constraint for a distributor-style exhibitor whose stand is a shop window for a catalogue, and it is invisible to a capital-goods exhibitor whose entire presence is one machine and one negotiation.

The positive list gate

Conversion is available only where the exhibit appears on the cross-border e-commerce retail import positive list. That is a gate, not a formality, and it is the single fact that decides whether the third destination exists for a given product.

The logic of a positive list is that eligibility is defined by enumeration rather than by exclusion. A product not on the list is not converted, however suitable it looks commercially. For an exhibitor, this means the question has to be answered before the goods move, not after: check the list, then check with the bonded zone operator what the conversion requires in practice.

The commercial consequence is asymmetric across exhibitors. For a consumer brand — cosmetics, food, small appliances, fashion — the channel is a way to test Chinese demand without committing to general-trade import, and it is often the reason a first-time exhibitor takes a small stand at all. For a machinery maker, the channel is irrelevant, and the retention route does the work instead.

What the customs file should hold

The four outcomes — sale during the show, retention, conversion, return — are decided item by item, and each one leaves a different paper trail. The mistake that costs money is treating the file as an afterthought to be assembled once the stand is down.

Three properties of the decision make early preparation necessary.

The first is that the decision is made by people who are not on the stand. The commercial lead knows which machine is going to which buyer; the customs broker knows which site will accept the goods; the bonded zone operator knows whether the product is on the list. On the last day, none of them is in the same conversation unless someone arranged it in advance.

The second is that the routes are not interchangeable after the fact. Goods that have been moved into a bonded site are under supervision, and moving them out again is a transaction with its own conditions. Goods that have been returned have left. A decision taken on day six is a decision taken with the fewest options available.

The third is that the favourable treatment has boundaries that have to be evidenced, not asserted. The reasonable-quantity standard for sales during the show, the USD 20,000 per-exhibitor threshold, and the positive-list test for conversion are all questions that will be answered by a record rather than by an explanation.

What that suggests in practice is a single file, opened before the show, with one line per item: what it is, which route it will take, who confirmed the route, and what evidence supports the choice. For a retained machine, that means the site and the operator’s acceptance. For a converted product, the list entry and the zone’s process. For a sold sample, the quantity and the running total against the cap. For a returned item, the closure of the temporary entry.

None of this is exotic. It is the ordinary discipline of a trade fair, applied to the one part of the fair that most exhibitors leave to the last afternoon.

The 365-day layer

Six days has an obvious limitation: a negotiation that starts on day five has nowhere to go. The 365-day layer is the answer, and it is the part of the expo that a first-time visitor tends to miss.

Greenland Global Commodity Trade Port, across the road from the “four-leaf clover” venue, runs year-round on a six-days-plus-365-days model, so discussions that did not close at the show can continue in a permanent setting with permanent staff. Shanghai added to that layer this year with a “Shanghai Custom” experience zone, a “CIIE debut new trend” series of events, and a “CIIE consumption map” that lets a ticket stub unlock offers at more than 100 merchants.

The three additions do different jobs. The experience zone gives a product somewhere to be handled rather than merely displayed. The event series keeps a launch alive after the press cycle ends. The consumption map extends the expo to retail, converting a trade visitor’s ticket into consumer traffic — which matters to a brand whose Chinese launch depends on consumer awareness rather than on a distributor contract.

None of this is customs policy, and none of it changes what happens to the goods. What it changes is the calendar. An exhibitor that treats the show as a six-day event and leaves on day seven has used perhaps a third of what the platform offers.

There is a pairing here that exhibitors rarely make explicit. A retained machine needs somewhere to be shown while a buyer decides; a year-round platform needs inventory that can be demonstrated on request. The bonded site supplies the first, the platform supplies the second, and neither is much use on its own.

What the tenth edition already shows

The tenth edition is not yet open, and 143 companies and institutions have signed for it, committing to more than 100,000 square metres.

Re-booking at that scale before this edition’s invoices are settled is the clearest signal available of how exhibitors treat the show: as a channel with an annual cycle rather than a one-off event. The behaviour is rational. Space in the enterprise halls is allocated by application, returning exhibitors tend to want the same position, and a company that waits until the spring to decide is negotiating from a weaker position than one that signed in November.

The 143 reported at this closing is a snapshot taken eleven months before the next edition opens, not a final total.

For a foreign company weighing attendance, the question is therefore less whether the expo works than where in the cycle to enter. There are two distinct entry points. One is as an exhibitor building a first contact: a small stand, a defined product, a target list, and a decision about retention made before the show rather than after it. The other is as a brand using the retention and conversion routes to test a market without committing inventory — a smaller presence, but one that requires the customs decision to be made earlier, not later.

A sequence for the last day

  • Decide, product by product, between sale during the show, retention, conversion and return; the four carry different tax and documentation outcomes.
  • For anything sold, confirm the reasonable-quantity treatment applies, and check where the USD 20,000 per-exhibitor threshold falls before the stand is dismantled.
  • For anything retained, name the customs special supervision area or bonded supervision site in advance and check what the operator requires at the gate.
  • If the product is on the cross-border e-commerce retail import positive list, ask the bonded zone operator about conversion before the goods move.
  • Book the 365-day platform conversation while the buyer is still in the building.
  • Record the customs decision in the same file as the commercial one; the two are settled by different people on different timetables.

Sources

  • 中国国际进口博览局·第九届进博会闭幕新闻通气会数据(新华财经转载):https://www.ciie.org/zbh/cn/2024/medium-item1/xinhua/20261110/66226.html
  • 中国国际进口博览局·人民日报闭幕评论(转载):https://www.ciie.org/zbh/cn/2024/medium-item1/people/20261111/66285.html
  • 上海市商务委员会(市口岸办)·第九届进博会闭幕与展后安排:https://kab.sww.sh.gov.cn/xwzx/001001/20261110/64af37e0-2da1-4294-b202-ab65636d2d20.html
  • 央视网·进博会闭幕报道(含海关展后留存通道表述):https://news.cctv.com/2026/11/11/ARTIG4yp4hgGJZObEJgbQay9261111.shtml
  • 央广网·第九届进博会闭幕:https://www.cnr.cn/shanghai/tt/20261110/t20261110_527868175.shtml

第九届进博会闭幕之后:五项数据、展品的三种去向与留在境内的通道

闭幕时的五个数字

第九届中国国际进口博览会于 2026 年 11 月 10 日闭幕,从开幕算起会期 6 天。闭幕通气会上,主办方给出的五项核心数据均创新高:参展企业 4142 家;展览展示总面积首次突破 44 万平方米,其中企业展面积首次达到 37 万平方米;参展的世界 500 强和行业龙头企业超过 300 家;全球首发、亚洲首秀、中国首展的代表性展品 485 项,其中全球首发首秀展品超过 200 件;按一年计意向成交额 854.5 亿美元,较上届增长 2.3%。

这五项之外还有一圈支撑数据。国家展有 162 个国家、地区和国际组织参与,其中 69 个国家和国际组织展示综合形象,”一带一路”共建国家 56 个,瑞典、加拿大等 12 个发达国家参展,5 个国家和 1 个国际组织首次参展。采购一侧是 51 个交易团、700 多个交易分团,注册观众超过 47 万人,80 场配套活动,其中首次举办的中欧医疗器械专场贸易对接会现场达成初步合作意向超过 100 项。创新孵化专区聚集了近 40 个国家和地区的 633 个科创项目,较上届增长近 20%,展品端首发新品近 50 项。人文交流板块展览面积超过 4 万平方米,展示机构超过 1300 家,中华老字号 296 个,非遗项目 374 项。

闭幕通气会是一种有固定写法的文体,而它的固定写法是先报纪录。对一家没有参展、或者参过展正在犹豫要不要再来的公司来说,真正有用的问题不是这些数字大不大,而是每一个数字数的究竟是什么。

五项纪录,五个不同的分母

这五个数字量的不是同一件事,也不会同向变动。

参展企业数,量的是广度:有多少机构认为这趟路值得走。面积,量的是容量:主办方能填满多少场馆,今年企业展在 44 万平方米的总盘子里占到 37 万平方米。世界 500 强和行业龙头企业的数量,量的是分布的上沿而不是中段,而且是 4142 家里的一个子集。485 项首发,量的是新意:有多少产品被选择在这里而不是别处亮相。854.5 亿美元量的是钱,也是五项里唯一给出增速的一项。

这个不对称值得停一下。参展商数量创纪录而成交额持平,描述的是一场热闹但不产出结果的展会;成交额创纪录而参展商数量下滑,描述的是一场正在收敛的展会。今年五个方向一致,这比其中任何一项都更有说服力,但说服力来自组合,而不是来自某一个数字的绝对大小。

对准备明年参展的企业,这五项的正确用法是诊断。以获取线索为目标的企业,该盯的是观众与交易团数据;以发布声量为目标的,该盯首发数量;以签约结果为目标的,该盯成交额,但更重要的是盯自己上一届从接触到签约的转化率。

国家展不是卖货的地方

国家展有 162 个国家、地区和国际组织参与,其中 69 个国家和国际组织展示综合形象,”一带一路”共建国家 56 个,瑞典、加拿大等 12 个发达国家参展,5 个国家和 1 个国际组织首次参展。

国家馆和企业展位干的是两件不同的事。国家馆推介的是一个国家——出口、旅游、投资环境,有时还有文化;里面的企业挂的是国家招牌,不是自己的名号。展台上很少签单,真正起作用的会面发生在企业馆里,或者场馆周边的酒店里。

但这不意味着国家馆与企业无关,理由有两条。

第一条是信息。国家展的构成,是一份关于”哪些政府当下正在为中国市场准入投入资源”的读本。12 个发达国家参展、其中包括瑞典和加拿大,说明欧洲与北美的存在没有变薄;56 个共建”一带一路”国家,说明参与面的增量来自哪里;5 个国家和 1 个国际组织首次参展,则是管线——第一次来设馆,通常意味着随后会有第一批企业展位。

第二条是实用,也是贸易官员值得在展会上见一面的原因。一个监管问题——标准对不上、注册卡住了、认证谁都不认——很少由商业对手方解决。它被解决的时候,往往是通过从参展国自己的贸易主管部门开始的那条线。如果一周之内,这个部门与监管机构在同一栋楼里,这一周就该用起来。

采购一侧:交易团、观众与 80 场配套活动

展会的采购一侧是被组织起来的,而首次参展的企业常常低估这一点。

51 个交易团、700 多个交易分团。交易团是把中国买家带到展会上的单位:省市、国有集团、行业协会把各自的采购人员组织起来,给他们一张日程,再按路线穿过各个馆。对展商来说,被排进某个交易团的路线,比拿到一个靠大门口的好位置更值钱,因为被组织来的观众是带着授权来的。

注册观众超过 47 万人,这个数字要和任何客流量数字一样小心地读。注册观众不等于合格买家;资格是在 80 场配套活动里、在会议室里、在那些能谈超过九十秒的展位上完成的。配套活动也是形式被测试的地方——今年新增的中欧医疗器械专场就是一例——而一场行业专场,通常比一个综合馆更值得花掉一个下午。

人文交流板块处在同一套系统的消费端:展览面积超过 4 万平方米,展示机构超过 1300 家,中华老字号 296 个,非遗项目 374 项。它是让本地公众走进场馆的那部分安排,也是场馆周边那些零售联动——消费地图、凭票根在上百家商户享优惠——有受众可用的原因。对中国首发的消费品牌来说,这个板块是一条聚集注意力的渠道,不是一条成交订单的贸易渠道。

“意向成交额”到底统计什么

意向成交是签下来的意向,不是发票,不是订单,也不是收入。

这个口径是”按一年计”的,这四个字透露了它的组装方式:一份覆盖数年的框架协议,会被折算成一年的等值;一笔一次性订单,按面值计入。于是这个数字是把形状各异的承诺压到同一个分母上的结果——用来做跨届比较很合适,用来预测任何具体事项则很粗糙。

从数字到现金之间有三道缝。第一道是履约:在展会上签下的意向,后面还要过尽职调查、内部审批,有时还会变卦。第二道是时间:按一年计的金额,并不告诉你钱落在哪个季度。第三道是范围:一笔意向可能覆盖经销权、服务协议或合作框架,而不只是货物。

这些都不让这个数字失去意义,只是把它从收入指标变成了需求指标。把它和海关总署给出的背景放在一起看——今年前 10 个月货物贸易进出口总值 44.81 万亿元,同比增长 20%——854.5 亿美元描述的是一种胃口,而不是一次交付。闭幕同期发布的《出口中国能力发展报告》从另一侧说了同一件事:中国吸纳全球出口的比重由 2000 年的 3.2% 升至 2025 年的 9.2%,共建”一带一路”国家对华出口占比升至 59.4%。

对外资企业,操作层面的结论很窄但有用。如果你的意向是这 854.5 亿美元里的一笔,那么对你有意义的不是总量,而是你自己那条管线在下一届展会上的转化率。

首发名单,采购方从中读什么

485 项代表性展品里,超过 200 件是全球首发,此前没有在任何地方展出过;其余是亚洲首秀或中国首展。

采购方从一次首发里读两件事:这个产品在别处是否已经存在,以及中国上市是否与全球同步。全球首发对这两个问题都给出了肯定回答,这也是这个分类被盯着看的原因。一个在中国与全球同步上市的产品,等于告诉经销商:供应商把这里当作主力市场,随之而来的是供货承诺。一个比欧洲晚两年才到的产品,讲的是相反的故事。

有一条边界需要说清楚:首发是营销分类,不是监管分类。一件以全球首发身份展出的器械,在能够销售之前,仍然要走完它所属类别要求的全部审批,展会并不缩短这条路。展会压缩的是另一件事——采购方在一周之内、在一个馆里,把互相竞争的说法摆在一起比较的时间。

也正因如此,485 这个数字本身不如它的构成重要。一个馆里全是微调型号,能凑出很大的数字却不产生多少信息;一个馆里有 200 多件产品是全球第一次亮相,就形成了一个真正的参照点,因为采购方可以在一周内看清每个品类当前的技术边界在哪里。

633 个项目背后的创新管线

创新孵化专区聚集了近 40 个国家和地区的 633 个科创项目,较上届增长近 20%,展品端首发新品近 50 项。

这个专区是管线,而它被单独划成一块区域而不是塞进主馆角落,原因值得说清。孵化展位是一种小承诺:面积小、搭建省、在场时间短。对一个有样机、有专利,但在中国没有主体、没有经销商、也没开始注册的项目,这正好是合适的承诺量级——足以试出市场上有没有人在意,又不必付出一整个展位的成本。

主办方的算盘很直白:今年小到需要孵化展位的项目,就是两三年后整摊展位的候选,专区的增速因此是企业展增速的先行指标。这也是为什么项目数量会和成交数量一起报出来。

对项目自身,这个专区是对三个问题的低成本测试:中国买家能不能不需要长篇解释就听懂这个产品;经销商或战略伙伴有没有兴趣;监管路径值不值得开始走。这三个问题都不会在展会上得到答案,但都能在一周内摸清轮廓。需要说明的是,专区是展台,不是审批通道——展位不改变一件产品要卖出去所必须满足的任何条件。

首次举办的中欧医疗器械专场

80 场配套活动里有一场是新的:中欧医疗器械专场贸易对接会,首次举办,现场达成初步合作意向超过 100 项。

专场对接和综合展会用的是两种不同的工具。综合展会把产品摆给路过的人看;对接会把一群确定的卖方和一群确定的买方放进同一间屋子,并给他们一张时间表。医疗器械比多数行业更依赖这种形式,因为这个行业通往市场的路上横着注册,而手里握着注册证或经销网络的,往往不是制造商。

值得记的不是那 100 多项意向,而是这场会办了。第一次办,本身就是对形式的测试:如果行业反应积极,这种模式通常会被复制并延伸到相邻品类。对今年没参展的器械企业,明年实际要问的问题是:自己的品类有没有专场;因为一张对接时间表,往往比一个展位更能用满四天。

展品的三种去向

对展商来说,展会最后一天的老问题从来没变:运回去、卖掉,还是核销掉。海关总署为本届推出的便利化措施,用一组明确的去向回答了它。

第一种是退运。货物怎么来就怎么走,暂时进境的手续就此关闭。对有本土市场承诺的设备,这是默认选项;对重量大的货物,这也是最贵的选项。

第二种是留存。展商如不将展品运回,可先将货物移入海关特殊监管区或保税监管场所。货物一旦进入,即算留在境内,无需原路退回。展台上演示过的那台设备,可以以保税资产的身份留在国内——留作后续演示、备件池,或者等一笔还没谈成的销售。

第三种是转换。展品如果属于跨境电商零售进口正面清单,可以在保税区内直接转为跨境电商商品上线销售。这类货物通过一条自成一体的渠道触达个人买家,而不以通常意义进入国内市场。

与这三种去向并列的,还有展期内本身的销售通道:展期内销售的合理数量进口展品,免关税、进口环节增值税和消费税。这不是一种去向,而是一个出口,也是消费品展商能在展台上直接卖样品的原因。

这三条路并不对称,也不像很多展商以为的那样可以随时反悔。退运会结案;留存是把案子在监管之下继续挂着;转换是把货物挪进另一套规则、另一套单证。真正的昂贵错误是决定得太晚——留存要办的手续,必须在货还在馆里的时候安排。

真正决定结果的是时点。商务谈判通常会拖到最后一天的最后一个小时;海关这边的安排必须在这之前就位,因为得有场所同意接货,也得有报关行去递单。如果两件事同一个上午才开始,通常只能办成一件事。

留存不做什么

留存是一种状态,不是一次清关。把货物移入保税场所,改变的是货物所在的位置和适用的规则,并不会结清内销本该承担的税负。

误解大多从这里开始。把一台设备移进保税仓的展商,解决的是物流问题,不是税务问题。这台设备日后卖给中国买家,那笔销售仍然是一次普通的进口,货物当初以展品身份进来这件事,不改变它的待遇。留存买到的是时间和选择权:货已经在对的国度里、在监管之下,而商业问题还在谈。

转换同理。把产品挪进跨境电商渠道,不等于把它变成可以在境内自由流通的货物,而是把它放进一条为个人零售设计的渠道,这条渠道有自己的商品清单、自己的单证要求。它是通往消费者的路,不是绕开产品规则的路。

对要在三者之间做选择的公司,最实用的检验是问货物用来做什么。用于演示和售后支持的,留存是自然选择;有零售价位的消费品,值得研究转换;已经另有市场承诺的,把仓储和监管成本算进去之后,退运往往比看上去便宜。

两万美元限额:管的是销售,不是留存

优惠待遇还有第二层。在展期内销售的合理数量之外,其他展品每家展商享受税收优惠的销售限额不超过 2 万美元。

这句话有两种常见误读,都可以避开。

第一种是把限额理解成”展商最多只能卖这么多”。并非如此:它限的是适用优惠的销售,而不是销售本身。超过这条线的销售不被禁止,只是按一般进口处理,该缴的关税和税照缴。

第二种是把限额套到展商自己留下的货物上。也不对。经保税渠道留存是另一套机制、另一套条件;2 万美元这个数字属于销售那一层。

计量单位同样值得注意:限额按展商计,不按单品、也不按展位计。一家在一个展位上摆六条产品线的公司,只有一个额度,不是六个。这对把展位当橱窗、以目录示人的经销型展商是实实在在的规划约束,而对整个参展就是一台设备、一场谈判的装备类展商,则几乎不可见。

正面清单这道闸

转换只在展品属于跨境电商零售进口正面清单时才可用。这是一道闸,不是一道手续,也是决定第三种去向对某个具体产品是否存在的那一个事实。

正面清单的逻辑是”列明即准入”,而不是”未列禁止”。不在清单上的产品不能转换,无论它在商业上多么合适。对展商,这意味着问题必须在货物移动之前回答,而不是之后:先查清单,再向保税区运营方问清转换在操作上需要什么。

商业后果在不同展商之间并不对称。对化妆品、食品、小家电、时尚这类消费品牌,这条渠道是在不承诺一般贸易进口的前提下测试中国需求的办法,也常常是一家首次参展的小展位之所以存在的原因。对机械制造商,这条渠道无关紧要,干活的是留存通道。

海关那份卷宗该装什么

四种结果——展期内销售、留存、转换、退运——是逐件决定的,每一种留下的纸面痕迹都不一样。真正花钱的错误,是把卷宗当成撤展之后再来收拾的收尾工作。

这个决定的三个属性,决定了准备必须提前。

第一,做决定的人不在展台上。商务负责人知道哪台设备给哪个买家;报关行知道哪个场所会收货;保税区运营方知道这个产品在不在清单上。到了最后一天,除非有人事先安排,这几个人不会出现在同一场对话里。

第二,这些路径事后并不互通。已经移入保税场所的货物处在监管之下,再移出来是一笔有自身条件的业务;已经退运的货物已经离境。第六天才做的决定,是可选余地最少的决定。

第三,优惠待遇有边界,而边界要靠证据,不能靠说明。展期内销售的合理数量、每家展商 2 万美元的限额、转换所依据的正面清单,这些都是要用记录来回答的问题,而不是用解释来回答的问题。

落到操作上,这意味着在展前就建一份卷宗,一件货一行:是什么、走哪条路、谁确认了这条路、依据是什么。留存的设备,写清场所与运营方的接收;转换的产品,写清清单条目与保税区的流程;卖掉的样品,写清数量与相对限额的累计;退运的货物,写清暂时进境手续的关闭。

这些都不稀奇。它就是参展的常规纪律,只是被用在了多数展商留到最后一个下午才处理的那一部分。

6 天之外的 365 天

6 天的日程有个显而易见的局限:第五天才谈起来的生意,没地方接着谈。常年这一层就是答案,而它恰恰是首次到访者最容易忽略的部分。

与”四叶草”场馆一路之隔的绿地全球商品贸易港,按”6 天+365 天”模式常年运行,展期内没谈完的事项可以在一个有常设人员、常设场地的环境里继续。今年上海又给这一层加了内容:”上海定制”体验区、”进博首发新潮流”系列活动,以及一份”进博消费地图”——凭进博票根可以联动上百家商户。

这三样各干各的活。体验区让产品有地方被上手摆弄,而不只是被看见;系列活动让一次发布在新闻周期结束之后还活着;消费地图把展会延伸到零售,把一张专业观众的票根变成消费客流——对一个中国首发要靠消费者认知而不是靠一纸经销合同立住的品牌,这一条尤其重要。

这些都不是海关政策,也不改变货物会发生什么。它们改变的是日历。把这场展会当成六天活动、第七天就走的展商,大概只用了这个平台三分之一的功能。

这里还有一组搭配,展商很少明说:留存的设备需要有地方在买家犹豫期间继续展示,常年平台需要有可以应要求演示的现货。保税场所提供前者,平台提供后者,单用哪一头都不够。

第十届已经显露的轮廓

第十届尚未开幕,已有 143 家企业及机构签约,展览面积超过 10 万平方米。

在本届账还没结清之前就按这个规模续订,是展商态度最直接的信号:他们把这当作一个有年度周期的渠道,而不是一次性活动。这个行为是理性的。企业展的展位按申请分配,老展商通常希望保住原来的位置,而等到来年春天才决定的公司,谈判位置比 11 月就签约的公司弱得多。

对正在权衡是否参加的境外公司,问题因此不在于这场展会是否有用,而在于从周期的哪一段切入。切入点是两个。一个是作为建立首次接触的展商:小展位、明确的产品、一张目标名单,以及一个在展前而不是展后做出的留存决定。另一个是作为借助留存与转换通道试水市场、暂不压库存的品牌:存在感更小,但要求海关那一侧的决定做得更早,而不是更晚。

需要说明的是,闭幕时报出的 143 家,是距下一届开幕还有十一个月时的一个截面,不是最终总数。

闭幕日的一份操作顺序

  • 按单品逐项决定:展期内销售、留存、转换,还是运回。四者的税务与单证结果各不相同。
  • 凡是要卖的,先确认合理数量的口径适用,并在撤展之前算清 2 万美元这条按展商计的线落在哪里。
  • 凡是要留的,提前锁定海关特殊监管区或保税监管场所,并问清运营方在卡口的要求。
  • 产品若在跨境电商零售进口正面清单内,在货物移动之前先向保税区运营方问清转换流程。
  • 趁买家还在馆内,把 365 天平台的后续约谈排好。
  • 把海关的决定和商业的决定记在同一份文件里;这两件事由不同的人、按不同的时间表了结。

来源

  • 中国国际进口博览局·第九届进博会闭幕新闻通气会数据(新华财经转载):https://www.ciie.org/zbh/cn/2024/medium-item1/xinhua/20261110/66226.html
  • 中国国际进口博览局·人民日报闭幕评论(转载):https://www.ciie.org/zbh/cn/2024/medium-item1/people/20261111/66285.html
  • 上海市商务委员会(市口岸办)·第九届进博会闭幕与展后安排:https://kab.sww.sh.gov.cn/xwzx/001001/20261110/64af37e0-2da1-4294-b202-ab65636d2d20.html
  • 央视网·进博会闭幕报道(含海关展后留存通道表述):https://news.cctv.com/2026/11/11/ARTIG4yp4hgGJZObEJgbQay9261111.shtml
  • 央广网·第九届进博会闭幕:https://www.cnr.cn/shanghai/tt/20261110/t20261110_527868175.shtml

Shanghai Rewrites the Customs Unit: From One Company to One Supply Chain

  • Shanghai issued the Supply Chain Headquarters Opinion as 沪商规〔2026〕4 号 on 20 October 2026, in force the same day and valid until 19 October 2031.
    上海以沪商规〔2026〕4 号印发《上海市关于鼓励企业设立供应链总部的若干意见》,2026 年 10 月 20 日印发并自当日起施行,有效期至 2031 年 10 月 19 日。
  • For the first time the city proposes moving customs supervision of a supply chain headquarters from the company as the unit to the supply chain as the unit.
    文件首次提出,探索把供应链总部的通关监管从”以企业为单元”转向”以供应链为单元”。
  • AEO cultivation is to move towards “1+N” group and industry-chain cluster certification, so that accreditation follows the chain rather than stopping at one legal entity.
    AEO 培育转向”1+N”集团式、产业链供应链式组团培育和认证,认证沿链条展开,不再止于单一法人。
  • Three thresholds apply: a managed type above 5 billion yuan in revenue or 2 billion in trade, a service type above 3 billion or 2 billion, and a platform type above 1 billion in platform revenue with 15 billion in GMV.
    三类门槛分别是:管理型年营收超 50 亿元或国际贸易规模达 20 亿元,服务型超 30 亿元或达 20 亿元,平台型年平台运营收入超 10 亿元且成交总额超 150 亿元。
  • Imported goods replaced free of charge because of damage, shortage, quality defects or specification mismatch are granted duty exemption under the rules.
    进出口货物因残损、短少、品质不良或规格不符,以无代价抵偿方式免费补偿或更换的,按规定实施免征关税。
  • Recognition is dynamic: exit, relocation of the main business, serious violation, or a failed re-assessment left uncorrected will each end it.
    认定实行的是动态管理:主动退出、主要经营业务迁出上海、重大违法违规,以及经评估不再适合且整改未达标,都会导致资格被取消。
  • Existing trade headquarters that cannot meet the new tests keep their old treatment only until 31 August 2027.
    已认定的贸易型总部若不满足供应链总部认定条件,原待遇仅保留至 2027 年 8 月 31 日前。

Shanghai Rewrites the Customs Unit: From One Company to One Supply Chain | 上海改写通关监管单元:从一家企业到一条供应链

Why Shanghai added a fourth category

Shanghai has run a headquarters economy for two decades, and for most of that time the city worked with a single template: the trade headquarters, recognised on trade volume and the scale of import and export it handled. The Supply Chain Headquarters Opinion, issued on 20 October 2026 by the municipal commerce commission together with the financial committee office, Shanghai Customs, the economic and digitalisation commissions, the talent bureau and the data bureau, adds a second template rather than replacing the first.

The relationship between the two is set out in the text rather than left to inference. A supply chain headquarters is described as the upgraded form of a trade headquarters. Where a trade headquarters aggregates one or several trade functions — buying, selling, or both — a supply chain headquarters coordinates the whole chain, from the purchase of inputs through storage and movement to the final sale, and does so across borders.

That difference shows up in what the opinion measures. The test is not volume alone but control: whether the entity fuses logistics, commercial flow, information flow and capital flow into what the text calls four flows in one, and whether it exercises strategic decision-making and management over affiliates across regions. A trading company that books large volumes but takes its instructions from an overseas parent will struggle on both counts; a smaller entity that genuinely directs procurement, inventory and settlement for a group can pass.

Six management functions are named — strategic planning, global procurement and distribution, warehousing and logistics, digital management, cross-border settlement and risk control — and an applicant must hold at least three of them. That list is the real gate. It asks a group to demonstrate that its Shanghai entity decides for the chain, not merely books for it, and it is the part of the file that a finance team cannot assemble on its own.

The timing explains the design. Shanghai’s five-year plan for the international trade centre sets out an intention to move the headquarters economy from a concentration of trade functions to management of the entire supply chain. The opinion is the instrument that turns that intention into a recognition with a certificate attached, and the certificate is what unlocks the customs, foreign exchange, data and visa measures that follow in the later articles.

Three types and their thresholds

The opinion splits supply chain headquarters into three types, and each has its own arithmetic.

A managed headquarters handles centralised procurement, logistics decisions and final product sales for a group. It needs annual operating revenue above 5 billion yuan, or international trade volume of 2 billion yuan. On top of that it needs at least two branches in China or abroad, and it must either purchase raw materials centrally for the group or sell finished goods on the group’s behalf. The substance of this type is a procurement and sales hub.

A service headquarters is set up by a professional supply chain management company and runs integrated trade, logistics and settlement for other clients. It needs revenue above 3 billion yuan or trade volume of 2 billion yuan, plus one of two alternatives: at least three branches or service outlets, or stable supply chain partnerships with five or more countries and regions where trade with any single one exceeds 100 million yuan. The substance here is managing chains for others.

A platform headquarters operates cross-border matching and supporting services online. It needs platform operating revenue above 1 billion yuan, at least 2,000 registered members or resident merchants with more than 20 percent of them from outside Shanghai, and annual platform transaction value above 15 billion yuan. Unlike the first two, all three conditions apply at once, which makes this the tightest of the three tests.

Two baselines sit underneath all of them: the enterprise must conduct business in Shanghai, and it must have no serious violation of laws or regulations in the previous three years. It must also hold branches or subsidiaries through investment, or manage them under authorisation from the parent. That last requirement looks procedural but decides which entity in a group may apply — the authorisation chain breaks at a particular level, and the applicant can only be the entity above the break.

There is also a discretionary route. Where a headquarters is of major strategic significance to Shanghai’s ambition to be a global supply chain hub, or has made an outstanding contribution, it may be recognised through the same filing process on a case-by-case basis. For a group whose figures sit just under one of the lines, that sentence is the difference between waiting a year and applying now, though it should be read as a route that requires a written case rather than a waiver of the tests.

The certification file, document by document

The procedure is deliberately short, but the file is specific, and each document maps onto a condition in the earlier articles.

The application form is the municipal Supply Chain Headquarters Application Form. The prior year’s financial report and supporting evidence, such as a third-party audit report, is what substantiates the revenue and trade figures. Proof of investment in branches or subsidiaries, or proof of parental authorisation to exercise cross-regional strategic decision-making and management over them, is what substantiates the control test. Material on the domestic and overseas distribution network and service outlets substantiates the branch or partnership conditions.

The global supply chain operations report is the substantive document. It is expected to cover the company profile, its global procurement, product sales, warehousing and cross-border settlement, the state of its supply chain management information system, its overseas supply chain footprint, and its trade scale, service capacity and management system. In practice this is where a group explains how the four flows actually connect, and where a reviewer can see whether the entity is a decision point or a booking centre.

A three-year undertaking of no serious violation completes the base file. Platform applicants add a report and supporting material on registered member numbers and prior-year transaction value, because those two figures carry two of the three platform conditions.

The review itself has four steps. The company files with the commerce authority of the district where it operates. The district carries out a comprehensive assessment and completes its preliminary review within 10 working days before forwarding the file. The municipal commission reviews with relevant units and industry experts. The passing list is published for five working days, and after that the commission issues plaques and publishes the recognised list.

Two practical notes follow from the timetable. Ten working days at district level means the file has to be complete when it arrives, because there is no long window for correspondence. And the five-day publication period is the point at which a name becomes public, which matters for groups that would rather not disclose the structure of their Chinese operations before a transaction closes.

The real change: a chain, not a company

Article seven carries the sentence that has drawn the most attention since publication: the city will explore shifting customs supervision of a supply chain headquarters from the company as the unit to the supply chain as the unit, forming a supervision model and clearance facilitation service built around the headquarters and its upstream and downstream.

Set against how Chinese customs has worked, that is a structural change rather than an added convenience. AEO status, credit management, inspection rates and bonded arrangements have all attached to a legal person holding a customs registration. A group with twelve subsidiaries therefore holds twelve separate credit profiles. A compliance failure at one does not travel to the other eleven, which is the reassuring half; the other half is that the good standing of the eleven does not travel either. Supply chain competence built at group level converts into no clearance benefit at all under a per-entity lens.

A chain-based unit changes what is measured. The perimeter becomes the headquarters plus the suppliers and distributors it can see and control, and the asset that earns facilitation becomes its visibility into that perimeter — order flows, inventory positions, counterparty credentials, exception alerts. That is why the same article pairs the idea with “1+N” group and industry-chain cluster cultivation and certification: one core enterprise plus N members, assessed together, with obligations and benefits running along the links.

For a group, the consequence is a change in what compliance money buys. Under a per-entity model, the rational spend is to make each legal person independently clean. Under a chain model, the rational spend is to make the chain legible — a single system that can show where goods are, who supplied them and what happened when something went wrong. The second is harder to buy at short notice, which is why it favours groups that already run integrated systems.

Two limits should be stated plainly. The text says explore, not implement: no timetable, scope or technical standard is attached, so the operative rules will come from Shanghai Customs in practice rather than from the opinion itself, and reading explore as already in force overstates the certainty available today. And the duty exemption in the same article — for goods compensated or replaced free of charge because of damage, shortage, quality defects or specification mismatch — applies according to existing rules; it is a statement of treatment, not a new relief.

What shifts in audit, inspection and guarantees

The three places where a unit change is felt first are audit, inspection and guarantees, and it is worth being concrete about each.

Audit. A per-entity audit asks whether this legal person kept proper books, classified goods correctly and paid what it owed. A chain-level audit asks a harder question: whether the headquarters’ account of its suppliers and distributors is accurate, and whether the same transaction is described consistently at both ends. Groups that reconcile their Chinese and overseas records only at quarter end will find that question uncomfortable.

Inspection. Inspection rates follow credit. Under a per-entity model, a newly established entity starts without history and is treated accordingly, even if its parent has traded for decades. Under a chain model, the headquarters’ record could in principle inform the treatment of members, which is precisely what cluster certification is meant to enable. Until the pilot rules exist, the practical advice is to assume the per-entity model still governs and to plan for the transition rather than around it.

Guarantees. Duty guarantees and bonded arrangements are sized against risk, and risk is assessed per entity. A chain view would allow a group to present collateral and history at the level where the risk actually sits. That is a meaningful difference for groups that hold inventory in several bonded locations, because it changes whether each location must stand on its own balance sheet.

None of this happens automatically. The opinion creates the direction and the recognition that goes with it; the customs authority creates the mechanics. A group that wants the benefit should expect to be asked for evidence of chain visibility, and should build that evidence before the conversation rather than during it.

Finance: two channels and how to choose

The financial measures are grouped by function, and the most useful way to read them is as two channels rather than a list.

The first channel is the cross-border trade high-level opening pilot. The commerce commission pushes the recognised list to the Shanghai branch of the foreign exchange regulator and to prudently run banks, so that headquarters can be included among the pilot’s quality enterprises. Inclusion brings exemption from registration for special refunds and netting of trade receipts and payments, with banks setting their own document review. For a group with imports, exports and re-exports running through the same entity, netting is usually the item that shows up fastest in the cash flow.

The second channel is the free trade account route for offshore trade. The opinion supports including headquarters in the city’s offshore trade pilot, with facilitation for international settlement and trade finance through free trade accounts, and the Lin-gang special area is to support offshore trade subsidiaries set up there by headquarters. The choice between the two is not technical preference but business model: a group that mostly moves goods across the border will live in the first channel, while a group that books transactions between overseas parties and needs to settle them from Shanghai will need the second.

Around those two channels sit the supporting items. There is coaching and training for cash pool arrangements, with the domestic main account able to handle current account receipts and payments for onshore and offshore members and to net a period’s transactions into a single entry. Headquarters may join the digital yuan cross-border payment pilot. Supply chain finance is offered as a menu — receivables, order, warehouse receipt and inventory pledge financing, and supply chain notes — with export credit insurance and policy financing from the Shanghai branch of Sinosure.

The sequencing matters more than the menu. Netting and refund registration are operational changes that a treasury team can adopt in a quarter. Cash pool restructuring and offshore trade routing touch the group’s legal structure and take longer. A group planning a Shanghai supply chain headquarters should therefore decide early whether the entity will be a settlement centre as well as a procurement centre, because that decision determines which channel it needs and how much work precedes the filing.

One further article sits next to finance and is often read as boilerplate: risk control. Headquarters are to be guided through supply chain resilience assessments, asked to hold safety stock for critical materials, and expected to prepare interruption contingency plans and alternatives. They are also encouraged to diversify sourcing so as to reduce dependence on any single country or region, and the city’s international trade public service platform is to provide early-warning information on global supply chain disruption, critical material supply and trade policy change.

Read alongside the customs article, that is not boilerplate. A headquarters that has done a resilience assessment, mapped its single points of failure and recorded alternatives is precisely the entity that can answer the question a chain-based supervision model asks. The risk article is, in effect, the documentation standard for the clearance article.

Logistics: warehouses, multimodal and rail capacity

The logistics article is the one that most directly rewards a company for treating Shanghai as a physical node rather than a billing address, and it is easy to skim past because it reads like a list of encouragements.

On warehousing, headquarters are encouraged to set up warehouses in Shanghai, to convert existing space and to install automation. They are also supported in establishing overseas warehouses, distribution centres and spare-parts bases in key regions along the Belt and Road, and may apply to join the relevant public overseas warehouse service system. For a group running cross-border e-commerce or after-sales operations, that combination — a Shanghai warehouse plus an overseas network — is the physical counterpart of the four-flows test.

On transport, the opinion provides facilitation for multimodal combinations: sea plus rail, sea plus road, air plus land. More concretely, it establishes a capacity priority mechanism for supply chain headquarters on China-Europe freight trains, with preference in slot allocation and departure frequency. That is a scarce resource allocated by list rather than by market, and it is one of the few measures in the document that a headquarters would feel in a specific week rather than in a general way.

The trade-network provisions sit alongside: exhibitions recommended by a headquarters may be given priority in the city’s key overseas exhibition programme, and headquarters participating in those exhibitions receive support. The logic is consistent — the city is willing to underwrite the cost of showing a Shanghai-based chain to overseas buyers, because the chain is what the policy is trying to attract.

Data: the fast channel inside the negative list

On data, the opinion does something narrower than it might appear: a fast service channel for outbound data transfer is to be built inside the negative-list regime that already operates in the Shanghai pilot free trade zone and Lin-gang special area, and in the national service-sector opening pilot in Shanghai. Headquarters may apply to the relevant departments for compliance guidance and advisory services.

The placement matters. The channel does not create a new exemption; it creates a faster route through a regime that already defines which data transfers need review and which do not. For a group, the practical gain is predictability: a named channel, a compliance guide, and a place to ask before a transfer is designed rather than after.

Two habits follow. First, map the data before designing the system. The negative list is organised by scenario, so a headquarters that can describe its transfers in the same categories will move faster than one that arrives with a data architecture diagram. Second, keep the compliance file with the transfer record. If the channel is used, the evidence of what was transferred and why should be retrievable, because the regime is built on documentation rather than on prior approval alone.

For most foreign groups, the transfers that matter are operational rather than strategic: order data, inventory positions, supplier master data, after-sales records. Those are exactly the categories a supply chain headquarters generates daily, which is why the data article sits in this opinion at all rather than in a digital policy document.

People: residence, visas and the travel card

The talent measures are written for the people who run a regional chain, and they are worth reading as a package.

Headquarters are to be recommended for inclusion in the municipal list of key institutions for talent introduction, which brings facilitation in household registration, medical care and applications for talent apartments. The company may also put staff forward for the city’s major talent programmes and for special awards in key industrial fields.

For foreign staff, three provisions matter. Foreign senior managers may be recommended to apply for permanent residence in China. Foreign employees may buy or sell foreign exchange for salary purposes within a threshold set by a prudently run bank without submitting supporting documents. And staff who travel frequently to APEC economies for business may apply for the APEC Business Travel Card, which replaces the need for a visa on each trip to participating economies.

There is also a visa route for visitors rather than employees: foreign business people whom a headquarters needs to invite for repeated short stays may apply to the public security exit-entry authorities for multiple-entry visas. For a regional headquarters that hosts supplier audits, training and technical visits, this is often more useful than a residence permit, because the people who need it are not on the payroll.

The common thread is that all four measures depend on the headquarters being on the recognised list. None is available on application by an individual alone. That makes the certification a personnel instrument as much as a trade one, and it explains why human resources tends to become an unexpected stakeholder in the filing.

Dynamic management and the four grounds for withdrawal

Recognition is not permanent, and the opinion says so in operational terms rather than as a caveat.

The municipal commission operates dynamic management. Recognition can be withdrawn on four grounds: the enterprise applies to exit; its main business relocates out of Shanghai; it commits a serious violation of law; or a dynamic assessment finds it no longer suitable and it fails to meet the conditions within a set rectification period.

Around that sit two obligations. District commerce authorities maintain a tracking mechanism and must report to the municipal commission when they learn of circumstances that may mean a headquarters no longer meets the conditions. And a headquarters must itself report adjustments to its business and material changes.

The practical reading is that the certificate is a statement about a current state of affairs, not a status conferred once. A group that restructures its Chinese operations, moves a regional function to another city, or changes the entity that holds the authorisation should expect to revisit the file. Groups that treat the plaque as a permanent asset and stop maintaining the underlying evidence are the ones most exposed to the fourth ground.

There is a softer layer of management as well. The opinion provides for an industry committee under a trade association to connect upstream and downstream companies, a communication mechanism between municipal and district commerce departments, at least one training session a year for newly recognised headquarters covering export control compliance, trade risk, legal and financial topics, and information sharing of the recognised list with customs, tax, foreign exchange and talent authorities.

For a compliance team, the list of documents to keep current is therefore short but specific: the authorisation or investment proof, the network map, the prior-year financials, and the global supply chain report. The first three change when the group restructures. The fourth changes when the business does. A group that refreshes all four annually will rarely be surprised by a re-assessment.

The transition from trade headquarters

Article fifteen is the one to diarise. Trade headquarters recognised under the earlier opinion on encouraging trade headquarters are to be managed under this opinion. Those that do not meet the supply chain tests may keep their existing treatment, but only until 31 August 2027.

Three things follow. First, the two regimes are connected rather than parallel: the new category is an upgrade path, and an existing trade headquarters is expected to move along it or to plan for the treatment to end. Second, the date is fixed, not a rolling grace period, so a group that intends to requalify has a defined window in which to restructure its Shanghai entity, its network and its evidence. Third, because the opinion took effect on the day it was issued, the countdown began on 20 October 2026, not on the November publication.

The design is consistent with how Shanghai has handled headquarters policy before: create a higher tier, grandfather the lower one for a period, and let the market decide who climbs. For a group, the question is not whether the old treatment is adequate today — it is whether the business it supports will still be adequate after the summer of 2027.

What a foreign group should verify first

Start with the applicant. Proof of investment in affiliates or parental authorisation decides which Chinese entity qualifies, and it is often a holding company rather than the operating company that carries the revenue. Where the two are separate, the authorisation has to be documented before the file is assembled, not after.

Then check where the numbers sit. Revenue, international trade volume, platform operating income and transaction value are all measured for the applicant. A group that books trade through a regional hub may find the Shanghai entity’s figures thinner than the group total suggests, and the answer is usually a question about business allocation rather than a conclusion that the threshold is out of reach.

Then check chain visibility. If supervision does move to a chain unit, facilitation will be calibrated to what the headquarters can evidence about its suppliers and distributors. Groups already running one system across the chain will be better placed than those reconciling spreadsheets, and the gap will not close in the weeks before a filing.

Then check the “1+N” question: whether upstream suppliers and downstream distributors, including overseas members, can be brought into a cluster certification, and who bears the compliance obligation if one member fails. The answer often decides whether the group moves real supply chain management authority to Shanghai or keeps it elsewhere and treats the entity as a booking centre.

Finally, put the dates in one place: the filing window, the five-day publication, the transition deadline of 31 August 2027, and the point at which the customs pilot rules are expected. A headquarters decision that ignores any one of them tends to be revisited within a year.

One note on sequence. The filing is the easiest part of this to describe and the hardest to do out of order. A group that starts with the form will spend its time assembling evidence; a group that starts with the three questions — who holds the mandate, where the numbers sit, and what the chain can show — will find that the form fills itself. The difference is usually four to six weeks, which is the difference between filing in a given quarter and filing in the next one.

That sequence also explains why the opinion is worth reading in full rather than as a summary of thresholds. The thresholds tell a group whether it qualifies. The other fifteen articles tell it what qualification is for, and several of them — customs, data, visas — only become useful if the entity has been built to use them.

Sources

  • 上海市商务委员会、中共上海市委金融委员会办公室、中华人民共和国上海海关、上海市经济和信息化委员会、上海市人才工作局、上海市数据局·《上海市关于鼓励企业设立供应链总部的若干意见》(沪商规〔2026〕4 号)全文:https://www.investsh.org.cn/cn/investsh/zctz/2026102813370022149406
  • 上海市外国投资促进中心·《”投资上海”一周资讯动态回顾(1109—1113)》:https://www.investsh.org.cn/cn/investsh/tzdt/2026111316344603129477
  • 澎湃新闻·《民生智库|商务领域动态热点速览(20261101—1106)》,2026-11-10:https://m.thepaper.cn/newsDetail_forward_34453134

上海改写通关监管单元:从一家企业到一条供应链

上海为什么再加一类总部

上海经营总部经济已有二十年,此前的主线基本只有一种模板:贸易型总部,按贸易额与进出口规模认定。2026 年 10 月 20 日,上海市商务委员会会同市委金融办、上海海关、市经济信息化委、市人才工作局、市数据局印发的《上海市关于鼓励企业设立供应链总部的若干意见》,加的是第二种模板,而不是替换第一种。

两者的关系,文件是写出来的,不靠推断:供应链总部是贸易型总部的能级提升形态。贸易型总部聚合的是一项或几项贸易功能——买、卖,或买卖兼有;供应链总部协调的是整条链条,从投入品采购,到仓储与运输,再到最终销售,而且跨越关境。

这个差别体现在文件测什么上。衡量标准不只是规模,还有控制力:这家机构是否把物流、商流、信息流、资金流做成文件所称的”四流合一”,是否对分支机构、子公司行使跨区域供应链业务的战略决策和管控职能。一家账面贸易额很大、但指令来自境外母公司的贸易公司,两条都不容易过;反过来,一家规模小一些、但真正替集团指挥采购、库存与结算的主体,反而能过。

文件列出六项综合管理功能:战略规划、全球采购与分销、物流仓储、数字化管理、跨境结算、风险管控。申请企业至少要占三项。这份清单才是真正的闸门:它要求集团证明,上海的这家主体是在替整条链做决定,而不只是替整条链记账;也正因为如此,这部分材料不是财务部门单独能凑出来的。

时点能解释设计意图。上海加快国际贸易中心建设的五年规划提出,要推动总部经济从贸易功能集聚向供应链全链条管理升级。这份意见就是把那句话变成一张带证书的认定,而证书正是后面各条——通关、外汇、数据、签证——的钥匙。

三种类型与三套门槛

文件把供应链总部分为管理型、服务型、平台型三类,各有各的算法。

管理型负责集团内部统一采购、物流管理等供应链决策,以及最终产品销售。门槛是年度营业收入超过 50 亿元,或国际贸易规模达到 20 亿元;同时在境内外拥有两个及以上分支机构;并且集团内部原材料由申请企业集中采购,或产成品由申请企业统一销售。这一类的实质,是集团采购与销售的枢纽。

服务型由专业供应链管理企业设立,为其他企业客户提供贸易、物流、结算一体化的全球供应链服务。门槛是年度营业收入超过 30 亿元或国际贸易规模达到 20 亿元,另加二者之一:境内外设有三个及以上分支机构或服务网点;或与五个及以上国家或地区的贸易伙伴建立稳定合作关系,且单个国家年贸易额超过 1 亿元。这一类的实质,是替别人管链。

平台型依托互联网等信息技术设立,从事跨境供应链撮合交易与配套服务。门槛是年度平台运营收入超过 10 亿元;平台注册会员或入驻企业超过 2000 家,且非本市企业占比超过 20%;面向企业的平台年度成交总额超过 150 亿元。与前两类不同,这三条要同时满足,是三类里最紧的一套。

三类之上还有两条共同前提:在上海市经营业务;近三年内无严重违反法律法规行为。此外,企业须通过投资形式,或经母公司授权,对分支机构、子公司行使跨区域战略决策和管控职能。这一条看似程序性,实际决定了集团里哪家主体有资格申报——授权链条断在哪一层,申报主体就只能落在断点之上。

还有一道酌情认定的口子:对上海建设全球供应链枢纽、提升国际贸易中心能级具有重大战略意义或作出突出贡献的,可按申报流程酌情认定。对数字刚好卡在某条线下方半步的集团,这句话决定了是再等一年,还是现在就能递材料;不过它要求一份写得出来的理由,而不是把门槛免掉。

认定材料,逐项拆开看

程序刻意做得短,但材料很具体,而且每一份都对应前面某一条门槛。

申请表是市定的《上海市供应链总部申请表》。上年度财务报告及相关佐证材料,例如第三方审计报告,是用来支撑营收与贸易规模的。对分支机构或子公司的投资证明,或母公司授权其对分支机构、子公司行使跨区域战略决策和管控职能的授权证明,是用来支撑控制力这一条的。境内外分销网络、服务网点布局情况和相关投资证明,是用来支撑分支数量或合作关系那一条的。

真正有分量的是全球供应链业务情况报告。文件要求它涵盖企业概况,开展全球采购、产品销售、物流仓储、跨境结算等业务的情况,供应链管理信息系统建设情况,境外供应链布局情况,以及贸易规模、服务能力、供应链管理服务体系等内容。实务上,这就是集团说明”四流”如何真正连起来的地方,也是审核方判断这家主体是决策点还是记账点的依据。

近三年无严重违法违规行为的承诺书,补齐基础材料。申报平台型的,另需就平台注册会员数、上年度成交总额出具情况报告及证明材料,因为这两个数字撑起了平台三条门槛中的两条。

审核本身分四步。企业向所在地的区商务主管部门提出申请;区级结合本区实际作综合评估,在 10 个工作日内完成初审后转报市商务委;市商务委会同有关单位和行业专家评审;通过名单公示 5 个工作日,无异议后统一授牌、对外公布。

时间表带来两点提醒。区级只有 10 个工作日,意味着材料递上去时就得齐备,没有长时间的往返补正空间。5 个工作日的公示,则是名字公开的那一刻——对不希望在一笔交易交割前披露中国架构的集团,这个节点要提前排进计划。

真正的变化:监管的是一条链,不是一家公司

第七条里那句被讨论最多的话是:探索将供应链总部通关监管从”以企业为单元”向”以供应链为单元”转变,形成以供应链总部上下游为单元的监管模式和通关便利服务。

放回中国海关既有的作业方式看,这是架构层面的调整,不是便利措施的叠加。AEO 资质、信用管理、查验比例、保税安排,历来都挂在一个有海关登记编码的法人身上。一家有十二个子公司的集团,就有十二份彼此独立的信用档案。其中一家出了合规问题,不会传导给其余十一家,这是让人放心的一半;另一半是,其余十一家攒下的良好记录,也不会自动惠及它。集团层面再强的供应链管理能力,在单法人口径下也换不来任何通关便利。

以链条为单元,改变的是被测量的范围。相关边界变成总部加上它看得见、管得住的供应商与分销商;能换来便利的资产,是总部对这个边界的可见度——订单流向、库存位置、上下游资质、异常预警。正因如此,同一条把这一设想与”1+N”集团式、产业链供应链式组团培育和认证并列:一家核心企业加若干成员,一起评估、一起认证,义务与便利沿链条传导。

对集团而言,结果是合规预算的用途变了。单法人口径下,理性的花法是让每个法人各自干净;链条口径下,理性的花法是让整条链可读——一套能说清货在哪里、谁供的、出问题时发生了什么。后一种能力很难临时买到,这也是它更有利于本就跑一套系统的集团的原因。

两处边界要说清楚。其一,文件写的是”探索”,不是”实施”,没有配套的时间表、适用范围或技术标准,真正的操作规则会在上海海关的实践中形成;把”探索”读成”已落地”,会高估眼下可得的确定性。其二,同一条里的无代价抵偿免税——因残损、短少、品质不良或规格不符而免费补偿或更换的进出口货物,按规定免征关税——是按现行规定执行,属于把既有待遇写进总部政策包,而不是新设一项减免。

稽查、查验、担保三个环节各差在哪

单元一换,最先有感觉的是稽查、查验、担保这三处,值得逐个说清。

稽查。单法人稽查问的是:这家法人账做得对不对,商品归类对不对,该缴的税缴了没有。链条稽查问的是更难的问题:总部对供应商与分销商的描述准不准,同一笔交易在两端是不是一个说法。习惯到季末才对一次境内外账的集团,会觉得这个问题不好答。

查验。查验比例跟着信用走。单法人口径下,新设主体没有历史,就按没有历史对待,哪怕母公司做了几十年生意。链条口径下,总部的记录原则上可以影响成员的待遇,这正是组团认证想打开的口子。在试点规则出来之前,稳妥的做法是先假定单法人口径仍然有效,为过渡做准备,而不是绕着它设计。

担保。关税担保与保税安排的额度按风险定,而风险是按主体评估的。链条视角允许集团在风险真正所在的那一层提供担保、呈现历史。对在多个保税地点压着库存的集团,这差别很实在:它决定每个地点是否都要独立撑起自己的资产负债表。

这些都不会自动发生。意见给出的是方向和随之而来的认定,机制由海关建立。想要这项便利的集团,应当预期会被要求提供链条可见度的证据,并且要在对话之前把证据建好,而不是在对话当中现找。

金融:两条通道,怎么选

金融措施按功能编排,最省力的读法是把它们看成两条通道,而不是一张清单。

第一条是跨境贸易高水平开放试点。商务部门把认定名单推送至国家外汇管理局上海市分局及审慎合规银行,支持银行把供应链总部纳入试点优质企业范围。进入范围后,可享特殊退汇免登记、贸易收支轧差结算,单证审核类型和方式由银行自主决定。对进口、出口、转口都跑在同一家主体上的集团,轧差通常是最快体现在现金流上的那一条。

第二条是自由贸易账户下的离岸经贸通道。文件支持把供应链总部纳入本市离岸经贸业务企业试点,通过自由贸易账户为国际结算、贸易融资提供便利;临港新片区对总部在区内设立的离岸贸易专营子公司按规定给予支持。两者的取舍不是技术偏好,而是业务模型:主要让货物跨关境的集团会活在第一条通道里;在境外交易对手之间成交、却要从上海完成结算的集团,需要第二条。

两条通道之外是配套项。资金池业务有政策辅导和培训,可通过国内资金主账户集中代理境内外成员企业的经常项目收支,把一段时期的收付合并为一笔轧差结算;可参与数字人民币跨境支付试点;供应链金融给出一整套工具——应收账款、订单、仓单和存货质押融资,以及供应链票据——中国信保上海分公司提供出口信用保险与保单融资。

顺序比清单更重要。轧差与退汇免登记是财务团队一个季度内可以落地的操作变化;资金池重构与离岸业务路径则牵动集团法律架构,周期更长。因此,计划设立上海供应链总部的集团,应当尽早决定这家主体是否同时充当结算中心——这个决定决定了它需要哪条通道,以及递材料之前还有多少活要干。

金融旁边还有一条常被当成套话的条款:风险管控。文件要求指导供应链总部开展供应链韧性评估,建立关键物资安全库存机制,制定供应链中断应急预案和替代方案;鼓励强化多元化布局,在全球建立关键零部件和原材料采购来源,降低对单一国家或地区的依赖;并依托本市国际贸易公共服务平台,提供全球供应链中断风险、关键原材料供应风险、贸易政策变化风险等预警信息服务。

与通关那一条放在一起读,这就不是套话。一家做过韧性评估、标出过单点故障、记录过替代方案的供应链总部,恰恰就是链条式监管想要问的那家主体。风险条款在效果上,是通关条款的文档标准。

物流:仓库、多式联运与班列舱位

物流这一条,最直接奖励的是把上海当成实体节点而不是开票地址的企业,但它读起来像一串鼓励性表述,很容易被跳过。

仓储方面,文件鼓励供应链总部在本市设立仓库、开展仓库改建和添置自动化设备;也支持在”一带一路”沿线国家等重点区域设立海外仓、物流配送中心等,并可申请纳入公共海外仓相关服务体系。对做跨境电商或售后业务的集团,这一组合——上海仓加海外网络——正是”四流合一”在物理层面的对应物。

运输方面,文件对”海运+铁路””海运+公路””空运+陆运”等多式联运提供便利服务;更具体的是,建立供应链总部中欧班列运力优先保障机制,在舱位配置、发运频次上给予倾斜保障。这是一种按名单分配而非按市场分配的稀缺资源,也是全文中少数能让总部在某一周就感受到差别的措施之一。

贸易网络条款与之并列:总部推荐的海外专业展会,可优先纳入本市重点海外展推荐项目;总部参加本市重点海外展,给予支持。逻辑是一致的——城市愿意为”把一条以上海为节点的链条展示给海外买家”承担成本,因为它要吸引的正是这条链。

数据:负面清单之内的快车道

数据这一条做的事,比看上去窄:在中国(上海)自贸试验区及临港新片区、国家服务业扩大开放综合试点地区(上海)的数据出境负面清单管理制度框架下,建立供应链总部数据出境快速服务通道;总部可向相关部门申请数据出境合规指引和相关咨询服务。

位置很重要。这条通道并不创设新的豁免,它是在一套已经界定”哪些出境需要评估、哪些不需要”的制度里,开出一条更快的路。对集团而言,实际收益是可预期:有一条指名可走的通道、一份合规指引、一个可以在设计传输之前先去问的地方。

由此有两个习惯。第一,先画数据再设计系统。负面清单是按场景组织的,能用同一套场景语言描述自己传输的总部,会比带着一张数据架构图来的总部走得快。第二,合规文件跟着传输记录走。用了这条通道,就要能随时取出”传了什么、为什么传”的证据,因为这套制度靠的是文档,而不只是事前批准。

对外资集团来说,真正要紧的多是运营数据而非战略数据:订单、库存位置、供应商主数据、售后记录。这些恰好是供应链总部每天产生的类别,也解释了为什么数据条款会出现在这份总部文件里,而不是一份数字政策文件里。

人才:居留、签证与旅行卡

人才措施是给管区域链条的人写的,值得当成一个整体来读。

供应链总部可推荐纳入本市人才引进重点机构名单,随附落户、医疗保障、申请人才公寓等便利。企业也可为人才申报本市重点人才支持项目,以及为重点产业领域人才申请专项奖励。

对外籍员工,三条最要紧。外籍高级管理人员可推荐申办在华永久居留。涉外员工在合规审慎银行核定的免审单金额范围内办理薪酬购付汇或收结汇,免予提交单证。经常赴 APEC 经济体从事商务活动的人员,可申办 APEC 商务旅行卡——一张卡替代每次出行前的签证。

还有一条是给来访者而不是雇员的:总部需要邀请、需多次临时入境的外籍商务人员,可向公安出入境管理部门申请办理多次签证。对要接待供应商审核、培训和技访的区域总部,这条常常比居留许可更实用,因为需要它的人并不在工资表上。

四条的共性在于,都挂在”总部在名单上”这个前提上,没有一条可以只由个人申请取得。这使认定同时成为一项人事工具,也解释了为什么人力资源部门往往会在申报过程中成为意料之外的当事方。

动态管理与取消资格的四种情形

认定不是一劳永逸,文件把这一点写成了可操作的条款,而不是一句提醒。

市商务委对经认定的供应链总部实施动态管理,四种情形下可取消资格:企业主动申请退出;企业主要经营业务迁出上海;发生重大违法违规行为;经动态评估不再适合认定,且经限期整改仍未满足认定条件。

两侧还有两条义务。各区商务主管部门建立跟踪机制,对企业报告或发现可能造成企业不满足认定条件的情况,及时上报市商务委。总部自身发生经营事项调整和重大变更的,应主动向相关商务主管部门报告。

实务上的理解是:证书描述的是当下的状态,而不是一次授予的身份。集团若重组中国业务、把某项区域职能迁到别的城市,或更换持有授权的主体,就应当预期要重新过一遍材料。把牌子当成永久资产、从此不再维护底层证据的集团,最容易踩到第四种情形。

管理还有一层软性的。文件依托行业商会设立供应链总部专委会,日常开展上下游企业服务对接;市区两级商务部门建立沟通联络与服务机制,围绕出口管制合规管理、贸易风险防控、法律、金融服务等议题,每年至少组织一次新认定总部的培训;商务部门定期把新认定总部信息共享至海关、税务、外汇、人才等部门,各部门协同给予便利化服务。

对合规团队而言,需要常备更新的材料清单因此不长,但很具体:投资或授权证明、网络布局图、上年度财务报告、全球供应链业务情况报告。前三项在集团重组时变,第四项在业务变化时变。四项每年滚一遍的集团,很少会在重新评估时被意外。

与贸易型总部的关系:升级,不是另起炉灶

第十五条最该记进日历。原按《上海市鼓励企业设立贸易型总部的若干意见》认定的贸易型总部,按本意见实施管理;已认定但不满足供应链总部认定条件的,可保留原贸易型总部待遇,但只到 2027 年 8 月 31 日前。

由此有三点。第一,两套制度是接续的,不是并行的:新类别是一条升级路径,存量贸易型总部要么沿这条路走上去,要么为待遇终止做准备。第二,这是一个确定的日期,不是滚动宽限期,打算重新认定的集团,有一个明确的窗口去调整上海主体、网络与证据。第三,由于意见自印发之日起施行,倒计时从 2026 年 10 月 20 日就已开始,而不是从 11 月对外公开时起算。

这个设计与上海以往的总部政策一脉相承:先立更高一档,把低一档保留一段时间,让市场自己决定谁往上走。对集团来说,问题不是旧待遇今天够不够用,而是它所支撑的业务,到 2027 年夏天之后还够不够用。

外资集团应先核对什么

先看申报主体。对分支机构的投资证明或母公司授权,决定集团里哪家中国主体有资格——往往是控股公司,而不是背着营收的经营公司。两者分离时,授权关系要在材料组装之前落实,而不是之后补。

再看数字落在哪里。营业收入、国际贸易规模、平台运营收入、成交总额,都是按申请企业计算。习惯把贸易记在新加坡或香港枢纽的集团,可能会发现上海主体的数字比集团整体薄;这时的正确反应通常是讨论业务归集,而不是直接判定门槛够不着。

再看链条可见度。如果监管真的转向以链条为单元,便利程度将取决于总部能就供应商与分销商提供多少可验证的信息。已经在全链条上跑一套系统的集团,会比靠表格对账的集团更占先手,而这个差距不会在申报前几周内补上。

再看”1+N”:上游供应商与下游分销商,包括境外成员,能否一并纳入组团认证;若其中一个成员失守,合规义务由谁承担、如何切割。这个答案往往决定集团是把区域供应链的实际管理职能迁到上海,还是把这家主体当成记账中心留在原地。

最后,把几个日期放在一处:申报窗口、5 个工作日公示、2027 年 8 月 31 日的过渡截止,以及海关试点规则可能出台的时点。任何一个被忽略,总部决策通常都会在一年之内被重新提起。

还有一个顺序问题。申报表是全文最容易描述、也最难打乱顺序去做的一步。先拿表格开工的集团,时间会花在凑证据上;先回答三个问题——谁握有授权、数字落在哪一层、链条能证明什么——的集团,会发现表格是自己填满的。这个差别通常是四到六周,也就是”这一季递”与”下一季递”的差别。

这个顺序也解释了为什么这份意见值得整篇读,而不是只看门槛摘要。门槛回答的是”够不够格”;其余十五条回答的是”够格之后能做什么”,而其中几条——通关、数据、签证——只有当主体是按用得上它们的方式搭起来的,才真正起作用。

来源

  • 上海市商务委员会、中共上海市委金融委员会办公室、中华人民共和国上海海关、上海市经济和信息化委员会、上海市人才工作局、上海市数据局·《上海市关于鼓励企业设立供应链总部的若干意见》(沪商规〔2026〕4 号)全文:https://www.investsh.org.cn/cn/investsh/zctz/2026102813370022149406
  • 上海市外国投资促进中心·《”投资上海”一周资讯动态回顾(1109—1113)》:https://www.investsh.org.cn/cn/investsh/tzdt/2026111316344603129477
  • 澎湃新闻·《民生智库|商务领域动态热点速览(20261101—1106)》,2026-11-10:https://m.thepaper.cn/newsDetail_forward_34453134

Hainan’s Island-Wide Customs Closure: How Bonded Cargo Moves Between the First and Second Lines

  • Hainan issued Qiong Cai Shui [2026] No. 986 on 2 November 2026, effective on publication, to settle how goods move between areas inside the island once the island-wide closure is in operation.
    海南省三部门于 2026 年 11 月 2 日联合印发琼财税〔2026〕986 号,自公布之日起实施,用于明确全岛封关运作后货物在岛内不同区域之间流转的管理口径。
  • Bonded cargo held in a customs special supervision area, a bonded logistics centre or a bonded warehouse may move into other parts of the island by reference to first-line import, provided no operation has changed the state of the goods.
    存放于海关特殊监管区域、保税物流中心或保税仓库的保税货物,在未发生改变货物状态情形的前提下进入自贸港其他区域,可以参照”一线”进口办理。
  • Circulation, processing, testing, repair and blending are named as the operations that break the chain, and the list is deliberately open-ended.
    流转、加工、检测、维修、混兑被点名为会打断这条链条的操作,且该列举本身留有兜底空间。
  • Quarantine, inspection and licence verification already completed are not carried out a second time when the cargo moves on.
    此前已经完成的检疫、检验、许可证件核验等管理措施,货物继续流转时不再重复实施。
  • Goods made by encouraged-category enterprises in the port from imported inputs, with value-added processing of 30 per cent or more, follow the second-line processing value-added rules when sold on the island.
    自贸港内鼓励类产业企业使用进口料件生产、加工增值达到或超过百分之三十的货物,在岛内销售的,参照经”二线”进入内地的加工增值政策执行。
  • The notice introduces no new tariff schedule and no new list; it points operators back to closure policy documents that already exist.
    这份通知没有新增税则,也没有新增清单,它的作用是把经营者指回已经生效的封关政策文件。
  • Delivery sits with provincial departments and with municipal finance, development and reform, and commerce bureaux, under a standing risk-screening and reporting duty.
    落实责任落在省直单位与市县财政、发改、商务部门身上,并配有常态化的风险排查与上报机制。

Hainan’s Island-Wide Customs Closure: How Bonded Cargo Moves Between the First and Second Lines | 海南全岛封关:保税货物如何在”一线”与”二线”之间流转

Why this notice matters

On 2 November 2026 the Hainan Provincial Department of Finance, the Hainan Provincial Development and Reform Commission and the Hainan Provincial Department of Commerce jointly issued the Notice on Management Matters Concerning the Island-Wide Customs Closure Operation of the Hainan Free Trade Port, numbered Qiong Cai Shui [2026] No. 986. It took effect on the day it was published, which is to say there was no transition period and no separate implementing circular to wait for.

The notice rests on a working arrangement issued by the General Office of the National Development and Reform Commission together with other bodies, numbered Fa Gai Ban Qu Yu [2026] No. 754. That document set the task; this one discharges it for a specific and rather narrow set of movements. Nothing in the text creates a new tax instrument, a new catalogue of goods or a new eligibility test for investors. What it does is decide which existing rule applies when cargo crosses an internal boundary.

That is a smaller question than it sounds, and a more consequential one. An island-wide closure turns Hainan into a territory that is treated separately from the mainland customs territory for a defined set of purposes. Once that is true, every movement of goods acquires a direction: in from overseas, out to the mainland, or across from one part of the island to another. The first two directions were addressed when the closure framework was published. The third — internal movement — is what this notice settles.

For a foreign group running a bonded warehouse, a regional distribution centre or a light processing line on the island, the internal boundary is not an abstraction. Goods arrive from overseas and sit in bond. Some of them are later sold or transferred to a buyer elsewhere on the island: a hotel group, a retailer, a construction contractor, a manufacturer. Before this notice, an operator had to reason by analogy about whether that transfer counted as a fresh import, with all the documentary consequences that follow. The notice replaces analogy with a rule.

The document contains two substantive provisions and two housekeeping ones. The first substantive provision covers bonded goods leaving a bonded premises for another part of the island. The second covers goods manufactured on the island and sold on the island. The housekeeping provisions assign implementation duties and set out a risk-control mechanism. Everything else in this article follows from those four elements.

Because the notice is administrative rather than fiscal, it is easy to skip. There is no headline rate in it, no new list of eligible activities, nothing that would appear in an investment pitch. That is precisely why it repays a careful read. A group that has already decided to use Hainan as a staging point for regional distribution has, in practice, already accepted the fiscal framework; what it now needs is to know how goods move, and by what paperwork. Those are the questions this notice answers, and they are the questions that determine whether a warehouse runs smoothly or spends its days explaining itself.

What “first line” means for bonded cargo

The closure framework uses two directional terms. The first line is the boundary between overseas and the island. The second line is the boundary between the island and the mainland. Duty, licence and inspection questions cluster at those two lines, because that is where a customs territory begins or ends.

The notice’s first provision deals with a movement that crosses neither line. Goods come in from overseas and are placed in bonded storage inside one of three types of premises: a customs special supervision area, a bonded logistics centre or a bonded warehouse. Later, instead of being exported or sent to the mainland, they move to another area of the island. On the face of it, that is a domestic movement inside a single customs territory. But because the island has its own closure arrangements, the destination area may be governed by closure policy documents that assume goods have just arrived from overseas.

The notice resolves this by directing that such goods, when they enter other areas of the port, are handled by reference to first-line import and apply the requirements of the closure policy documents. The operative phrase is “by reference to” rather than “deemed to be”. The drafters are not reclassifying the goods or creating a legal fiction that they never entered. They are telling the operator and the local customs office which set of documents to work from.

The provision then closes the door on any wider reading. Goods other than those described continue to be managed under the rules currently in force. So the notice is not a general liberalisation of internal movement. It is a targeted answer for bonded cargo that has not been touched since it was stored.

It is worth pausing on the fact that all three premises types are named together. A customs special supervision area, a bonded logistics centre and a bonded warehouse occupy different positions in the supervision architecture, and an operator familiar with one does not automatically understand the others. By listing all three in a single clause, the notice signals that the operative test is bonded status rather than the shape of the enclosure. That is a useful signal for a group deciding where to site a new facility: the notice does not push the decision toward one form over another, and the choice can be made on commercial and logistics grounds.

There is a practical reason the answer had to be given in this form. Bonded storage exists precisely so that goods can be held without duty being settled, and the commercial logic of a distribution hub depends on being able to move stock to where the buyer is. If every internal transfer required the goods to be treated as a fresh import, the bonded model on the island would lose much of its point. If every internal transfer were simply waved through, the closure arrangements would be undermined. The notice picks a middle course: honour the bonded status, but only while the goods remain in the condition in which they were stored.

The three conditions that must all hold

The provision is written as a single sentence with three cumulative conditions. All three must be satisfied. Fail any one and the goods fall back into the ordinary rules.

The first condition concerns where the goods are. They must be held in bonded storage inside a customs special supervision area, a bonded logistics centre or a bonded warehouse. These are the three named premises types. A yard, a general warehouse outside supervision, or a factory store is not on the list. The condition is about the legal status of the storage location, not about the physical security of the site.

The second condition concerns what has happened to the goods. There must have been no circulation, processing, testing, repair, blending or any other circumstance that changes the state of the goods. The list is illustrative rather than exhaustive — the closing words “and other circumstances of any kind that change the state of the goods” make that explicit. So the test is functional: has the goods’ condition altered since it was stored? An operation that leaves the goods exactly as they were does not break the chain. An operation that changes them does, whether or not it appears in the list.

The third condition concerns where the goods are going. The destination must be another area of the free trade port. Movement to the mainland is not covered here; that is a second-line question governed by a different set of documents. Movement back overseas is an export. The notice is only about the internal leg.

Read together, the three conditions describe a fairly narrow corridor. Cargo enters from overseas, goes into bond in a recognised premises, is not touched, and is then transferred to another point on the island. That is the case the notice was written for. Anything more elaborate — a blend, a test, a repair, a repack that changes the goods — steps outside it and the operator must look to the general rules and to whatever specific policy covers the operation in question.

Operators should also note what the notice does not say. It does not require the goods to be sold, nor does it require a buyer to be identified in advance. It does not impose a time limit on how long the goods may sit in bond before the transfer. Those matters, if they are regulated at all, are regulated elsewhere.

Where the corridor ends: three situations

Abstract conditions become easier to apply when they are run against concrete operations. Three recur often enough in bonded practice to be worth spelling out. In each, the question is the same — has the goods’ state changed — but the answer is not always intuitive.

The first is blending. Liquid bulk, agricultural commodities and some chemical inputs are frequently stored in bond precisely so that different parcels can be combined to meet a buyer’s specification. Blending is named in the notice as a state-changing operation, and it is easy to see why: after blending, the consignment is no longer the consignment that entered. The customs identity of the goods has been altered, and the record that travelled with the original parcels no longer describes what is in the tank. An operator who blends in bond and then moves the result to another part of the island is outside the corridor, however clean the original clearance was.

The second is repair and testing. A bonded maintenance or quality-control operation takes goods that entered in one condition and returns them in another, or at least establishes something about them that was not established before. Both are named. This is the case that most often surprises operators, because repair and testing are services rather than manufacturing, and the commercial logic feels like an extension of storage. The notice does not draw that distinction. What matters is the effect on the goods, not the label on the contract.

The third is the consignment that is split. Here the answer depends on what splitting involves. If a pallet is broken down into cases and the cases are the same goods in the same condition, with only the handling unit changed, the argument that nothing has changed is available. If the split involves repacking into retail packaging, relabelling that alters the declared description, or any process that touches the goods themselves, the argument becomes much harder. The dividing line is the condition of the goods, and a pack that changes them falls on the far side.

There is a fourth situation that is not really an exception but is often confused with one: the goods leave the island for the mainland. That is a second-line movement, governed by the second-line documents, and the first provision of this notice has nothing to say about it. An operator who assumes that a clean bonded file automatically carries a consignment across the second line is reading the notice for more than it says.

What “no repeated inspection” saves

The second sentence of the first provision is short and, in day-to-day terms, probably the most valuable: where quarantine, inspection and verification of licence documents have already been completed, they are not carried out again.

Consider what that means for a consignment that entered Hainan from overseas, cleared quarantine and inspection at the first line, had its permits checked, and went into bond. When it later moves to another area of the island, the receiving end would ordinarily want to satisfy itself that the regulatory formalities had been met. Under the notice, it does not repeat them. The file travels with the goods.

The saving is measured in three currencies. Time is the obvious one: a second round of inspection means a second appointment, a second queue, a second release. Cost is the second: inspection fees where they apply, but more often demurrage, storage beyond the free period, and the labour of handling the consignment twice. The third is condition. Every additional handling event is a chance for damage, and for temperature-controlled cargo it is a chance for a break in the cold chain.

The wording matters as much as the benefit. The notice says the measures are not repeated; it does not say the goods are exempt. A measure that was never carried out in the first place is not covered, and a measure that was carried out incompletely is not cured. The relief assumes a clean file. Where the original clearance was partial, or where a permit was issued subject to a condition that has not yet been discharged, the operator should expect the gap to be closed rather than overlooked.

For that reason the operational discipline the notice implies is documentary. The relief is available to the operator who can show what was already done. Certificates, inspection records, release notices and permit numbers should travel with the consignment or be retrievable at short notice, and the internal transfer record should cross-reference them. A warehouse that cannot produce the first-line paperwork will find that the second-line or internal checkpoint simply starts again from the beginning.

The record that makes the rule work

Both substantive provisions in the notice are conditional, and both conditions are proved by documents. That is not an accident of drafting. A rule that grants a benefit on the basis that something has not happened, or that something has already been done, can only be administered against a file.

Three records carry most of the weight. The first is the bonded entry record: what entered, when, from where, in what quantity, under which declaration, and with which permits. This is the document that establishes the starting condition against which everything later is measured. The second is the state log: an internal record, kept by the operator, of everything done to the consignment while it was in bond. Its value is that it is contemporaneous. A log written at the end of the storage period, reconstructing what happened, is worth much less than one written as events occurred. The third is the transfer note: the document that links the origin premises to the destination area, names the date, and cross-references the first two.

Where the notice’s relief depends on prior measures having been completed, the entry record and the permits it cites do the work. Where the relief depends on nothing having been done, the state log does it, supported by the absence of any processing, testing or repair instruction in the operator’s own systems. Both are more persuasive when they were created before anyone asked.

There is a further, quieter benefit to keeping these records well. The risk-screening mechanism described in the notice draws on information shared across departments. A consignment whose bonded record, state log and transfer note agree is easy to clear and uninteresting to a risk model. A consignment where one of the three is missing or inconsistent invites a question, and a question at the point of movement is exactly what the notice is designed to avoid.

The second-line rule for island sales

The notice’s second substantive provision moves from storage to manufacturing. Where an encouraged-category industrial enterprise in the port produces goods that contain imported materials and components, and where processing on the island has raised the value-added by 30 per cent or more, those goods, when sold within the island, are handled by reference to the processing value-added policy documents that apply to second-line entry into the mainland.

Three elements have to be present. The producer must be an enterprise in an encouraged industry — the classification that determines whether the closure’s industrial preferences apply at all. The goods must contain imported materials or components; entirely domestic input does not engage the rule. And the value added by processing on the island must reach or exceed 30 per cent. That threshold is the same one used for second-line entry to the mainland, and the notice borrows the machinery rather than inventing a parallel test.

What changes is the destination. Until now, the processing value-added rule was associated with goods leaving the island for the mainland. The notice extends the same reference point to goods that stay: a sale to a buyer on the island is now handled as though the goods had gone through the second line, for the purposes of the processing value-added policy documents. For a manufacturer whose customers are Hainan hotels, retailers, builders or public projects, that removes a gap between the export-facing and domestic-facing sides of the business.

The provision is drafted as a reference, not a grant. It says the processing value-added policy documents apply by reference; it does not restate the treatment those documents provide, and it does not create an independent entitlement. An operator who wants to know the actual fiscal outcome of an island sale must therefore read the underlying processing value-added documents, and must be able to substantiate the 30 per cent calculation in the way those documents require.

The first of the three elements deserves separate attention, because it is the one most often assumed rather than checked. Encouraged-category status is a classification of the enterprise’s business, not a description of what it makes on any given day. A company that sits inside the catalogue for one product line may find that a second line falls outside it. The notice does not extend the reference to island sales to enterprises that are not in the encouraged category, and it does not extend it to goods that contain no imported materials at all. Both points should be settled before a pricing decision is built on the assumption that the rule applies.

The commercial consequence of the reference is easiest to see at the invoicing stage. A sale to a buyer on the island that relies on the notice should be able to point to the calculation behind it — the imported input values, the processing costs incurred on the island, and the resulting proportion — in a form that a reviewer can follow without reconstructing the business. Where the same product is also sold to the mainland, keeping one methodology across both channels is simpler than maintaining two, and it reduces the risk of the two sets of figures drifting apart.

That calculation is where most of the work sits. Value-added is not a matter of assertion; it is a computation over input values, processing costs and output values, and it has to survive scrutiny. Enterprises that already run the calculation for mainland-bound sales will have the methodology in place and can extend it to island sales. Enterprises that have never had to compute it should build the model before, not after, the first island invoice is issued.

Who implements, who supervises

The notice divides implementation between two levels. Provincial departments carry out their parts according to their assigned responsibilities. Below them, the finance, development and reform, and commerce departments of the cities and counties exercise territorial duties: interpreting the policy, publicising it, and providing service support to enterprises in their area.

That division matters for anyone seeking guidance. The provincial text is short and leaves the operational detail to the levels that deal with individual consignments and individual enterprises. A question about whether a particular warehouse qualifies, or how a particular transfer should be documented, is most likely to be answered by the municipal or county bureau rather than by the provincial drafters.

Alongside implementation sits a risk-control mechanism with three layers. The first is routine screening: risks are to be identified and assessed on an ongoing basis rather than in response to incidents. The second is analysis, judgement and early warning — the machinery for spotting a pattern before it becomes a loss. The third is prevention and handling, which is where an identified risk is actually dealt with.

The notice also requires information sharing and coordinated action across departments, which is the practical precondition for any of the above. A closure regime that depends on knowing what is in bond, where it came from and where it went cannot function if each department holds a fragment of the picture. And where a major risk or hidden danger is identified, it must be reported immediately through the prescribed procedures and channels. The emphasis on immediacy, and on using the established channels rather than ad hoc ones, is worth noting: it signals that escalation is expected to be fast but orderly.

For operators, the supervisory architecture has a straightforward implication. The reliefs in the notice are premised on the goods being what the file says they are. A risk-screening system that works across departments will notice a consignment whose bonded record and internal transfer record do not match. The compliance posture that makes the notice useful is therefore the same posture that makes an operator uninteresting to a risk model: complete records, consistent declarations, and no gaps between what was stored and what moved.

The two-level split also has a timing consequence worth planning around. Provincial documents arrive first and set the principle; local practice develops afterwards, city by city. A group with facilities in more than one city should expect that the same transfer may be handled with slightly different expectations about paperwork in each place, at least in the early period, and should build its internal standard to the more demanding of the two rather than to the more convenient. Where a group runs a single hub, the relevant standard is the one applied where that hub sits, and it is worth confirming in writing before the first transfer rather than after.

What to ask the local bureau

Because the notice is short and the implementing levels are close to the cargo, most of the practical work happens in a conversation rather than in a reading of the text. Five questions are worth putting in writing, early, to the finance, development and reform, or commerce bureau of the city or county where the premises sit.

First, confirm the classification of the premises. The notice names three types, and the treatment of a transfer depends on which one the goods are in. A warehouse that has been described loosely in internal documents should be re-checked against its formal status before a transfer is planned around it.

Second, describe the intended operation and ask whether it is regarded as changing the state of the goods. This is the question on which most cases turn, and it is better asked in advance, with the process flow attached, than discovered at the point of movement. A written answer, even an informal one, is worth keeping.

Third, ask what documentation the local level expects for an internal transfer. The notice sets a principle; the form of the transfer note, the evidence expected for the state log, and the way prior inspection results should be referenced are matters that will be settled locally.

Fourth, ask whether the destination area has requirements of its own. The notice governs the movement by reference to first-line import and the closure policy documents; the area receiving the goods may have its own operational expectations, and a transfer that satisfies the origin may still need something at the destination.

Fifth, ask how a discrepancy should be handled. If the file and the cargo do not match — a missing permit, an unrecorded handling operation — the useful thing to know in advance is whether the matter can be regularised and how, rather than discovering the answer while the consignment is standing still.

None of these questions has a single national answer. They are local questions about a local notice, and the operators who get the most out of it will be the ones who asked them before the first movement rather than after.

Practical checklist for bonded operators

  • Confirm which of the three premises types your goods actually sit in, and keep the documentary proof of that status. A bonded warehouse, a bonded logistics centre and a customs special supervision area are distinct, and the notice names all three but does not treat them as interchangeable in every other respect.
  • Map the movement before it happens. Identify the origin premises, the destination area and the route, and record the date. The notice governs a movement to another area of the port, so a movement to the mainland or back overseas needs a different file.
  • Keep a state-of-goods log. If nothing has been done to the consignment since storage, say so in writing and be able to show it. If something has been done, classify it against the list in the notice — circulation, processing, testing, repair, blending, or anything else that changes the goods — and assume the exception applies until you have confirmed otherwise.
  • Pull the first-line file forward. Quarantine records, inspection results and licence verification should be indexed and linked to the internal transfer, because the “no repetition” benefit depends on being able to prove what was already done.
  • For island sales, build the value-added model early. Confirm encouraged-industry status, isolate the imported content, and document the calculation that takes you to 30 per cent or above, in the format the processing value-added documents expect.
  • Read the underlying closure documents rather than the notice alone. The notice is a pointer; the treatment it points to lives in the policy documents it references.
  • Watch for follow-up guidance from provincial and municipal departments, particularly on documentation standards for internal transfers. The notice sets the principle; the forms will come from the implementing levels.

None of this requires a new business model. It requires that the file matches the cargo, and that the operator knows in advance which of the notice’s corridors a given movement falls into. For a distribution hub, that is a modest administrative discipline in exchange for a rule that no longer has to be inferred.

It is also worth reading the notice for what it reveals about how the closure will be administered. A framework that is willing to answer a narrow question precisely, by reference to documents already in force rather than by inventing a parallel regime, suggests that subsequent questions will be handled the same way: one movement type at a time, in short documents, with the operational detail left to the levels that see the cargo. Operators who keep their files in order and their questions specific will find that pattern easier to work with than one that produces a single comprehensive code and then amends it.

Sources

  • 海南省人民政府,《关于海南自由贸易港全岛封关运作有关管理事项的通知》(琼财税〔2026〕986 号),2026-11-02。https://www.hainan.gov.cn/hainan/0101/202611/926377e2319c47eb8ff62e034689d5a0.shtml
  • 海南省人民政府,《海南自由贸易港全岛封关运作有关管理事项政策解读》,2026-11。https://www.hainan.gov.cn/hainan/tingju/202611/9a26d8be2ecf4cada93fd4ea210a6809.shtml
  • 三亚市中级人民法院(转载),《海南自由贸易港全岛封关运作有关管理事项通知及政策解读》,2026-11。https://syzy.hicourt.gov.cn/article/detail/2026/11/id/9560564.shtml
  • 中国经济网,《海南明确全岛封关运作有关管理事项》,2026-11-03。https://www.ce.cn/xwzx/gnsz//gdxw/202611/t20261103_3281236.shtml

海南全岛封关:保税货物如何在”一线”与”二线”之间流转

这份通知为什么值得看

2026 年 11 月 2 日,海南省财政厅、海南省发展和改革委员会、海南省商务厅联合印发《关于海南自由贸易港全岛封关运作有关管理事项的通知》,文号为琼财税〔2026〕986 号。通知自公布之日起实施,没有过渡期,也没有另一份实施口径需要等待。

这份通知的依据,是国家发展改革委办公厅等部门印发的《关于做好海南自由贸易港全岛封关运作有关工作的通知》,文号为发改办区域〔2026〕754 号。后者布置任务,前者就一组相当具体的货物流转作出交代。通知全文没有创设新的税种、新的商品目录,也没有为投资者设置新的资格条件。它做的是一件很窄的事:当货物跨越岛内的一条内部边界时,决定适用哪一套现成规则。

这个问题听起来不大,实际分量不轻。全岛封关之后,海南在一组被明确界定的用途上,被当作与内地关境相区别的区域对待。一旦如此,货物的每一次移动都有了方向:从境外进来、往内地出去,或者在岛内从一个区域转到另一个区域。前两个方向在封关制度框架公布时已有交代,第三个方向——岛内流转——正是这份通知落笔的地方。

对在岛上经营保税仓库、区域分拨中心或轻加工线的外资集团来说,这条内部边界不是抽象概念。货物从境外运来,存放在保税状态;其中一部分后来卖给或转给岛内其他区域的买方:酒店集团、零售商、施工承包商、制造商。在通知出台之前,经营者只能靠类推去判断这次转移算不算一次新的进口,以及随之而来的单证后果。通知用一条规则替换了类推。

文件包含两条实体规定和两条事务性规定。第一条实体规定管的是保税货物离开保税场所、进入岛内其他区域。第二条管的是在岛上生产、在岛上销售。事务性规定一条分配落实责任,一条布置风险防控机制。本文余下的内容,都从这四个要素展开。

正因为这份通知是管理性的、而非财税性的,它很容易被跳过。里面没有醒目的税率,没有新的可享受优惠的业务目录,也没有哪句话能写进招商材料。恰恰因为如此,它值得细读。一个已经决定把海南当作区域分拨中转点的集团,实际上已经接受了那套财税框架;它此刻需要知道的,是货怎么走、凭哪些单证走。这些问题正是通知回答的问题,也正是决定一间仓库是顺畅运转、还是整天忙于解释自己的问题。

“一线”对保税货物意味着什么

封关制度使用两个方向性概念。”一线”是境外与海南之间的边界,”二线”是海南与内地之间的边界。关税、许可证件、检验检疫的问题都聚拢在这两条线上,因为关税区的起点与终点在那里。

通知的第一条实体规定,处理的却是一次两条线都不跨越的移动。货物从境外进来,存入三类场所之一并处于保税状态:海关特殊监管区域、保税物流中心、保税仓库。此后,它既没有复运出境,也没有进入内地,而是转到岛内另一个区域。表面看,这是同一关税区内的国内移动;但由于全岛封关有自己的制度安排,目的地所在区域可能适用那些”假定货物刚从境外抵达”的封关政策文件。

通知的解法是:这类货物后续进入自贸港其他区域的,参照”一线”进口办理,适用封关政策文件的要求。要害在”参照”二字,而不在”视同”。起草者没有给货物重新定性,也没有拟制一个”从未入境”的法律事实;他们只是告诉经营者和属地海关,该按哪一套文件办。

紧接着,通知把口子收住:上述情形以外的货物,继续按现行规定管理。所以这不是对岛内流转的普遍放开,而是针对”入区后未被触动过的保税货物”给出的一条定向答案。

三类场所被并列写进同一句,这一点值得停一下。海关特殊监管区域、保税物流中心、保税仓库在监管架构里所处的位置并不相同,熟悉其中一类,并不等于自动理解另外两类。通知把三者一并列出,等于说明:起作用的判断标准是保税状态本身,而不是围栏长什么样。对一个正在决定把新设施设在哪里的集团来说,这是个有用的信号——通知并不把选址往某一种形态上推,选择可以按商业与物流的需要来做。

答案必须写成这个样子,有它的现实原因。保税仓储存在的意义,本就是让货物在不结税的状态下被持有;分拨中心的商业逻辑,也依赖于能把库存挪到买方所在之处。如果每一次岛内转移都要当作一次全新进口,岛上的保税模式就失去了大半意义;如果每一次岛内转移都直接放行,封关安排又会被架空。通知走的是中间路线:承认保税状态,但只在这批货物仍保持入库时状态的前提下承认。

必须同时成立的三项条件

这条规定写成一个长句,里面叠着三项条件,三者必须同时满足。任何一项不成立,货物就回到一般规则里去。

第一项条件关乎货物在哪里。它必须存放在海关特殊监管区域、保税物流中心或者保税仓库,处于保税仓储状态。这是被点名的三类场所。堆场、监管区之外的一般仓库、工厂的自用料仓,都不在名单上。这项条件看的是存放场所的法律属性,不是场地的物理安保水平。

第二项条件关乎货物经历过什么。不得发生流转,也不得发生加工、检测、维修、混兑等各类改变货物状态的情形。这份列举是示例性的,收尾的”等各类改变货物状态情形”把话说明白了。所以判断标准是功能性的:入库之后,货物的状态有没有被改变?没有改变状态的操作,不打断这条链条;改变了状态的操作,无论是否出现在列举里,都会打断它。

第三项条件关乎货物去哪里。目的地必须是自贸港内的其他区域。运往内地不在此列,那是”二线”的问题,由另一组文件管;复运出境属于出口。通知只管岛内这一段。

三项条件合起来,划出的是一条相当窄的通道:从境外进境,进入被认可的场所保税存放,期间未被触动,随后转到岛上另一点。这就是通知为之而写的情形。更复杂的动作——混兑、检测、维修、改变货物形态的再包装——都落在通道之外,经营者需要回到一般规则,并查看针对该业务的具体政策。

还有一点值得留意:通知没有要求货物必须已经售出,也没有要求事先确定买方;它没有对货物在保税状态下可以存放多久才发生转移设定时限。这些事项如果另有规定,规定在别处。

通道在哪里断开:三种情形

抽象条件放到具体作业里,才更容易用。保税实务中有三种情形出现频率足够高,值得单独说清。三种情形面对的是同一个问题——货物状态有没有改变——但答案并不总是直觉能给的。

第一种是混兑。液体散货、农产品以及部分化工原料之所以常年在保税状态下存放,正是为了让不同批次的货能够合并,达到买方要求的规格。通知把混兑列为改变货物状态的操作,理由也直观:混兑之后,这一票货不再是入库时那一票货。货物的海关身份被改写了,随原始批次而来的那套记录,已经不能描述罐里现在装着什么。在保税状态下完成混兑、再把结果转到岛内其他区域的经营者,落在通道之外,无论原始验放多干净。

第二种是维修与检测。保税维修或质量控制作业,把入库时某一状态的货物,以另一状态交还,或者至少确立了此前未曾确立的某种结论。这两项都在列举之中。这类情形最容易让经营者意外,因为维修与检测在商业上属于服务,而不是制造,感觉像是仓储的延伸。通知并不作这个区分。起作用的是对货物的实际影响,不是合同上的名目。

第三种是拆票。答案取决于拆的是什么。把整板拆成箱,箱里还是同一批货、同一状态,改变的只是搬运单元,那么”什么也没变”这个说法是站得住的。如果拆的过程中重新包装成零售规格、更换标签以致申报品名发生变化,或者任何触及货物本身的处理,这个说法就难得多。分界线在货物状态,改变状态的包装落在另一侧。

还有第四种情形,严格说不是例外,却常被误当成例外:货物离开海南运往内地。那是一次”二线”移动,由”二线”文件管,通知的第一条实体规定对此不作任何交代。以为一份干净的保税档案就能自动把货物送过”二线”的经营者,是从通知里读出了它没有写的东西。

“不再重复实施”省下的是什么

第一条实体规定的第二句很短,但在日常操作层面可能是最值钱的一句:此前已经完成检疫、检验、许可证件核验等管理措施的,不再重复实施。

设想一票货物从境外进入海南,在”一线”完成了检疫与检验,许可证件也核验过了,随后入库保税。当它后来转到岛内另一区域时,接收端通常想自行确认监管手续是否齐备。按通知,它不再重复做一遍——文件跟着货物走。

省下来的东西有三种计价方式。时间最直观:第二轮检验意味着第二次预约、第二次排队、第二次放行。成本其次:确有检验费用的场合是费用,更常见的是滞箱费、免堆期之外的仓储费,以及把货物经手两遍的人工。第三是货物状态:每多一次搬倒,就多一次货损机会;对温控货物而言,还多一次冷链中断的机会。

措辞与好处同样重要。通知说的是”不再重复实施”,不是”免于实施”。一开始就没做的措施,不在覆盖范围内;做得不完整的措施,也不会因此补齐。这项便利的前提是档案干净。原始验放本身是部分的,或者许可证件附带了尚未履行的条件,经营者应当预期缺口会被补上,而不是被忽略。

正因如此,通知隐含的操作纪律是单证纪律。便利属于能够证明”此前做过什么”的经营者。证书、检验记录、放行通知、许可证件编号,应当随货同行,或者能在短时间内调取;岛内转移记录应当与这些材料互相勾连。拿不出”一线”手续的仓库,会发现下一道关口只能从头再来。

让规则运转起来的那份档案

通知的两条实体规定都是附条件的,而两个条件都要靠单证来证明。这不是起草上的偶然。一条以”某事没有发生”或者”某事已经做过”为由给予便利的规则,只能对着档案来执行。

三份记录承担了大部分重量。第一份是保税入库记录:进了什么、何时、从哪来、数量多少、凭哪份申报、附哪些许可证件。这份文件确立了此后一切判断所依据的初始状态。第二份是状态台账:经营者自己留存、记载货物在库期间被动过什么的内部记录。它的价值在于同步——在存储期末尾凭回忆补写的台账,价值远低于随事件发生逐笔写下的台账。第三份是转移单:把起运场所与目的区域连起来、写明日期、并与前两份互相引用的文件。

通知的便利若依赖于”此前措施已完成”,起作用的是入库记录及其引用的许可证件;若依赖于”什么也没做”,起作用的是状态台账,再加上经营者自身系统里确实找不到任何加工、检测或维修指令这一事实。两者在没人追问之前就已形成,说服力最强。

把这些记录做扎实,还有一层不声张的好处。通知所描述的风险排查机制,依赖跨部门共享的信息。保税记录、状态台账、转移单三者一致的货物,验放顺畅,对风险模型而言也不显眼。三者之中缺一份、或者彼此对不上,就会招来一个问题,而在移动环节被问问题,正是通知想要避免的事。

岛内销售的二线规则

通知的第二条实体规定,从仓储转向制造。自贸港内属于鼓励类产业的企业,生产的货物含有进口料件,且在自贸港内加工增值达到或者超过百分之三十的,这些货物在岛内销售的,参照经”二线”进入内地所适用的加工增值政策文件要求办理。

三个要素必须同时具备。生产者必须是鼓励类产业企业——这个归类决定了封关的产业优惠是否适用。货物必须含有进口料件,纯境内料件不触发这条规则。加工增值必须达到或超过百分之三十。这个门槛与”二线”进入内地所用的门槛是同一个,通知借用现成机制,而不是另造一套并行的算法。

变化发生在目的地这一端。此前,加工增值规则与”货物离开海南运往内地”联系在一起;通知把同一个参照点延伸到了留在岛内的货物:向岛内买方销售,在加工增值政策文件的意义上,按已经过”二线”处理。对客户是海南本地酒店、零售商、建筑商或公共项目的制造商来说,这抹平了外向业务与内销业务之间的一道缝。

条文写的是”参照”,不是”授予”。它说加工增值政策文件参照适用,没有复述那些文件给的是什么待遇,也没有创设一项独立的权利。因此,想知道一次岛内销售的实际财税结果,必须去读底层的加工增值文件,并且必须按那些文件要求的方式,把百分之三十的测算做实。

三个要素中的第一个,值得单独提一句,因为它最常被默认,而最少被核对。鼓励类产业身份是对企业业务的归类,不是对某一天在产什么的描述。一家企业在某条产品线上落在目录之内,另一条产品线未必也在。通知没有把参照适用延伸到不属于鼓励类产业的企业,也没有延伸到完全不含进口料件的货物。这两点在依据”规则适用”作出定价决定之前,都应当先落定。

这项参照在商业上的后果,最容易在开票环节看出来。一笔依赖通知的岛内销售,应当能够指向它背后的测算——进口料件价值、在岛上发生的加工成本、由此得出的比例——并且以复核者不必重建整个业务就能看懂的形态呈现。同一产品如果同时销往内地,两个渠道共用一套方法,比维护两套更简单,也能减少两组数字日后走样的风险。

测算才是大部分工作量所在。增值不是一句声明,它是对投入价值、加工成本与产出价值的计算,并且要经得起核查。已经为内销内地做过这套测算的企业,方法学是现成的,扩用到岛内销售即可;从未算过的企业,应当在开出第一张岛内发票之前把模型搭起来,而不是之后。

谁来落实,谁来监督

通知把落实分成两层。省直单位按职责分工落实各自那部分。其下,市县财政、发改、商务部门履行属地职责:做好政策解读、宣传引导和服务保障。

这个分工对寻求指引的人有意义。省级文本很短,操作细节留给直接面对具体货物、具体企业的层级。某个仓库是否属于那三类场所、某次转移应当怎样留痕,这类问题多半由市县部门而不是省级起草者回答。

与落实并行的是一套三层风险防控机制。第一层是常态化排查:风险要持续识别、持续评估,而不是等出了事再回头看。第二层是研判预警:在苗头变成损失之前发现规律的机制。第三层是防控处置:已经识别的风险如何实际处理。

通知还要求跨部门信息共享与协同联动,这是上述任何一层能够运转的现实前提。一个依赖”知道什么在保税、从哪来、到哪去”的封关制度,如果每个部门只握着一块碎片,是转不动的。而一旦发现重大风险隐患,必须第一时间按规定程序和渠道上报。”第一时间”与”按规定渠道”并提,值得注意:它表明升级要快,但也要有序。

对经营者而言,这套监管架构的含义很直接。通知给的便利,前提是货物与档案一致。一个跨部门运转的排查系统,会注意到保税记录与岛内转移记录对不上的那票货。让这份通知变得有用的合规姿态,恰恰也是让经营者对风险模型而言”不显眼”的姿态:记录完整、申报一致、存放与移动之间没有缺口。

两级分工还带来一个时序上的后果,值得提前安排。省级文件先到,定下原则;地方做法随后形成,且是一地一地形成。在不止一个城市设有设施的集团,应当预期同一类转移在各地对单证的要求会略有差别,至少在初期如此;内部标准宜按两者中更严的那个来定,而不是按更省事的那个。只有一个枢纽的集团,标准就是枢纽所在地执行的那一套,而这套标准最好在第一次转移之前书面确认,而不是之后。

该向属地部门问什么

通知很短,落实层级又贴近货物,所以实务工作多半发生在一场对话里,而不是在文本阅读里。有五个问题值得早早用书面方式,提给场所所在市县的财政、发改或商务部门。

第一,确认场所的归类。通知点了三类,而一次转移如何对待,取决于货物身处哪一类。内部文件里一直写得含糊的仓库,在围绕它安排转移之前,应当对照其正式状态重新核一遍。

第二,把打算做的作业描述清楚,问它是否被视为改变货物状态。多数个案就卡在这个问题上,提前带着工艺流程图去问,好过在移动环节才发现。得到的答复,哪怕是口头的,也值得留档。

第三,问属地层级对岛内转移要求哪些单证。通知给的是原则;转移单的格式、状态台账要留哪些证据、此前的检验结果应当如何引用,这些由地方定。

第四,问目的区域是否有自己的要求。通知是参照”一线”进口和封关政策文件来管这次移动的;接收地可能另有操作上的期待,满足了起运地的要求,未必等于在目的地也已齐备。

第五,问出现不一致时怎么处理。档案与货物对不上——缺一份许可证件、漏记一次处理作业——事先知道能否补正、怎么补正,比等货物停在那里时才知道答案要有用得多。

这些问题没有全国统一的答案。它们是关于一份地方文件的地方性问题;把这份通知用足的人,多半是在第一次移动之前就问过,而不是之后。

保税经营者的实操清单

  • 先确认货物究竟存放在哪一类场所,并留存能证明该状态的文件。保税仓库、保税物流中心、海关特殊监管区域是三种不同存在,通知三者都点名,但在其他方面并不等同视之。
  • 转移之前先画路线。标明起运场所、目的区域、路径,并记录日期。通知管的是进入自贸港其他区域的移动;运往内地或复运出境,需要另备一套文件。
  • 建立”货物状态台账”。入库后未做任何处理,就书面写下来并能举证;做过处理,就对照通知的列举归类——流转、加工、检测、维修、混兑,或者任何改变货物状态的其他操作——在确认不属于例外之前,先按例外对待。
  • 把”一线”档案往前调。检疫记录、检验结果、许可证件核验材料应当编好索引,与岛内转移记录相互关联,因为”不再重复”这项好处,取决于能否证明此前做过什么。
  • 涉及岛内销售的,尽早搭好增值模型。确认鼓励类产业身份,单独剥离进口料件价值,并按加工增值文件要求的口径,把测算到百分之三十以上的过程记录下来。
  • 读底层的封关文件,而不只是读这份通知。通知是路标,它指向的待遇写在被它引用的政策文件里。
  • 留意省市两级后续的操作指引,尤其是岛内转移的单证标准。通知给的是原则,表格会从落实层级出来。

以上都不要求改变商业模式,只要求档案与货物对得上,并且经营者事先清楚:某一次具体移动,落在这份通知划出的哪一条通道里。对一个分拨中心来说,这点行政纪律换来的,是一条不必再靠推测的规则。

这份通知也值得从它透露的施政方式去读。一套愿意把窄问题答准、并且靠引用现行文件而不是另造并行制度来作答的框架,暗示后续问题会以同样的方式处理:一次一类移动,用不长的文件,操作细节留给看得见货物的层级。把档案理清、把问题问得具体的经营者,会发现这种路数比”先出一部大而全的法规、再反复修订”更好用。

来源

  • 海南省人民政府,《关于海南自由贸易港全岛封关运作有关管理事项的通知》(琼财税〔2026〕986 号),2026-11-02。https://www.hainan.gov.cn/hainan/0101/202611/926377e2319c47eb8ff62e034689d5a0.shtml
  • 海南省人民政府,《海南自由贸易港全岛封关运作有关管理事项政策解读》,2026-11。https://www.hainan.gov.cn/hainan/tingju/202611/9a26d8be2ecf4cada93fd4ea210a6809.shtml
  • 三亚市中级人民法院(转载),《海南自由贸易港全岛封关运作有关管理事项通知及政策解读》,2026-11。https://syzy.hicourt.gov.cn/article/detail/2026/11/id/9560564.shtml
  • 中国经济网,《海南明确全岛封关运作有关管理事项》,2026-11-03。https://www.ce.cn/xwzx/gnsz//gdxw/202611/t20261103_3281236.shtml

China’s Segmented Pharmaceutical Production Rules Move From Pilot to Standing Procedure

  • The National Medical Products Administration issued a consultation draft on 13 November 2026 that would convert segmented pharmaceutical production from a case-by-case pilot into a standing procedure, with comments due by 30 November 2026.
    国家药监局于 2026 年 11 月 13 日发布征求意见稿,拟把药品分段生产从”一事一议”的试点转为常规程序,反馈截止 2026 年 11 月 30 日。
  • The draft runs to six parts and twenty-three articles and rests on the revised Regulation for the Implementation of the Drug Administration Law, in force since 15 May 2026, while the biologics pilot it supersedes runs out on 31 December 2026.
    征求意见稿共六部分二十三条,上位依据是 2026 年 5 月 15 日起施行的新修订《药品管理法实施条例》;它所接替的生物制品试点将于 2026 年 12 月 31 日到期。
  • Four categories of product may be split across sites, and the fourth — an originator drug already approved in China but made abroad, with some or all steps moved onshore — is the new addition.
    可分段生产的品种共四类,其中第四类为新增:已取得境内上市许可、在境外生产、部分或全部工序转移至境内分段生产的原研药品。
  • At least one side of a sterile arrangement must hold three years of commercial sterile manufacturing experience in the same dosage form, and only two of the four relaxations listed in the 2025 No. 134 announcement are carried over.
    无菌药品安排中,委托方与受托方至少一方须具备同剂型三年以上商业化生产经验;2025 年第 134 号公告所列四项放宽情形中,仅两项被沿用。
  • Cross-border commissioning opens in one direction only: a foreign Marketing Authorization Holder may move steps into China, while a domestic holder may not in principle commission manufacturing abroad, save for the fourth category.
    跨境委托只单向开放:境外持有人可以把工序放到境内,境内持有人原则上不得委托境外生产,第四类原研药除外。
  • Antibody-Drug Conjugate intermediates that are not managed as drugs or as active pharmaceutical ingredients are excused from holding a drug manufacturing licence, but still face GMP compliance, inspection and a licence annotation.
    抗体偶联药物中不按药品或原料药管理的中间产品,可免于办理药品生产许可证,但仍须符合 GMP、接受检查并在许可证上载明。
  • The obligations land squarely on the holder: two or more technical staff stationed on site for each product in production, one full-system GMP audit of every site each year, and a quality risk review chaired by the legal representative or the enterprise head.
    义务落在持有人身上:每个在产品种驻厂技术人员不少于两名,每年对全部生产场地至少开展一次全体系 GMP 审核,并由法定代表人或企业负责人主持质量风险会商。

China’s Segmented Pharmaceutical Production Rules Move From Pilot to Standing Procedure | 中国药品分段生产:从试点走向常规程序

Why this matters now

On 13 November 2026 the general office of the National Medical Products Administration published a consultation draft titled Announcement on Matters Concerning Segmented Production of Entrusted Drugs. Comments close on 30 November 2026 and go to a dedicated mailbox. The text is a draft. Nothing in it binds anyone until the final version is issued, and the final version may differ in wording, in scope, and in the transitional arrangements it attaches to products already in the pipeline.

What makes the draft worth reading now rather than in December is the calendar behind it. The segmented production pilot for biologics opened in October 2024 and closes on 31 December 2026. The revised Regulation for the Implementation of the Drug Administration Law has been in force since 15 May 2026. Between those two dates sits a gap: the Regulation provides the legal basis for splitting manufacturing across sites, but the operational route has been a pilot run on a case-by-case basis, with entry negotiated rather than published. The draft fills that gap by writing the criteria down.

Segmented production, in plain terms, means one product whose manufacturing steps are carried out at more than one site, sometimes by more than one legal entity, under a single quality assurance system that the marketing authorization holder owns. It is not outsourcing in the ordinary sense. The holder does not hand over responsibility along with the process; it keeps the responsibility and distributes the work. That distinction drives almost every clause in the draft.

For a foreign group, three features of the draft stand out. The first is the fourth eligible category, which for the first time gives an approved originator product made overseas a route to move some or all of its steps into China without re-registering as a new drug. The second is the direction of the cross-border rule: the door swings inward and, with one exception, not outward. The third is the carve-out for certain antibody-drug conjugate intermediates, which relieves a specific and expensive licensing burden without relieving the quality obligations that sit around it.

Each of those has a commercial consequence. A company that has been running a global supply chain with China as a finishing and packaging site can now contemplate the reverse: bulk or intermediate manufacture onshore, with the regulatory footprint that follows. A company that has been waiting for the pilot to become permanent can plan investment against published criteria rather than against a negotiating position. And a company in the antibody-drug conjugate space can stop treating a mature small-molecule intermediate as if it needed a drug manufacturing licence, provided it can satisfy the conditions attached.

The draft also narrows as it clarifies. New applicants must be the holder’s own development or a development the holder participated in end to end. The first category has no post-approval change route, which means a product that is already on the market and falls only within that category cannot use this announcement to restructure its manufacturing after the fact. Reading the draft as a general liberalisation would misread it. It is a procedural consolidation with a deliberately bounded scope.

From pilot to standing procedure

The pilot that the draft replaces was launched in October 2024 for biologics. Its logic was experimental: take a small number of products, work out how a split manufacturing chain could be supervised, and learn from the cases. That model has real advantages when the regulator is still deciding what the standards should be. It has an obvious disadvantage when a company needs to commit capital: entry depends on a conversation, not on a rule, and the outcome of the conversation is not published.

A standing procedure changes the nature of the decision. Instead of asking whether a product might be allowed, a company asks whether it meets published conditions. The answer becomes auditable, and a negative answer becomes something that can be addressed — by changing the product scope, by changing the site configuration, or by choosing a different route.

The draft is structured as six parts and twenty-three articles. The parts follow the life of an application rather than the structure of the statute: general provisions, eligible products, the application and approval route, the holder’s obligations, cross-border arrangements, and supplementary provisions. That ordering is itself informative. Cross-border arrangements sit in a part of their own, near the end, which reflects the fact that they are treated as a distinct policy question rather than as one more variation on a domestic application.

The legal foundation is the revised Regulation for the Implementation of the Drug Administration Law, effective 15 May 2026. The Regulation settled the principle: a holder that genuinely needs to may entrust qualified enterprises with segmented production, subject to a unified quality assurance system covering the whole process and every site. What the Regulation did not do was specify which products qualify, what an application looks like, how licensing and registration interact, or what happens when a site is outside the country. Those are the questions the announcement answers.

The timing is not accidental. A pilot that expires on 31 December 2026 and a standing procedure that has not yet been finalised in mid-November leave a narrow window. Companies with products in the pilot need continuity; companies outside it need to know whether to prepare an application. The thirty-day comment period is short for a text of this density, which suggests the regulator is working to a fixed schedule and expects comment to be targeted rather than exhaustive.

There is also a signal in what the draft does not do. It does not extend segmented production to every product that could technically be split. It does not create a general right to commission manufacturing abroad. It does not remove the requirement for a drug manufacturing licence except in one defined case. The consolidation is real, but it is a consolidation of a narrow permission, not a broadening of it.

For companies, the practical difference between a pilot and a procedure shows up in the diligence file. Under a pilot, a legal opinion on manufacturing structure could describe the regulator’s practice and little more; under a procedure, it can cite a condition and test the facts against it. That shift is small in words and large in what it lets a board approve, because a condition can be assessed before money is committed, while a practice can only be assessed after a conversation that may or may not happen.

The four eligible categories

The draft lists four categories of product that may be produced in segments. The list is exhaustive, and a product that does not fall within one of the four cannot use the procedure however sensible the manufacturing logic may be.

The first category is innovative drugs whose production process or facilities and equipment carry special requirements. The wording points at products where the manufacturing itself is the constraint: a process that needs dedicated equipment, an environment that cannot be shared, a step that does not fit inside a general-purpose plant. For these products, splitting the chain is not a cost exercise but a feasibility one.

The second category covers drugs and vaccines that departments under the State Council have identified as clinically urgently needed, and drugs and vaccines needed for responding to public health emergencies or for reserve purposes. This is the category that carries the public-interest rationale most visibly. It also carries the most external dependency, because the identification comes from outside the drug regulator.

The third category is multi-valent and multi-component vaccines. The logic here is structural: a combination vaccine may require several antigen components, each with its own manufacturing chain, and consolidating all of them at a single site is not always realistic. Segmentation lets the components be made where they are best made and combined where the combination is validated.

The fourth category is the new one, and it is the one most likely to change behaviour. It covers an originator drug that has already obtained a marketing authorization in China, is manufactured outside the country, and has some or all of its steps transferred onshore for segmented production. In substance this is an onshoring route for products that are already approved. The product does not become a new drug; it becomes a product with a different manufacturing footprint.

The legal gateway for the fourth category is the third item of Article 32 of the Regulation, which leaves room for other drugs designated by the national drug regulator. The draft uses that room. The significance is procedural as much as substantive: it shows that the reserved category is not decorative, and that the regulator intends to use it to solve specific problems rather than to keep a general discretion in reserve.

Two limits sit on top of the four categories, and both point in the same direction. A new application for marketing must concern a product that the holder developed itself or in whose development the holder participated throughout. A licensee that acquired a product and now wishes to restructure its manufacturing does not automatically qualify. And the first category has no post-approval change entry at all: only products within the second, third and fourth categories may apply to change an existing authorization into a segmented arrangement.

The combined effect is a procedure that favours companies with an unbroken development history and a live pipeline. For a group that in-licenses assets into China and then looks for manufacturing flexibility, the draft offers less than a first reading might suggest. For a group that develops in China, or that developed globally and holds the full development record, the draft offers a route that did not previously exist in published form.

Two-track procedure and the sterile-experience threshold

An application runs on two tracks at once, and confusing them is the most common way to lose time.

The licensing track runs through the provincial drug regulator where the site is located. The holder, or the entrusted manufacturer, applies for the issue or the amendment of a drug manufacturing licence covering the site and the process. The licence is the permission to manufacture; it is not the permission to market the product made there.

The registration track runs through the drug evaluation centre for products not yet approved, and through the post-approval change framework for products already on the market. For the latter, the holder assesses the change category under the Measures for the Administration of Post-Approval Changes to Drugs (Trial) and files accordingly. A change in manufacturing site is not a uniform event: the same physical move can fall into a different category depending on the product, the process and the validation data. The classification determines the depth of review, and getting it wrong is a substantive error rather than a formatting one.

The two tracks are not sequential in a strict sense. Licensing work can proceed while the registration question is being resolved, and in practice it must, because a registration application that names a site without a licence is incomplete. But the two tracks answer to different reviewers, and a company that treats a licence amendment as the whole project will find itself with permission to manufacture a product it is not cleared to place on the market.

On organisation, the draft sets a principle rather than an absolute rule: the same process step should, in principle, be entrusted to one manufacturer. The principle is stated in the application and recorded in the approved process. The practical effect is that a holder cannot quietly split a single step between two contractors to manage capacity, and cannot move a step between contractors without the change being visible in the approved documentation.

Intermediates carry their own requirement. Where a product passes between sites in an unfinished state, the intermediate must have a clear quality standard that has been approved. This is the clause that turns a logistics arrangement into a regulatory object. An intermediate without an approved standard is not a legitimate handover point; it is an unregulated material in transit.

The sterile threshold is the hardest condition in the draft. For sterile products, at least one of the two parties — the holder or the entrusted manufacturer — must have three years or more of commercial production experience in the same dosage form. The requirement is deliberately relational: it does not demand that both sides be experienced, but it does demand that the experience exist somewhere in the pair. A new entrant cannot be paired with another new entrant, however strong their respective quality systems may look on paper.

The draft then addresses relaxations. A 2025 announcement, numbered 134, set out four situations in which the experience requirement could be relaxed. The draft recognises only the second and the fourth: the case where the parties sit within the same group under a unified quality system, and the case of an entrusted manufacturer that has been materially involved in developing the product and in manufacturing clinical trial supplies. The first and third relaxations are not carried over.

The omission is significant and easy to miss. A company that has been relying on one of the two excluded relaxations during the pilot should assume that the reliance does not survive into the standing procedure. Whether the final text restores them is a question worth putting in a comment, because the draft as published does not.

Where the holder’s obligations land

The draft distributes work but not responsibility. The holder remains the party answerable for the product, and the obligations listed in the text are written to make that answerability operational rather than nominal.

The first is presence. For each product in production, the holder must station two or more technical staff at the manufacturing site. This is not an audit visit and not a periodic review; it is a continuous presence during production. For a foreign holder using a domestic contractor, it means either seconding staff or contracting for a resident technical function that the regulator will treat as the holder’s own.

The second is audit. At least once a year, the holder must conduct a full-system GMP audit covering all manufacturing sites. The phrase “all sites” is the operative one: an audit programme that covers the domestic contractor but not the overseas site, or that covers the main site but not a site performing a single step, does not meet the requirement.

The third is the cold chain. Where bulk drug substance or another intermediate travels a long distance, the holder must validate the transport, monitor temperature, and inspect the material on arrival. The three obligations are cumulative. Validation without monitoring produces a document; monitoring without incoming inspection produces a record of a failure discovered too late.

The fourth is governance. A product quality risk consultation must be chaired by the legal representative or the enterprise head. Placing the chair at that level is a deliberate choice: it removes the possibility of treating a cross-site quality question as a technical matter settled below the level at which trade-offs are made.

The fifth is data. At steps where humans and machines interact, video data must be collected. The requirement reflects a supervisory reality: in a segmented chain, the reviewer cannot rely on walking the line, so the record of the line becomes the evidence. Companies should expect the retention, accessibility and integrity of that video to be examined alongside the batch records.

None of these obligations is novel in substance — a manufacturer operating a single site faces equivalents — but their combination is demanding for a chain that has been cut into pieces. The audit obligation looks outward to every site; the presence obligation looks inward to a single line; the data obligation looks backward at what actually happened. A holder that runs them as three separate programmes usually discovers, at the worst moment, that the evidence each produces does not line up.

Supervision follows the same logic. Provincial regulators are to manage on a basis described as one file per enterprise and one approach per product, with at least one GMP compliance inspection per product each year. For cross-province arrangements, the two sides are to notify each other periodically of inspection results, sampling results, adverse reaction information and major changes, and to conduct joint investigations or assist each other where a major risk emerges.

The cross-province mechanism deserves attention from foreign groups because it is where delay usually appears. A product split between two provinces has two supervisory relationships, and the notification duty is what keeps them from diverging. A holder that maintains its own parallel record of what each province has been told will find the arrangement easier to manage than one that relies on the two regulators to reconcile.

Cross-border arrangements: a one-way opening

The fifth part, spanning Articles 17 to 20, deals with arrangements that cross the border. It is the part most relevant to foreign holders, and it is the part where the draft is most explicit about direction.

Inbound is permitted. A foreign holder may entrust some or all of its process steps to segmented production inside China, provided the product falls within one of the four categories. The condition is not waived by the cross-border character of the arrangement; it is applied to it. A product that is not in one of the four categories does not become eligible because the holder is foreign.

Outbound is restricted. A domestic holder may not, in principle, entrust a foreign manufacturer with segmented production. The single exception is the fourth category, the originator drug that was approved in China while made abroad and is now being brought onshore. In that case the arrangement is the reverse of the general prohibition, because the product’s history already involves an overseas site.

The asymmetry is a policy choice, and it is worth stating plainly rather than reading a motive into it. The draft opens a route into the country and keeps the route out of the country closed except where a specific product history justifies it. Companies planning a regional manufacturing footprint should not assume that the inbound permission implies a reciprocal outbound one.

A group weighing an inbound arrangement should also look at its contract architecture earlier than it usually does. Joint liability shifts the risk allocation between the holder and the domestic party, and the annual audit of overseas sites shifts the cost base. Both are easier to negotiate before a licence annotation exists than after, because after the annotation the domestic party knows it cannot walk away without a regulatory consequence.

On documentation, the draft offers a practical concession. A cross-border application may use production qualifications and marketing authorization materials issued by domestic or foreign regulators to substitute for application materials that do not fit the arrangement. This matters most where a foreign site holds approvals that have no Chinese equivalent document. The concession is about form, not about substance: it does not reduce the quality standard, it reduces the paperwork that would otherwise be impossible to produce.

Annotation differs by direction, and the difference is operational. Where a foreign holder uses a domestic entrusted manufacturer, the arrangement is recorded on the copy of the domestic party’s drug manufacturing licence. Where a domestic holder commissions a foreign manufacturer, the foreign enterprise’s name, the site address and the product are recorded on both the original and the copy of the licence. The distinction reflects who holds the licence: the annotation follows the licensed entity.

The most consequential clause concerns responsibility. Where a foreign holder commissions segmented production inside China, a domestic responsible person performs the foreign holder’s obligations and bears joint liability. Joint liability is not a formality. It means the domestic party is exposed to the same regulatory consequences as the holder for failures in the arrangement, and it changes the negotiation of any contract that gives effect to the arrangement.

The audit obligation travels with it. The domestic responsible person must conduct at least one full-system GMP audit each year covering the overseas sites as well. For a foreign group, this is the clause that turns a Chinese entity into a genuine quality gate for global manufacturing, not merely a local representative. Groups should consider whether the entity they nominate has the technical depth and the mandate to audit a site it does not control.

The ADC carve-out

Article 22 addresses a problem specific to antibody-drug conjugates. The manufacturing chain for these products includes small-molecule intermediates that are chemically well characterised, stable in handling, and controllable in quality, and that play a decisive role in the quality and efficacy of the finished product. Some of these intermediates are not managed as drugs and not managed as active pharmaceutical ingredients.

Under a literal reading of the licensing rules, a site making such an intermediate could be treated as needing a drug manufacturing licence. The draft resolves the question by excusing the site from holding one, on conditions.

The conditions are cumulative. The site must comply with GMP. It must be inspected by the provincial regulator where the holder is located. The site and the process must be recorded on the licence. And the site remains subject to routine supervision and to extended supervision. Where the site is in another province, the provincial regulator at the site must additionally issue an opinion consenting to the entrusted production.

Read together, the carve-out is narrower than a headline reading suggests and more useful than a cautious one. It removes a licensing requirement that was poorly matched to the material, and it replaces it with an inspection and annotation regime that keeps the site inside the supervisory perimeter. The site does not become unregulated; it becomes regulated through the holder’s licence rather than through a licence of its own.

For groups with conjugates in development, the practical question is whether a given intermediate meets the description. Maturity of process, stability of physicochemical properties, controllability of quality, and criticality to quality and efficacy are the stated criteria. A company that wants the carve-out should be able to document each of them, because the inspection will ask for exactly that.

The clause also has a supply-chain consequence. Conjugation and payload manufacturing are often spread across specialist sites, some of which are not pharmaceutical facilities in the ordinary sense. The carve-out allows those sites to remain in the chain without each acquiring a licence, which reduces a structural barrier to locating conjugate work in China. It does not reduce the holder’s obligation to audit, to annotate, and to answer for the result.

Vaccines and the sequencing question

Vaccines appear in the second and third categories, and the draft adds a sequencing rule for a subset of them. Where the circumstances in Article 33 of the Regulation apply, the applicant must first seek approval from the national drug regulator under the Vaccine Administration Law, and only then proceed to licensing and registration.

The sequencing matters because it inverts the ordinary order of work. In a standard application, licensing and registration run in parallel with review. For the vaccines caught by Article 33, the approval comes first, and the licensing and registration steps follow it. A project plan that assumes parallel tracks will be wrong for these products, and the error will surface late.

The draft does not restate the substance of Article 33, and companies should read the Regulation directly rather than infer the scope from the announcement. What the announcement contributes is the procedural consequence: approval under the Vaccine Administration Law is a precondition, not a parallel filing.

What foreign holders should do before 30 November

The comment window closes on 30 November 2026, and the mailbox is the one named in the notice. A comment that identifies a specific article and a specific operational consequence is more likely to be useful than a general statement of support or concern. Four questions are worth answering internally before that date.

First, does the portfolio map onto the four categories, and onto which one? A product may appear to fit the first category because its process is unusual, while in fact fitting the fourth because its history is an overseas approval followed by an onshore transfer. The categories carry different consequences, including whether a post-approval change route exists at all.

Second, is the development history clean? For a new application, the holder must have developed the product itself or participated throughout. Groups that acquire assets should establish, before planning a segmented arrangement, whether the development record supports the claim. Where it does not, the route is closed regardless of the manufacturing logic.

Third, does the sterile experience sit on the right side? At least one party must hold three years of commercial experience in the same dosage form, and only two of the four relaxations are recognised. A group relying on an excluded relaxation should test the assumption now rather than at filing.

Fourth, who will be the domestic responsible person, and can that entity audit? Joint liability plus an annual full-system audit of overseas sites is a demanding combination. The entity needs technical capability, access to the sites, and a mandate that survives commercial pressure.

Beyond the four questions, three preparatory steps are worth taking. Map every site in the proposed chain against the licence annotation rules, because the annotation differs by direction and an omission is a defect. Confirm that each intermediate handover point has an approved quality standard, since an unapproved intermediate is not a valid boundary. And for conjugates, assemble the evidence that the intermediate meets each of the four criteria in Article 22, because the carve-out is granted on documentation and inspection rather than on description.

Outlook

The draft is a procedural document, and its effects will be felt in individual applications rather than in a single announcement. The parts most likely to be adjusted before the final text are the two excluded relaxations and the treatment of products whose development history is shared between an originator and a Chinese licensee. Both are the kind of question that a comment period exists to surface.

For planning purposes, the safe assumption is that the four categories, the two-track procedure, the one-way cross-border rule and the antibody-drug conjugate carve-out will survive in substance. The variables are the transitional arrangements for products currently in the pilot, and the precise wording of the experience requirement.

Companies with pilot products should treat the 31 December 2026 expiry as a real deadline and plan continuity rather than assume an automatic rollover. Companies without pilot products now have, for the first time, a published description of what a segmented arrangement looks like — which is itself the main change. The permission was always there in principle. What was missing was the procedure, and the procedure is now on the table for comment.

Sources

  • 国家药监局综合司·《关于药品委托分段生产有关事项的公告(征求意见稿)》征求意见页面(反馈截止 2026-11-30):https://www.nmpa.gov.cn/xxgk/zhqyj/zhqyjyp/20261113111413106.html
  • 中国政府网·国务院公报 2026 年第 4 号《中华人民共和国国务院令(第 828 号)中华人民共和国药品管理法实施条例》(2026-01-16 公布,2026-05-15 施行,含第三十二条分段生产条款):https://www.gov.cn/gongbao/2026/issue_12546/202602/content_7057461.html
  • 蒙城县人民政府(转载)·《中华人民共和国药品管理法实施条例》(2026 年修订,含第三十二条、第三十三条分段生产条款):https://www.mengcheng.gov.cn/XxgkContent/show/3008349.html

中国药品分段生产:从试点走向常规程序

为什么此刻值得看

2026 年 11 月 13 日,国家药监局综合司发布《关于药品委托分段生产有关事项的公告(征求意见稿)》。反馈截止 2026 年 11 月 30 日,意见投递到公告指定的邮箱。这份文本是征求意见稿,在正式稿印发之前不对任何人产生约束力,正式稿的措辞、范围以及对在审品种的过渡安排,都可能与现在不同。

让这份稿子值得现在读、而不是等到十二月再读的,是它背后的时间表。生物制品分段生产试点 2024 年 10 月启动,2026 年 12 月 31 日到期。新修订的《药品管理法实施条例》2026 年 5 月 15 日起施行。两个日期之间留着一道缝:条例给了分段生产的法律依据,但落地路径一直是试点,逐案商议,入口靠谈、不靠公布。这份公告要做的,就是把标准写下来。

用平实的话说,分段生产指的是同一个产品的不同生产工序分散在多个场地完成,有时还分散在多个法人主体手里,而这些场地与主体共同处在一套由上市许可持有人拥有的质量保证体系之下。它不是通常意义上的外包。持有人并没有随工序一起把责任交出去,它保留责任,只把工作分出去。这条区分,几乎决定了公告里的每一条。

对外资集团而言,这份稿子有三处值得注意。第一是第四类品种,它第一次为”已在境内获批、在境外生产”的原研药开出一条把部分或全部工序搬到境内、又不必按新药重新注册的路。第二是跨境条款的方向:门朝里开,除一种情形之外不朝外开。第三是抗体偶联药物某些中间产品的豁免,它卸掉了一项与物料性质并不匹配、代价却很高的许可负担,但并没有卸掉围绕着它的质量义务。

这三点各有商业含义。一家长期把中国当作成品加工与包装基地的全球企业,现在可以设想反过来做:把原液或中间产品放在境内生产,并接受随之而来的监管足迹。一家一直在等试点转正的企业,可以按公布的条件而不是按谈判地位来安排投资。而一家做抗体偶联药物的企业,可以不再把某个工艺成熟的小分子中间体当作需要药品生产许可的对象——前提是它能满足附在后面的条件。

这份稿子在澄清的同时也在收窄。新申请上市的品种,必须是持有人自主研发或者全程参与研发的。第一类没有上市后变更入口,也就是说,一个已经上市、且只落在第一类的产品,不能借这份公告在事后重组生产布局。把稿子读成一次普遍放开,是误读。它是一次程序上的收口,范围被有意画得很清楚。

从试点到常规程序

这份稿子接替的试点,2024 年 10 月针对生物制品启动。它的逻辑是试验性的:挑少量品种,摸索一条被切开的制造链该怎么监管,再从个案里学。在监管方还没决定标准应该是什么的时候,这种模式有实实在在的好处。但当企业需要押上资本时,它的短板也很明显:能不能进,取决于一次谈话;谈出什么结果,不会公布。

常规程序改变的是决策的性质。企业不再问”这个产品有没有可能被允许”,而是问”它是否满足已经公布的条件”。答案变得可核查,否定性的答案也变成可以处理的东西——调整品种范围、调整场地配置,或者换一条路走。

稿子的结构是六部分、二十三条。各部分的划分顺着一个申请的流程走,而不是顺着法条的体例走:总则、可分段品种、申请与审批、持有人义务、跨境安排、附则。这个次序本身带着信息。跨境安排单独成篇、位置靠后,说明它被当作一个独立的政策问题处理,而不是境内申请的一个变体。

法律基础是 2026 年 5 月 15 日施行的新修订《药品管理法实施条例》。条例把原则定了下来:确有需要的持有人可以委托符合条件的企业分段生产,前提是建立覆盖全过程、全部生产场地的统一质量保证体系。条例没有做的是:说明哪些品种够格、一份申请长什么样、许可与注册怎么衔接、场地在境外时怎么办。这些正是公告要回答的问题。

时间上的安排并非偶然。一个 2026 年 12 月 31 日到期的试点,加上一份十一月中旬还没定稿的常规程序,中间的窗口很窄。已在试点里的品种需要连续性,试点之外的品种需要知道要不要准备申报。对一份密度这么高的文本给出三十天反馈期,说明监管方按固定日程推进,并且预期意见是点对点的,而不是面面俱到的。

稿子没做什么,也是一种信号。它没有把分段生产扩展到所有技术上可以拆分的品种;它没有创设一项委托境外生产的普遍权利;除了一种明确情形,它没有免除药品生产许可。收口是真的,但收的是一项范围很窄的许可的口,而不是把它放宽。

对企业而言,试点与程序之间的实际差别,会出现在尽职调查文件里。试点之下,一份关于生产结构的法律意见只能描述监管惯例,再多也就到此为止;程序之下,它可以引用一项条件,并拿事实去比对。这个转变在字面上很小,在”董事会能批什么”上却很大——条件可以在掏钱之前评估,惯例只能在一场不知会不会发生的谈话之后评估。

四类可分段品种

稿子列出四类可以分段生产的品种。这份清单是穷尽式的,一个品种只要不落在这四类里,无论制造逻辑多么合理,都用不上这套程序。

第一类是生产工艺、设施设备有特殊要求的创新药。这个表述指向的是制造本身构成约束的产品:需要专用设备的工艺、不能共用的环境、塞不进通用厂房的一道工序。对这类产品,拆链不是成本问题,而是可行性问题。

第二类是国务院有关部门提出的临床急需药品及疫苗,以及应对突发公共卫生事件急需或者储备需要的药品及疫苗。这是公共利益理由最外显的一类。它同时也最依赖外部:认定来自药品监管系统之外。

第三类是多联多价疫苗。这里的逻辑是结构性的。一支联合疫苗可能要由几种抗原组分构成,每种组分各有自己的制造链,把它们全部塞进一个场地并不总是现实。分段让各组分在最适合的地方生产,再在组合工艺经过验证的地方完成组合。

第四类是新加的,也是最可能改变行为的一类。它指已取得境内上市许可、在境外生产、把部分或者全部工序转移至境内分段生产的原研药品。实质上,这是一条针对已获批产品的境内落地通道。产品不会因此变成新药,它只是换了一套生产足迹。

第四类的法律入口,是《实施条例》第三十二条第三项——为”国家局规定的其他药品”留的那道口子。稿子用掉了这道口子。意义一半在程序、一半在实体:它说明这项保留条款不是摆设,监管方打算用它解决具体问题,而不是留着一项笼统的裁量。

四类之上还压着两条限制,方向一致。新申请上市的品种,必须是持有人自主研发、或者全程参与研发的。一家把品种买进来、现在想重组生产的持证企业,并不自动够格。第一类则完全没有上市后变更入口:只有落在第二、第三、第四类的产品,才可以把已有的批件改成一种分段安排。

两条限制合起来,形成一套偏向”研发史完整、在研管线在手”的企业的程序。对一个把资产引进中国、再去找制造灵活度的集团来说,这份稿子给的东西比初读时感觉的要少。而对一个在中国做研发、或者在全球做研发并握有完整研发记录的集团来说,这份稿子给的是一条此前没有以公布形式存在的路。

两条线并行与无菌门槛

一份申请同时跑两条线,把两条线混为一谈,是最常见的误时方式。

许可线走场地所在地的省级药品监管部门。持有人或者受托生产企业,就场地与工艺申请《药品生产许可证》的核发或者变更。这张证是生产的许可,不是把那里生产出来的产品拿去上市的许可。

注册线走两条分支:尚未上市的品种报药品审评中心;已经上市的品种,按《药品上市后变更管理办法(试行)》走上市后变更。后一条分支里,持有人要先判定变更类别,再据以申报。生产场地的变化不是一件整齐划一的事:同样一次物理搬迁,因为品种、工艺、验证数据不同,可能落在不同类别。类别决定审评的深浅,判错类别是实体错误,不是格式错误。

两条线在严格意义上并不是先后关系。许可工作可以在注册问题解决的同时推进,实践中也必须如此,因为一份写明了场地却没有相应许可的注册申请是不完整的。但两条线面对的是不同的审评人,把许可证变更当成整个项目的企业,最后会拿到一张”可以生产、但不能上市”的许可。

在组织方式上,稿子给的是原则而不是铁律:同一道工序原则上委托一家生产企业。这条原则要在申请中说明,并在核准的工艺中载明。实际效果是,持有人不能为了调节产能把一个工序悄悄分给两家受托方,也不能在没有留下文件痕迹的情况下把一道工序从一家换到另一家。

中间产品有自己的要求。产品以未完成状态在场地之间流转的,中间产品必须有明确、且经过核准的质量标准。这一条把一项物流安排变成了监管对象。没有核准标准的中间产品,不是一个合法的交接点,而是一批处在监管真空里的物料。

无菌门槛是稿子里最硬的条件。对无菌药品,委托方与受托方之中至少一方,必须具备同剂型三年以上的商业化生产经验。这个要求是有意设计成”关系型”的:它不要求双方都有经验,但要求这份经验确实存在于这一对组合之中。两个新手凑在一起,无论各自的质量体系在纸面上多漂亮,都不行。

稿子接着处理放宽。2025 年的一份公告(第 134 号)列了四种可以放宽经验要求的情形。稿子只认其中第二项和第四项:双方同处一个集团、实行统一质量体系的;以及主要参与该药品研发、并生产过临床试验用药品的受托方。第一项与第三项没有沿用。

这个省略很重要,也很容易被忽略。在试点期间依赖那两项被排除的放宽情形的企业,应当假定这份依赖不会带进常规程序。正式稿会不会恢复,是一个值得写进意见的问题,因为按现在公布的文本,答案是否定的。

义务落在持有人身上

稿子把工作分了出去,但没有把责任分出去。持有人始终是对产品负责的一方,文本列出的义务,是为了让这份责任可执行,而不是停留在名义上。

第一项是在场。每个在产品种,持有人要在生产场地派驻两名以上技术人员。这不是一次审计访问,也不是一次阶段性复核,而是生产期间的持续在场。对使用境内受托方的境外持有人来说,这意味着要么派人,要么安排一项常驻的技术职能,而监管方会把这项职能视为持有人自己的。

第二项是审核。持有人每年至少开展一次覆盖全部生产场地的全体系 GMP 审核。”全部场地”是这句话的关键。一套只覆盖境内受托方、不覆盖境外场地的审核方案,或者只覆盖主场地、不覆盖承担单道工序场地的方案,都不满足要求。

第三项是冷链。原液或者其他中间产品长距离运输的,持有人要验证运输、监控温度,并在入厂时检验。三项义务是叠加的。做了验证不做监控,得到一份文件;做了监控不做入厂检验,得到一份”发现得太晚”的记录。

第四项是治理。产品质量风险会商必须由法定代表人或者企业负责人主持。把主持人定在这一层是有意的选择:它排除了把跨场地质量问题当成技术细节、在做出取舍的那一层之下解决的可能。

第五项是数据。在人机交互的环节,要采集视频数据。这条要求反映的是一种监管现实:在一条被切开的链上,检查员没法靠走线来确认,于是产线的记录就成了证据。企业应当预期,这些视频的留存、可调取性与完整性,会和批记录一起被看。

这几项义务单拎出来都不新鲜——只经营一个场地的生产企业面对的是同类要求——但叠在一条被切开的链上,要求就变得苛刻。审核义务朝外看,看的是每一个场地;在场义务朝内看,看的是某一条产线;数据义务朝后看,看的是当时到底发生了什么。把三者当成三个互不相干的项目的持有人,往往会在最糟糕的时候发现:三套证据对不上。

监督遵循同样的逻辑。省级监管部门按”一企一档、一品一策”管理,每个品种每年不少于一次 GMP 符合性检查。跨省安排的,两地要定期互相通报检查结果、抽检结果、不良反应信息与重大变更;出现重大风险的,联合调查或者相互协查。

跨省机制值得外资集团多看一眼,因为拖延通常就出在这里。一个跨两省的产品有两段监管关系,而通报义务是防止两段判断分叉的东西。自己另建一份”哪个省被告知了什么”的台账的持有人,会比指望两地监管方自行对账的持有人更省事。

跨境安排:只朝一个方向开

第五部分(第十七条至第二十条)处理跨越关境的安排。这是与境外持有人最相关的一部分,也是稿子把方向讲得最明白的一部分。

进来可以。境外持有人可以把部分或者全部工序委托境内分段生产,前提是品种落在四类之中。这个条件不会因为安排具有跨境性质而免除,而是被适用到跨境安排上。不在四类里的品种,不会因为持有人是境外的就变得够格。

出去受限。境内持有人原则上不得委托境外生产企业分段生产。唯一的例外是第四类——那个在境内获批、在境外生产、如今正在往境内搬的原研药。这种情形下,安排本身就是一般禁令的反面,因为产品的历史里本来就有一个境外场地。

这种不对称是一项政策选择,值得平实地陈述,而不必去揣测动机。稿子开了一条进国的路,同时把出国的路关着,只在特定产品历史足以支撑时例外。规划区域制造布局的企业,不应当从”进来可以”推出”出去也可以”。

考虑把工序搬进来的集团,还应当比惯常更早去看合同架构。连带责任改变的是持有人与境内一方之间的风险分配,对境外场地的年度审核改变的是成本结构。两件事在许可证载明之前谈,都比在载明之后谈容易——载明之后,境内一方清楚自己没有退路,因为退出本身就是一项监管后果。

在材料上,稿子给了一项务实的让步。跨境申请可以用境内、境外监管部门出具的生产资质与上市许可材料,替代那些不适用于该安排的申请材料。这在境外场地持有境外批件、而该批件在境内没有对应文件时最有用。这项让步是形式上的,不是实体上的:它不降低质量标准,只是减少了本来无法产出的那部分纸面工作。

载明方式按方向不同,而这个差别是操作性的。境外持有人委托境内生产的,在境内受托方《药品生产许可证》副本上载明。境内持有人委托境外生产的,境外企业名称、场地地址、产品要在许可证正本与副本上同时载明。区别的来源是谁持证:载明跟着持证主体走。

最有分量的条款是责任。境外持有人跨境委托境内分段生产的,由境内责任人履行境外持有人的义务,并承担连带责任。连带责任不是手续。它意味着境内一方要为这项安排中的失误承担与持有人同等的监管后果,也意味着任何落实这项安排的合同,谈判方式都要改。

审核义务随之同行。境内责任人每年至少要做一次覆盖境外场地的全体系 GMP 审核。对境外集团来说,这一条把一家中国主体变成了全球制造体系里一道真实的质量关口,而不只是一个本地代表处。集团需要考虑:拟指定的这家主体,有没有技术深度和授权去审核一个它并不控制的场地。

抗体偶联药物的例外

第二十二条处理的是抗体偶联药物特有的一个问题。这类产品的制造链里有小分子中间体,化学性质清楚、处置中稳定、质量可控,并且对成品的质量与疗效起关键作用。其中一些中间体既不按药品管理,也不按原料药管理。

按许可规则的直白读法,生产这种中间体的场地可能被当作需要药品生产许可证。稿子用”附条件豁免”回答了这个问题。

条件是叠加的。场地要符合 GMP;要经持有人所在地省级监管部门检查;场地与工艺要在许可证上载明;场地仍要接受日常监管与延伸监管。场地在另一省的,还须由场地所在地省级监管部门出具同意受托生产的意见。

合起来读,这项例外比标题式读法要窄,也比保守式读法要有用。它卸掉了一项与物料性质并不匹配的许可要求,换上一套检查加载明的安排,把场地留在监管边界之内。场地不会因此变成无人监管,它变成通过持有人的许可证接受监管,而不是自己持一张证。

对有偶联药物在研的集团来说,实际问题是一项给定的中间体是否落入这个描述。工艺成熟、理化性质稳定、质量可控、对质量疗效起关键作用,是文本给出的四项标准。想要用上这项例外的企业,应当能为每一条留下文件,因为检查要的正是这些。

这一条还有供应链上的后果。偶联与载荷的生产常常分散在专业场地之间,其中一些并不是通常意义上的制药车间。例外让这些场地可以留在链上而不必各自去拿一张证,从而降低了把偶联环节放在境内的结构性障碍。它不降低持有人的审核、载明与最终担责义务。

疫苗与先后顺序

疫苗出现在第二类和第三类里,稿子又为其中一个子集加了顺序规则。符合《实施条例》第三十三条情形的,要先按《疫苗管理法》向国家药监局申请批准,再办许可与注册。

这个顺序之所以重要,是因为它把通常的工作次序倒了过来。在一般申请里,许可与注册同审评并行推进。对落入第三十三条的疫苗,批准在前,许可与注册在后。按并行双线做的项目计划,在这些品种上是错的,而错误会在很晚的时候才暴露。

稿子没有复述第三十三条的实体内容。企业应当直接去读条例,而不是从公告反推适用范围。公告贡献的是程序后果:按《疫苗管理法》取得批准,是前置条件,不是一条并行的申报。

境外持有人应在 11 月 30 日之前做什么

反馈窗口 2026 年 11 月 30 日关闭,邮箱是公告里写明的那个。点名到具体条文、并说明具体操作后果的意见,比泛泛的支持或担忧更有用。在此之前,有四个问题值得先在内部回答清楚。

第一,产品组合落在四类的哪一类?一个产品可能看起来像第一类,因为工艺特殊;实际上却属于第四类,因为它的历史是”境外获批、再往境内搬”。不同类别的后果不同,包括到底有没有上市后变更入口。

第二,研发史是否干净?新申请上市,持有人必须自己研发、或者全程参与研发。收购资产的企业,在规划分段安排之前,应先确认研发记录能否支撑这个主张。支撑不了,无论制造逻辑多顺,路都是关着的。

第三,无菌经验是否落在对的一侧?至少一方要有同剂型三年以上商业化生产经验,而四项放宽只有两项被承认。依赖被排除的那两项的企业,现在就应当检验这个假设,而不是等到申报。

第四,境内责任人是谁,它审得动吗?连带责任加上对境外场地的年度全体系审核,是很硬的一组要求。这家主体需要技术能力、需要能进入那些场地、还需要一份不会被商业压力磨掉的授权。

四个问题之外,还有三件准备工作值得做。把拟设链条上的每个场地对照许可证载明规则过一遍,因为载明方式按方向不同,漏一处就是缺陷。确认每一个中间产品交接点都有核准的质量标准,因为没有核准标准的中间产品不是一个有效的边界。做偶联药物的,把中间体满足第二十二条四项标准的证据攒起来,因为这项例外是凭文件与检查给出的,不是凭描述给出的。

展望

这是一份程序性文件,它的效果会体现在一个个具体申请里,而不是体现在某一次发布里。正式稿最可能调整的地方,是那两项被排除的放宽情形,以及研发史由原研方与境内被许可方共享的品种怎么处理。这两类问题,正是反馈期存在的理由。

做规划时,稳妥的假设是:四类品种、双线程序、单向跨境、抗体偶联药物例外,在实体上会保留下来。变量在于试点内品种的过渡安排,以及经验要求的具体措辞。

手上有试点品种的企业,应当把 2026 年 12 月 31 日当作真实期限,按”需要接续”来准备,而不是假定会自动顺延。没有试点品种的企业,第一次拿到了一份关于”分段安排长什么样”的公开描述——这本身就是主要的变化。许可在原则上一直存在。缺的是程序,而程序现在摆上了桌面,等着被评论。

来源

  • 国家药监局综合司·《关于药品委托分段生产有关事项的公告(征求意见稿)》征求意见页面(反馈截止 2026-11-30):https://www.nmpa.gov.cn/xxgk/zhqyj/zhqyjyp/20261113111413106.html
  • 中国政府网·国务院公报 2026 年第 4 号《中华人民共和国国务院令(第 828 号)中华人民共和国药品管理法实施条例》(2026-01-16 公布,2026-05-15 施行,含第三十二条分段生产条款):https://www.gov.cn/gongbao/2026/issue_12546/202602/content_7057461.html
  • 蒙城县人民政府(转载)·《中华人民共和国药品管理法实施条例》(2026 年修订,含第三十二条、第三十三条分段生产条款):https://www.mengcheng.gov.cn/XxgkContent/show/3008349.html

Chongqing’s Free Trade Zone Now Issues Hong Kong and Macao Talent Endorsements to Mainland Staff

  • From 12 November 2026 the China (Chongqing) Pilot Free Trade Zone issues talent endorsements for Hong Kong and Macao travel under the National Immigration Administration’s ten innovation measures.
    自 2026 年 11 月 12 日起,中国(重庆)自由贸易试验区正式试点实施往来港澳人才签注,依据是国家移民管理局关于实施支持扩大开放服务高质量发展十项创新举措的公告。
  • The scheme covers six categories of mainland talent — outstanding, research, education and culture, health, legal, and others — and is not a visa for foreign nationals.
    这项政策面向六类内地人才——杰出人才、科研人才、文教人才、卫健人才、法律人才和其他人才——并不是给外国人的签证。
  • Outstanding talent rated A or B may receive a multiple-entry endorsement valid for one to five years; C and D ratings and the other five categories receive one to three years.
    杰出人才中经认定为 A、B 类的,可签发一至五年多次签注;C、D 类杰出人才及其余各类人才,可签发一至三年多次签注。
  • Each stay in Hong Kong or Macao is capped at thirty days, which suits repeated short trips rather than relocation.
    每次在香港或者澳门停留不超过三十天,适合频繁短程往返,而不是长期驻留。
  • The six categories rest on four different proof systems: a talent code, a title plus a filed list, a consent letter from the justice bureau, or a recommendation letter from the zone.
    六类人才背后是四套不同的证明体系:人才码、职称加备案名单、司法局同意函,以及自贸试验区出具的推荐函。
  • A combined permit and endorsement application takes seven working days, an endorsement alone five, and applicants registered outside Chongqing twenty days.
    通行证与签注一并申请七个工作日内签发,单独申请签注五个工作日内签发,非重庆户籍人才二十日内签发。
  • Fees run from sixty yuan for the travel permit to 240 yuan for a multiple-entry endorsement valid three years or more.
    费用从往来港澳通行证每证六十元,到三年以上多次签注每件二百四十元。

Chongqing’s Free Trade Zone Now Issues Hong Kong and Macao Talent Endorsements to Mainland Staff | 重庆自贸试验区:把港澳往返的通道发给内地骨干

A pilot confined to the free trade zone

On 12 November 2026 the China (Chongqing) Pilot Free Trade Zone began issuing talent endorsements for travel to Hong Kong and Macao. The policy implements part of the National Immigration Administration’s ten innovation measures, which provide for talent endorsements in pilot free trade zones.

The holders are mainland residents: executives, researchers, teachers, clinicians, lawyers and technical staff who work in the zone and travel to Hong Kong or Macao regularly. It is not a visa. It does not bring a foreign national into China, and it does not help a Hong Kong or Macao resident enter the mainland, since they travel on their own permits.

For a joint venture or a Hong Kong-invested company in Chongqing, that distinction decides everything: the endorsement serves the mainland half of the workforce. The people who negotiate with the Hong Kong parent, attend an arbitration in Hong Kong, inspect a Macao clinic, or spend a week at a Hong Kong campus are usually mainland nationals on the company’s own payroll, and Two documents are involved, and confusing them is the most common early mistake. The travel permit is the booklet that identifies the holder and is valid for years; the endorsement is the permission, printed or stored against that booklet, that allows a specific number of entries to a specific place. The talent endorsement changes the second, not the first. An employee whose permit is close to expiry gains little from a five-year endorsement, because the endorsement cannot outlive the document it sits on.

That is worth putting on a checklist before anything else. Permit validity, endorsement validity and passport validity are three different clocks, and the one that expires first governs the trip.

Why the zone boundary decides eligibility

The policy is a pilot, and a pilot has a perimeter. Eligibility is tied to employment inside the China (Chongqing) Pilot Free Trade Zone, not to the city as a whole, which makes the employer’s registered address the first question rather than the last.

That sounds like a formality until a company discovers that its Chongqing office sits just outside the boundary while its warehouse sits inside. The zone publishes a regional identification system precisely because the perimeter is not obvious from a map, and the check takes minutes. Companies should run it before they brief staff, not after an application is returned.

The boundary also explains why the policy is described as a pilot rather than a scheme. A pilot can be adjusted: categories can be widened, lists can be revised, and the perimeter can be redrawn. Anyone building this into a multi-year mobility plan should note which parts are fixed by the national announcement and There is a practical way to test the perimeter that does not depend on the map at all: look at which entity signs the employment contract. A group may run a plant inside the zone, a sales office outside it and a shared services centre in a third district, all under different legal entities. The endorsement follows the employing entity, so the question “does our company qualify” is really the question “which of our entities is inside, and who is employed by it”.

Document the answer while you are at it. A short internal memo that records the entity name, its registered address, the result of the zone check and the date of the check is worth more than a screenshot, because it survives staff turnover and can be attached to an application file.

The memo also settles a question that comes up in every batch: whether a person employed by a group service company can claim a category based on work performed for an operating subsidiary. The answer depends on the entity, not on the desk the person sits at, and Where a group has a choice, the memo also forces a decision that is often avoided: whether to move a role into the entity that sits inside the zone. That is a commercial question with tax and labour consequences, and it should not be answered by an immigration policy. But it should be asked, because the alternative — a qualified person employed by the wrong company — One more note on structure, because it is the question most likely to be asked in a board paper. The endorsement is issued to a person, not to a company, and it does not travel with a job. That means the company cannot bank it as an asset, cannot transfer it when a role moves between entities, and cannot count it in a headcount calculation of mobile staff without noting that the count will decay as people leave.

What the company can do is build a register and keep it current. The register is the asset; the endorsement is only the evidence that the register is being used.

A register also makes the policy legible to people outside human resources. Finance can see the cost, legal can see the exposure, and the business units can see which of their people are mobile and until when. That is a better conversation than one conducted in anecdotes about who once got a five-year endorsement and who did not.

None of this is complicated, and most of it is not even new. What is new is that a mainland employee in a named category can now hold a document that makes a year of short trips possible, and that the paperwork behind it is specific enough to be planned for.

For a company deciding whether to invest a day in this, the answer is in the trip log. If the log shows repeated short visits by people with titles, the day pays back inside a quarter. If it shows one visit a year, the ordinary business route is still the right tool.

Six categories of talent, defined

Eligibility runs through six categories, defined administratively rather than aspirationally. Whether a person fits is decided by a title, a recognition result, the nature of the employer, or a combination of degree and years of service.

Outstanding talent. Those recognised as class A, B, C or D under the New Chongqing talent service measures, coded in the Yucaihui system. This category does not look at titles at all; it looks at a municipal recognition result.

Research talent. Staff with an associate senior title or above at named research bodies: national laboratories and their bases, national key laboratories, national technology innovation centres, national clinical research centres, national Belt and Road joint laboratories, national manufacturing innovation centres, national industrial technology engineering centres and national emerging industry innovation centres; plus centrally affiliated and municipal institutes, Chongqing laboratories, municipal key and Belt and Road joint laboratories, municipal new-type research institutions, and units building major national science infrastructure.

Education and culture talent. Staff with an associate senior title or above at ministry-affiliated and municipal universities, Chongqing campuses of universities from other provinces, and incorporated Sino-foreign and mainland-Hong Kong or Macao cooperative institutions.

Health talent. Staff with an associate senior title or above at tertiary medical institutions, incorporated Sino-foreign joint venture or cooperative medical institutions, and Hong Kong or Macao sole-owned medical institutions.

Legal talent. Mainland arbitrators in Hong Kong or Macao proceedings, mainland mediators handling mainland-Hong Kong or Macao investment disputes, lawyers with foreign-related practice who must travel, and managers and secretaries of arbitral institutions with Hong Kong or Macao business.

Other talent. Managers at Fortune Global 500 headquarters and their Chongqing headquarters; middle management and above plus key technical staff at advanced manufacturing and modern services firms; mainland staff hired in priority fields and innovation team members; doctoral degree holders who have worked in Chongqing for three consecutive years; The categories also reveal what the city is trying to keep. Research institutes, universities, hospitals, arbitration and law, advanced manufacturing, modern services and corporate headquarters are the institutions named; agriculture, retail and construction are not. That is a statement about the industrial map of the zone rather than about the value of individual professions, and it tells a company where its own staff are likely to fit.

One term in the definitions does most of the work: associate senior. It is a rung on the professional title ladder, above intermediate and below full senior, and it is the threshold used for research, education and health talent alike. A person who holds it clears the category; a person one rung below does not, however long the service record.

That single threshold explains why the categories skew towards institutions that run title evaluations. Hospitals, universities and research institutes award titles as a matter of routine; a trading company does not, which is why its senior staff usually enter through the “other talent” route instead.

Four proof systems, not one

The categories look like a list of professions. Operationally they are a list of evidence routes, and the routes do not resemble each other.

Outstanding talent is proved by a code. The Yucaihui system assigns the class and the applicant shows it; there is no document to assemble and no title to verify, because the recognition has already happened elsewhere.

Research, education and health talent are proved by two documents and one cross-check: an employment certificate, a title certificate, and the officer’s comparison against lists filed by the municipal science and technology, education and health commissions. The decisive fact is not what the applicant says but whether the employer and the title appear on a list the commission has already received.

Legal talent is proved by a consent letter from the municipal justice bureau, which reflects the fact that the relevant population is small and its status is registered elsewhere.

Everyone else is proved by a recommendation letter from the designated office of the zone’s regional management body. That is the most discretionary route of the four, and the one where The four routes also differ in lead time, and that is what a planner should note. A talent code can be shown the same day. A title plus a filed list depends on a record that may need correction first. A consent letter depends on another bureau’s calendar. A recommendation letter depends on an internal decision about who the company is willing to put its name behind.

Matching staff to routes is therefore not a classification exercise but a scheduling one: the people on the slowest route should start first, even if they travel least.

The filing list, and what to do before applying

The filing lists deserve separate attention, because they are where most applications quietly succeed or fail.

A list is maintained by a commission, not by the applicant. An employer that has never filed its research staff, or a title that was awarded but never recorded, can leave a genuinely qualified person outside the check. The remedy is administrative and slow: the record has to be corrected upstream before the endorsement application makes sense.

The preparation is therefore unglamorous. Confirm that the institution is on the relevant list; confirm that the title certificate matches the name and the employing entity exactly; confirm that the person’s actual role is consistent with the category being claimed. Where the employer has several candidates, it is worth assembling the file once and checking all of them against it, In practice, three questions to the commission cover most of it: is our institution already on the list, if not what does it take to be added, and does the list record titles individually or by institution. The third matters most for larger employers, because it determines whether one filing covers a department or whether each person needs a separate entry.

Write the answers down and date them. A named contact at each commission is worth more than a filing cabinet. The science and technology commission, the education commission and the health commission each maintain their own list, and each has its own practice about what a list entry looks like. Knowing who to call when an entry is missing turns a month into a week.

Where the institution genuinely is not eligible to be listed, the honest answer is that the category does not apply, and the file should move to a different route rather than wait for a rule change that may never come. In practice the two processes can run at once, and the endorsement file can be assembled while the record is corrected.

Keep the correspondence. Keep the file in one place. The evidence that matters is not the endorsement itself but the trail behind it: the employment certificate as issued, the title certificate as recorded, the list entry as confirmed, the recommendation or consent letter as received. When a renewal comes up three years later, the person who handled the first application may have moved on.

A one-page index per employee — category, route, documents, dates, contact at the issuing body — costs an hour to build and saves a week on the second cycle.

Validity periods: one to five years, one to three years

A and B class outstanding talent may receive a multiple-entry endorsement valid for one to five years. C and D classes, and the other five categories, one to three years.

The split follows the logic of the recognition behind it. The A and B classes are the city’s highest recognition tier, and the longer validity saves those individuals a recurring administrative errand. For everyone else, three years covers a posting cycle and is short enough to be reviewed.

For workforce planning, the practical consequence is that a company should not assume a uniform expiry date across its mainland staff. A mobility calendar that treats the endorsement as a single renewal event will drift out of step within two years, and the staff most likely to be affected are the senior ones A simple calendar solves most of this. List every endorsement holder, the class or category, the issue date and the expiry date, and set a review point well before the earliest expiry. Add a column for the underlying permit, because the two rarely expire together.

The exercise also surfaces a succession problem. If the only person with a five-year endorsement is also the only person who handles Hong Kong arbitration, the company has a single point of failure that a mobility policy should address before it becomes a scheduling crisis.

The thirty-day cap: what the document is not

Every stay in Hong Kong or Macao is capped at thirty days, which suits short, repeated visits — a board meeting, a hearing, a site inspection, a family weekend — and does not support a secondment.

That limit is the clearest signal of the policy’s purpose. It is designed for people whose work requires presence in Hong Kong or Macao for days at a time and repeatedly, not for people who need to live there. An engineer supervising an installation for three months, or a manager seconded to a Hong Kong affiliate for a year, needs a different arrangement entirely.

Read the other way, the cap is generous where it matters. Thirty days is ample for the pattern that actually causes friction: the same person crossing four or five times a quarter, each time for a few days, Companies should also plan for the trips the endorsement cannot cover. A long installation, a semester of teaching, a year-long secondment or a family relocation all sit outside a thirty-day stay, and none of them is improved by holding a talent endorsement. The document is a solution to frequency, not to duration, and internal guidance that blurs the two will produce problems at the border rather than at the desk.

A travel log makes this test objective. For each candidate, list the Hong Kong and Macao trips of the past twelve months with dates and nights, then look at the shape rather than the total. Ten trips of three nights is a different pattern from two trips of twenty nights, and only the first is what the endorsement is built for.

Turnover is the reason the calendar matters more than the first application. A deputy general manager who leaves takes a five-year endorsement with them, and the replacement may be neither in a category nor eligible for one for a year. Companies that treat mobility as an attribute of a role rather than of a person plan for that; companies that treat it as a personal benefit discover it at the worst moment.

The remedy is not complicated: when a role is defined, note which category it would fall into and what the holder would need to qualify. That note belongs in the job description, not in a policy folder.

Legal talent: the narrowest category

The legal category is worth reading closely, because it is the narrowest of the six and the easiest to misapply.

It covers mainland arbitrators sitting in Hong Kong or Macao arbitration proceedings, mainland mediators handling investment disputes between the mainland and Hong Kong or Macao, lawyers with foreign-related practice who need to travel, and managers and secretaries of arbitral institutions with Hong Kong or Macao business. In each case the qualifying fact is not the profession but the connection to a Hong Kong or Macao matter.

A litigator with no cross-border caseload does not fit, however senior. A young arbitration secretary at an institution with a Hong Kong caseload may. That inversion surprises firms that think in terms of seniority, and it explains why the consent letter comes from the justice bureau rather than from the employer: For law firms and arbitration institutions, the implication is to nominate by caseload rather than by seniority. The partner who handles domestic litigation only may not qualify; the counsel who runs the Hong Kong-seated arbitrations almost certainly does; the secretary who manages those files may qualify where a more senior colleague does not.

Institutions should also keep the underlying evidence — the case references, the appointment letters, the institution’s own records of Hong Kong or Macao business — because the consent letter is issued against a connection that someone has to be able to describe.

Documents, fees and times: one service path

The basic file is a valid travel permit, a compliant photograph and the resident identity card, plus proof of category as described above. Serving military personnel and registered state functionaries need an approval letter under their own personnel rules.

A travel permit costs 60 yuan. A multiple-entry endorsement valid for one year or less costs 80 yuan; more than one year and up to two years, 120 yuan; more than two and under three years, 160 yuan; three years or more, 240 yuan.

Processing takes seven working days where permit and endorsement are applied for together, five where only an endorsement is sought. Urgent business, research exchange or a visit to a critically ill relative can use a green channel; in special circumstances the time may be extended to twenty working days with approval. Applicants registered outside Chongqing are issued within twenty days.

Applications go to the exit and entry administration hall of the Chongqing Public Security Bureau, 311 Jinshi Avenue, Liangjiang New Area; the enquiry line is 12367. For a company handling its first batch, the useful sequence is: confirm the zone address, Batching helps in a specific way. The hall handles applications individually, but the preparation is shared: the same template for employment certificates, the same translation conventions, the same cover note explaining the company and the zone address. A company that runs its first three applications as a batch learns the format once and reuses it.

It also produces a useful by-product: a list of who is eligible, who is nearly eligible, and who is not — which is exactly the document a mobility policy needs.

Two practical notes on the hall itself. Appointments cluster around quarter ends and around the start of the academic year, when families travel; and the hotline is more useful for confirming which documents a specific category needs than for booking a slot.

Who this is really for

Read the categories together and a profile emerges: a joint venture deputy general manager with a senior title, a laboratory lead at a municipal institute, a partner handling cross-border disputes, a chief physician at a Sino-foreign hospital.

For them the gain is time: a trip that once required a fresh application becomes a document valid for years. Companies should treat it as a mobility benefit, and check whether key mainland staff hold titles that fit a category.

It is also worth being explicit about who is not covered, because the point is easy to lose in internal communication. Foreign nationals in the same office are outside the scheme; their travel to Hong Kong and Macao runs on their own visa and permit arrangements. Hong Kong and Macao residents do not need an endorsement to go home. Human resources teams should say this plainly in internal announcements. A message that describes the policy as a new benefit for “staff who travel to Hong Kong” will be read by foreign employees as applying to them, and the correction afterwards costs more credibility than the original caution would have.

The accurate framing is narrower and more useful: a document for mainland employees in named categories, inside a named perimeter, for short and repeated trips. Said that way, it also tells people what to do if they fall outside it, which is to ask about their own arrangements rather than to wait for a policy that will not come.

How it differs from the APEC card and other channels

The APEC Business Travel Card, which China issues to eligible business travellers, serves a different purpose: it facilitates entry to participating economies, including short stays, and it is aimed at crossing international borders rather than at travel between the mainland and Hong Kong or Macao.

The talent endorsement does the opposite. It is issued to mainland residents by a mainland authority, for travel to two parts of the country, and its qualifying test is a professional category rather than a business need. A company that holds both instruments for the same executive has not duplicated anything; it has covered two different journeys.

The same logic separates the talent endorsement from the ordinary business endorsement. Both are endorsed on the same travel permit, but the business route is tied to a company’s filing and a particular trip, Put in a single sentence: the business endorsement answers “this company needs this person to make this trip”; the talent endorsement answers “this person belongs to a category the zone has decided to make mobile”. The first is a transaction, the second is a status, and the paperwork follows from that difference.

There is also a timing difference worth knowing. The talent endorsement is issued against a status that persists; the business endorsement is issued against a trip that has been planned. A company that discovers a need on Monday morning will find the business route quicker for a person who is not yet in a category, and the talent route cheaper over a year for a person who is.

A simple decision rule follows. If the person is in a category and travels more than a handful of times a year, the talent route pays for itself in administrative time. If the person is outside the categories, or the trip is exceptional, the business route is the correct instrument and no amount of category analysis will change that.

Where both could apply, the tie-breaker is duration of need: a status-based document is better for a role that will exist for years, and a trip-based document is better for a project that will end. Mixing the two produces applications that are correct in form and wrong in route.

What a company should do first

Start with the boundary. Confirm in writing that the employing entity’s address falls inside the zone, and identify which staff are employed by that entity rather than by an affiliate elsewhere in Chongqing.

Then sort the population into the six categories and note which proof route applies to each. The exercise usually reveals that the research, education and health groups are larger than expected and the outstanding talent group smaller, because the latter depends on a municipal recognition that most people never applied for.

Then fix the records. Titles that exist but are not filed, job descriptions that describe a role more narrowly than it is performed, and institutions that have never submitted a list are the three failures that recur. A final point on sequencing. The policy is new, the lists are administrative, and the first applications from any company will teach it something about its own records. Building a quarter of slack into the first batch is not pessimism; it is an accurate reading of how a pilot behaves in its first months.

One last practical note: the enquiry line, 12367, is the fastest way to settle a question about a specific category before assembling a file. Asking whether a given title qualifies, and what evidence the officer would expect, is a five-minute call that frequently saves a fortnight of preparation.

What a company can do before it calls is narrow down the question. Instead of asking whether the policy applies, ask which of the six categories a given role most resembles, which proof route that category uses, and whether the supporting record already exists. Those three answers turn a general enquiry into a specific one, and specific questions get specific answers.

Run those tests once a year, and the policy stops being news and becomes a control. The boundary check, the category map, the endorsement calendar and the list confirmations together take a day; skipping them costs a quarter of improvisation the first time a senior hire needs to be in Hong Kong next week.

For a company new to the zone, the order is the same as for a company that has been there for years. A closing observation about pilots. The categories, the lists and the perimeter are all administrative decisions, and administrative decisions in a pilot tend to be revisited once the first year of applications has produced some data. Companies that keep a dated record of what they asked, what they were told and what was approved are in a better position when the rules move than companies that relied on a verbal answer.

That is the practical case for treating this as a compliance process rather than a perk. The perk is real — a document that lasts years and replaces a stack of applications — but it is the process around it that determines whether a given employee actually holds one when the trip comes up.

Sources

  • 重庆市公安局·中国(重庆)自由贸易试验区往来港澳人才签注政策及《申请指南》:https://gaj.cq.gov.cn/sy_245/bmdt/202611/t20261112_16401908.html
  • 重庆市人民政府·重庆自贸试验区试点实施往来港澳人才签注政策:https://www.cq.gov.cn/ywdt/zwhd/bmdt/202611/t20261113_16404467.html
  • 重庆日报·往来港澳人才签注受理范围与办理指引:https://www.cqrb.cn/jicengshijian/2026-11-12/2819796_pc.html

重庆自贸试验区:把港澳往返的通道发给内地骨干

只在自贸试验区里试点

2026 年 11 月 12 日起,中国(重庆)自由贸易试验区正式试点实施往来港澳人才签注。这项政策落地的是国家移民管理局《关于实施支持扩大开放服务高质量发展十项创新举措的公告》中的一项安排——在自贸试验区实施往来港澳人才签注政策。

持有人是内地居民:在自贸试验区里工作、又需要经常去香港或者澳门的经营管理人才、科研人员、教师、临床医生、律师和技术骨干。它不是签证。它不能让外籍人士入境中国,也与港澳居民进入内地无关——后者用的是自己的证件。

对一家中外合资企业或者港资企业来说,这个区分决定了全部意义:签注服务的是员工队伍里的内地那一半。与香港母公司谈条款、去香港参加仲裁、到澳门看诊所、去香港校区待一周的人,通常是公司自己发薪的内地员工,这里涉及两份证件,把它们混为一谈是最常见的早期错误。通行证是证明持有人身份、多年有效的那本证件;签注是印在或者记录在这本证件上的许可,规定可以前往哪里、可以去几次。人才签注改的是后者,不是前者。通行证临近到期的员工,从五年签注里得不到多少好处,因为签注不能比它所在的那份证件活得更久。

这一条应当放在清单的最前面。通行证有效期、签注有效期、护照有效期是三只不同的钟,先停的那一只决定行程。

片区边界决定资格

这是一项试点,试点有边界。资格与用人单位是否落在片区之内挂钩,而不是与整座城市挂钩。因此,雇主的注册地址是第一个问题,而不是最后一个。

听起来像形式,直到某家公司发现自己的重庆办公室刚好在界线之外,而仓库在界线之内。自贸试验区之所以公布区域识别系统,正是因为这条界线从地图上并不容易看出来,而核对只需要几分钟。企业应当在向员工说明政策之前先查一次,而不是在材料被退回之后再查。

边界也解释了这项政策为什么被称为试点而不是制度。试点可以调整:类别可以放宽,名单可以修订,范围可以重划。任何把它写进多年期人员流动方案的人,都应当分清哪些内容由国家层面的公告固定,有一个不依赖地图的检验办法:看劳动合同跟谁签。一家集团可能在片区里有工厂、在片区外有销售办公室、在第三个区有共享服务中心,各自是不同的法律主体。签注跟着用人主体走,所以”我们公司符不符合”这个问题,实际上是”我们哪一个主体在片区内、谁与它签约”。

把这件事一次性以书面形式弄清楚的集团,顺便把答案写下来。一份简短内部备忘,记下主体名称、注册地址、区域识别系统的核对结果与核对日期,比一张截图有用,因为它能熬过人员更替,也能附在申请材料后面。

这份备忘还能定下一件每批都会遇到的事:由集团服务公司雇佣的人,能否以其为经营子公司所做的工作主张某一类别。答案取决于签约主体,而不取决于这个人坐在哪张桌子上;把这件事写在备忘里,比在会议上争论更省时间。

六类人才,逐类定义

适用对象分为六类,定义是行政口径,不是倡导性表述。一个人是否符合,取决于职称、认定结果、单位性质,或者学历加年限的组合。

杰出人才。 按《新重庆人才服务管理办法》的认定规则,由“渝才荟”系统赋码认定为 A、B、C、D 类的人才。这一类完全不看职称,看的是市级人才认定结果。

科研人才。 科研机构中具有副高级以上职称的人员,机构范围包括国家实验室及基地、全国重点实验室、国家技术创新中心、国家临床医学研究中心、国家“一带一路”联合实验室、国家制造业创新中心、国家产业技术工程化中心、国家新兴产业创新中心,以及中央在渝科研院所、市属科研院所、重庆实验室、市级重点实验室、市级“一带一路”联合实验室、市级新型研发机构,以及国家重大科技基础设施建设和管理单位。

文教人才。 高等院校中具有副高级以上职称的人员,包括部属与市属高校、市外高校在渝办学机构,以及具有法人资格的内地与港澳合作办学机构、中外合作办学机构。

卫健人才。 医疗卫生机构中具有副高级以上职称的人员,包括三级医疗卫生机构、具有独立法人资格的中外合资合作医疗机构,以及港澳独资医疗机构。

法律人才。 参与香港或者澳门法律仲裁程序的内地仲裁员、处理内地与港澳投资争端的内地调解员、具有涉外法律服务业务需要往返港澳的律师,以及在港澳有业务往来的仲裁机构管理人员和仲裁秘书。

其他人才。 世界五百强企业总部及其重庆总部管理人员;先进制造业、现代服务业企业中层及以上管理人员和骨干专业技术人才;重点发展领域、行业聘雇的内地人才和创新创业团队成员;具有博士研究生学历并在重庆连续工作满三年的人员;这六类也透露了这座城市想留住什么。科研院所、高校、医院、仲裁与律师、先进制造业、现代服务业、企业总部是被点名的机构;农业、零售、建筑没有被点名。这是对片区产业版图的判断,而不是对职业价值的排序,它能告诉一家企业,自己的员工大概率落在哪里。

对合资企业来说,实际的读法是:技术与临床岗位通常覆盖,一般商务岗位通常不覆盖,定义里有一个词承担了大部分工作:副高级。它是职称序列中的一档,位于中级之上、正高级之下,科研、文教、卫健三类都用它作门槛。有这一档的人过关,差一档的人不过关,服务年限再长也不行。

这一道门槛也解释了为什么六类会向开展职称评审的机构倾斜。医院、高校、科研院所把评职称当日常;贸易公司没有这条线,所以它的资深员工通常从”其他人才”这条路径进来。

四套证明体系,而不是一套

六类看起来像一份职业清单,操作上却是一份证明材料清单,而且这几条路径彼此并不相像。

杰出人才靠一个码。系统赋码,本人出示;没有材料要凑,也没有职称要核,因为认定在别处已经完成。

科研、文教、卫健三类靠两份材料加一次比对:在职证明、职称证明,以及出入境人员对照市科技局、市教委、市卫生健康委备案名单所做的核查。起决定作用的不是申请人怎么说,而是用人单位与职称是否出现在主管部门已经备案的名单上。

法律人才靠市司法局的同意函,这反映出一个事实:这一类人数量少,身份状态在别处已有登记。

其余人才靠自贸试验区区域管理机构明确的工作机构出具的推荐函。这是四条路径中最依赖判断的一条,也是内部说明写得清不清楚最要紧的一条:这个人做什么、为什么必须往返。

四条路径的提前量也不一样,这才是做计划的人该记的。人才码当天就能出示;职称加备案名单取决于一条可能需要先修正的记录;同意函取决于另一个部门的日程;推荐函取决于公司愿意为谁背书这一内部决定。

所以把人和路径对上号,不是分类工作,而是排期工作:走最慢那条路的人应当最先启动,哪怕他们出差最少。

备案名单,以及递交之前该做什么

备案名单值得单独说,因为大多数申请是在这里悄悄通过或者悄悄卡住的。

名单由主管部门维护,不由申请人维护。一家从未报送过科研人员信息的单位,或者一项评下来却没有登记的职称,都可能把一个完全合格的人挡在比对之外。补救是行政性的,而且慢:得先把源头的记录修正,签注申请才有意义。

所以真正的准备并不体面。确认机构在相应名单内;确认职称证书上的姓名与用人单位名称一字不差;确认这个人的实际岗位与所要主张的类别一致。同一单位有多名候选人的,应当把材料放在一起核一遍,如果卡住的正是名单,有用的做法是问主管部门需要什么才能把机构或者职称记录下来,并让这件事与签注准备并行,而不是串行。实务中两条线可以同时走,材料一边凑,记录一边补。

把往来记录留好。机构何时申请列入、提交了什么,这份带日期的记录,把材料放在一处。真正有用的不是签注本身,而是它背后的痕迹:开具的在职证明、登记的职称证书、确认过的名单条目、收到的推荐函或者同意函。三年后续办时,当初办这件事的人可能已经离职。

给每位员工建一页索引——类别、路径、材料、日期、发证机构联系人——建起来花一小时,第二个周期能省下一周。

一至五年,与一至三年

杰出人才中经认定为 A、B 类的,可以签发一至五年多次签注;C、D 类以及其余五类人才,可以签发一至三年多次签注。

这个切分与背后的认定逻辑一致。A、B 类是本市最高一档的认定,更长的有效期替这些人省掉一项反复出现的行政琐事;对其余的人,三年足够覆盖一个任期,也短到可以复核。

对用工规划来说,实际后果是企业不能假设所有内地骨干的签注在同一天到期。把人才签注当成一次性的续办事件来排期,两年之内就会失序,而最容易受影响的,一张简单的表就能解决大部分问题。列出每位签注持有人、类别或档次、签发日与到期日,并把复核点设在最早到期日之前足够早的位置。再加一列写通行证,因为两者很少同时到期。

这项练习还会暴露一个接班问题。如果唯一持有五年签注的人,同时也是唯一处理香港仲裁的人,这家公司就有一个单点故障,人员流动政策应当在它变成排期危机之前处理掉。

三十天停留上限:它不是工作许可

每次在香港或者澳门停留不超过三十天。这个上限决定了文件怎么用:它适合短程、高频的往返——开一次董事会、出一场庭、看一次现场、过一个周末——而不支撑派驻。

这条限制是政策用途最清楚的信号。它服务的是需要隔三差五在南边待上几天的人,而不是需要住在那里的。在现场盯三个月安装的工程师,或者被派到香港关联公司常驻一年的经理,需要的是完全不同的安排。

换个角度看,这个上限在真正要紧的地方是宽松的。三十天足以覆盖那种实际产生摩擦的模式:同一个人一个季度往返四五次,每次几天,企业也要为签注覆盖不了的行程做安排。为期数月的现场安装、一个学期的授课、一年的派驻、举家搬迁,都超出了三十天的停留上限,手里有一份人才签注也不会让它们变得可行。这份文件解决的是频率,不是时长;内部指引若把两者混为一谈,问题会在口岸而不是在窗口出现。

检验办法很简单:数一数住宿的夜数。两到五夜、反复出现,符合;以月计的,不符合。

法律人才:范围最窄的一类

法律这一类值得细读,因为它是六类中最窄的,也最容易被误用。

它覆盖参与香港或者澳门法律仲裁程序的内地仲裁员、处理内地与港澳投资争端的内地调解员、具有涉外法律服务业务需要往返港澳的律师,以及在港澳有业务往来的仲裁机构管理人员和仲裁秘书。每一类里,起决定作用的不是职业身份,而是与港澳案件之间的连接。

没有跨境案源的诉讼律师,资历再深也不符合;而一家有香港案源的仲裁机构里,年轻的仲裁秘书可能符合。这种倒挂会让习惯按级别判断的企业感到意外,也正是同意函要由市司法局而不是用人单位出具的原因:对律所与仲裁机构来说,含义是按案源而不是按级别报人。只做国内诉讼的合伙人可能不符合;主办香港仲裁的顾问几乎肯定符合;管理这些案卷的秘书,可能比更资深的同事更符合。

机构也应当保留底层证据——案号、指定函、机构自身关于港澳业务的记录——因为同意函是就一个必须有人能描述清楚的连接出具的。

材料、收费与时限:一条办事动线

基础材料是有效的《往来港澳通行证》、符合标准的照片、居民身份证原件,加上前面所说的类别证明。现役军人和登记备案的国家工作人员,还需按人事管理权限取得审批同意的出入境证件函。

往来港澳通行证每证六十元。多次签注按期限分档:不超过一年每件八十元;一年以上两年以下一百二十元;两年以上三年以下一百六十元;三年以上二百四十元。

通行证与人才签注一并申请,七个工作日内签发;单独申请人才签注,五个工作日内签发。紧急商务、科研交流、探望危重病人等情形可以走绿色通道;特殊情况经批准可以延长至二十个工作日。非重庆户籍人才办理通行证及签注,二十日内签发。

受理地点是重庆市公安局出入境管理局办证大厅,地址在两江新区金石大道三百一十一号,咨询热线 12367。第一次批量办理的企业,可用的顺序是:先确认片区地址,再确认类别与对应名单,费用通常不是约束。二十人一批、按最高档收费,相对于它所替代的那些行程价值有限;批量办理的好处很具体。窗口按人受理,但准备工作是共享的:同一套在职证明模板、同一套翻译口径、同一份说明公司与片区地址的封面函。把前三份申请当成一批来做,格式学一次,之后反复用。

它还会带来一个有用的副产品:谁符合、谁接近符合、谁不符合的名单——这正是一份人员流动政策需要的那份文件。

关于办证大厅,两点实务提醒。预约量在季度末与开学前后会集中,那正是家庭出行的时段;热线更适合用来确认某一类别需要哪些材料,而不是用来抢号。

这项政策真正服务的是谁

把六类放在一起看,画像就出来了:合资企业里拿着副高级以上职称的副总经理、市属科研院所的实验负责人、做跨境投资争议的合伙人、中外合资医院的主任医师。

对这些人的实际收益是时间:过去每次出差都要重新申请一遍的事,变成一份多年有效、可以多次往返、停留期明确的证件。企业不妨把它当作一项流动便利,同时核对一下自己的内地骨干里,有谁的职称刚好落在某一类里。

也需要把不覆盖谁讲清楚,因为这一点在内部沟通中最容易走样。同一间办公室里的外籍员工不在政策范围内,他们往返港澳走的是自己的签证与证件安排;港澳居民回港澳,本来也不需要内地签注。一则把政策描述成”经常去香港的同事的新福利”的通知,会被外籍员工理解为与自己有关,事后更正所损失的信任,比当初那句提醒更多。

准确的表述更窄,也更有用:这是给特定类别、特定片区内的内地员工、用于短程高频往返的一份证件。这样讲,也顺带告诉落在范围外的人该做什么——去问自己的安排,而不是等一项不会到来的政策。

与 APEC 商务旅行卡等通道的区别

APEC 商务旅行卡是另一件工具,服务另一个目的:它便利持卡人进入各参与经济体,包括短期停留,针对的是跨境国际旅行,而不是内地与香港、澳门之间的往来。

人才签注正好相反。它由内地机关签发给内地居民,用于前往本国的两个特别行政区,判断标准是专业类别而不是商务需要。同一位高管同时持有两件工具,不是重复,而是覆盖了两段不同的行程。

同样的逻辑也把人才签注与普通商务签注区分开。两者都签在同一本通行证上,但商务通道与单位的备案和某一次具体行程绑定,用一句话概括:商务签注回答的是”这家公司需要这个人跑这一趟”,人才签注回答的是”这个人属于本片区决定要让他流动起来的那一类”。前者是一笔交易,后者是一种身份,材料的不同由此而来。

同时管理两条通道的企业,合理的分工是:范围外员工的一次性出差走商务通道,范围内的常旅客走人才通道。两者在时间上也有差别,值得知道。人才签注是针对一种持续存在的身份签发的;商务签注是针对一次已经安排好的行程签发的。周一早上才发现需要的公司会看到:对一个尚未进入类别的人来说,商务通道更快;而对一个已经进入类别、一年要跑多次的人来说,人才通道更省。

两者互不替代,也都不替代通行证本身。由此可以得出一条简单的判断规则。人在某一类别里、一年要跑好几次的,人才通道在行政时间上就回本了;人不在类别里,或者这一趟属于例外,商务通道才是正确工具,再怎么分析类别也改变不了。

两者都可能适用时,用”需要持续多久”来定:以身份为基础的文件适合会存在多年的岗位,以行程为基础的文件适合会结束的项目。

企业第一步该做什么

从边界开始。以书面方式确认用人单位的地址落在片区之内,并分清哪些员工是与该主体签约,而不是与重庆其他关联公司签约。

然后把人分进六类,逐类标注适用哪一条证明路径。这项练习通常会暴露一个事实:科研、文教、卫健三组的人数比预想的多,杰出人才这一组比预想的少,因为后者依赖一项多数人从未申请过的市级认定。

最后修记录。有职称但没备案、岗位说明写得比实际职责更窄、机构从未报送过名单,是反复出现的三种失败。每一种都可以补救,最后说一句排序。政策是新的,名单是行政性的,任何一家企业的第一批申请,都会教它重新认识自己的档案。给第一批留出一个季度的余量,不是悲观,而是对试点头几个月运行方式的准确判断。

过了这一段,流程会变成常规:每年核一次边界,每年对一次签注日历,最后一条实务提醒:在凑材料之前,热线 12367 是问清某一类别的最快方式。问一句某项职称是否够格、经办人会看哪些证明,通常五分钟,往往省下两周的准备。

本文不构成对个案的任何法律意见。六类是行政口径,由经办人对照材料适用;某一个人到底符不符合,答案在办证大厅,而不在对政策的解读里。

来源

  • 重庆市公安局·中国(重庆)自由贸易试验区往来港澳人才签注政策及《申请指南》:https://gaj.cq.gov.cn/sy_245/bmdt/202611/t20261112_16401908.html
  • 重庆市人民政府·重庆自贸试验区试点实施往来港澳人才签注政策:https://www.cq.gov.cn/ywdt/zwhd/bmdt/202611/t20261113_16404467.html
  • 重庆日报·往来港澳人才签注受理范围与办理指引:https://www.cqrb.cn/jicengshijian/2026-11-12/2819796_pc.html

Hengqin’s New Recognition List: Overseas Qualifications, Filing, and a Subsidy Window Closing on 31 December

  • Hengqin’s talent bureau issued the 2026 edition of the overseas professional qualification recognition list on 11 November 2026, effective on publication and replacing the 2025 edition.
    横琴粤澳深度合作区人才发展和社会保障局于 2026 年 11 月 11 日印发境外职业资格便利执业认可清单(2026 年版),自发布之日起实施,同时废止 2025 年版。
  • The list lets holders serve in the zone under the relevant implementation measures, through practice filing, registration or exemption from specific examination subjects.
    清单允许持证人员按相关实施办法在合作区提供专业服务,路径包括执业备案、登记以及专业科目免考。
  • Overseas practice experience may be treated as equivalent to mainland experience, which affects how a career is counted rather than who may sign.
    境外从业经历可视同境内从业经历,影响的是职业年限如何计算,而不是谁有权签字。
  • A one-off subsidy of 30,000 yuan for Macao professionals is open for applications from 11 November to 31 December 2026, and late filings are not accepted.
    澳门专业人才一次性补贴 3 万元,申报期为 2026 年 11 月 11 日至 12 月 31 日,逾期不予受理。
  • Four conditions must all hold: Macao residency, a one-year contract with six months of tax payments, six months of qualifying work, and a listed qualification with a permit to serve.
    四项条件须同时满足:澳门居民身份、一年及以上劳动合同并连续缴纳个税六个月以上、从事相关工作六个月以上、持清单内资格并取得执业许可。
  • The list is maintained by annual replacement, so a professional should treat the current edition as the only operative reference.
    清单按年度替换维护,专业人员应把现行版本当作唯一有效依据。

Hengqin’s New Recognition List: Overseas Qualifications, Filing, and a Subsidy Window Closing on 31 December | 横琴新版认可清单:境外资格、执业备案与年底关窗的补贴

What changed on 11 November

On 11 November 2026 the talent development and social security bureau of the Hengqin Guangdong-Macao Deep Cooperation Zone issued the 2026 edition of the list of overseas professional qualifications recognised for convenient practice in the zone. It takes effect from the date of publication. The 2025 edition, issued as 粤澳深合经发通〔2025〕36 号, is repealed on the same day.

Two features deserve attention before anything else. It is a replacement, not an addition: from 11 November the operative document is the 2026 edition, and a file that cites the 2025 notice cites a repealed instrument. Project proposals, cooperation agreements and internal compliance manuals drafted earlier in the year will all need the citation updated, and a surprising number of organisations discover this only when a counterparty asks for the current version.

The second feature is structural. The list is published as an attachment to the notice, so its scope is settled qualification by qualification rather than by the covering text. A professional who wants to know whether a particular licence is covered has to open the attachment, not read the notice. That sounds like a formatting detail, but it changes the order in which people work: check the list first and then the measures, or read the measures first and come back to the list, and the two routes can produce different conclusions.

The reason is that three documents have to line up. The covering notice tells you the mechanism — filing, registration, exemption. The attachment tells you whether your qualification is inside the scope. The implementation measures for the profession tell you what to do next and in what order. Where any one of the three is missing, a service sits in a zone with no clear basis.

The legal basis in the Master Plan

The notice grounds itself in a single sentence of the Master Plan for the Guangdong-Macao Deep Cooperation Zone. Professionals holding Macao and other overseas practice qualifications in finance, construction, planning and design may, where they meet industry regulatory requirements, provide services in the zone after filing, and their overseas practice experience may be treated as equivalent to mainland experience. The same sentence supports convenient measures encouraging medical professionals with Macao and other overseas qualifications to obtain mainland practice qualifications lawfully.

Two ideas sit inside that sentence, and they are not the same idea. Filing is a route to service without a mainland licence: the professional registers with the competent authority and works within a defined scope. Experience equivalence is a rule about counting: it affects how years of practice are treated when a person applies for a title, a grade, or a post that requires domestic experience.

There is also a difference in grammar worth noticing. For finance, construction, planning and design, the Master Plan says professionals may serve after filing. For the medical field, it says the zone should encourage professionals to obtain mainland practice qualifications lawfully. The first limb lands on filing; the second lands on obtaining a licence. A reader who treats the list as a single mechanism will overestimate what it does for the second group.

Read together, the pair lowers the cost of carrying a practice across the border without pretending that the mainland licensing requirement has disappeared. That distinction matters more than it looks, because the two are frequently conflated in material written about the zone, and the conflation produces two opposite mistakes: assuming filing equals exemption from examination, or assuming equivalence of experience equals automatic qualification.

What the list actually grants

The notice permits holders of qualifications on the 2026 list to provide professional services in the zone “in accordance with the relevant implementation measures”. That phrase carries the weight. The list is the door; the profession’s measures are the handle.

Three mechanisms recur across the professions covered: practice filing, registration, and exemption from specific subjects of a professional examination. Which one applies depends on the profession and its regulator, and the notice does not choose for you. A qualification on the list does not by itself create a right to sign opinions, appear before a court, or hold a licensed post. It creates eligibility for whichever route that profession’s measures describe.

The practical reading is that the list is a scope document. It answers whether someone with a given qualification may serve here at all, and leaves the terms to the profession-specific rules. For an employer building a team in the zone, the due-diligence question is therefore not only whether the qualification appears, but which mechanism applies and whether the candidate has completed it.

This also explains a confusion that surfaces in almost every intake conversation. Two qualifications on the same list can differ sharply in how quickly they convert into practice. One may require only a filing with complete documents; another may require an entry on a register that depends on a regulator’s review cycle. The list does not show that difference, because it is not designed to; only the profession’s measures do.

Filing versus registration: two different doors

Because the notice groups three mechanisms into one phrase, it is worth separating them, since the paperwork and the timeline differ.

Filing produces a record. The professional submits documents to the competent authority, the authority accepts the record, and the professional may then work within the scope the filing describes. The controlling documents are the filing itself and the scope it names. A filing is usually the fastest of the three, and it is also the one most sensitive to scope: work outside the filed scope is not covered by the filing.

Registration produces an entry on a register. The professional is listed, and the right to practise follows from being on that list. Registration typically involves a review rather than a receipt, so the timeline depends on the regulator rather than on the applicant’s readiness.

Examination exemption produces a decision on specific subjects. It is the narrowest of the three in the sense that it addresses part of a qualification process rather than the right to serve directly, and it matters most where the profession requires a mainland examination as the route to a licence.

Two of the three routes lead to service without a mainland licence; the third is a step inside the licensing route. A professional who needs to advise a client next month and a professional who is building a ten-year practice in the zone may therefore need different routes even where they hold the same qualification. Reading the list alone will not tell them which.

Finding out which applies is a matter of asking the right authority the right question. The useful question is not “am I recognised” — the list answers that — but “what does my profession’s implementation measure require me to complete before I can serve”. In some professions the answer is a filing with a defined scope; in others it is an entry on a register maintained by a regulator; in others it is a decision on exempted subjects followed by the ordinary licensing steps. The same qualification can sit on the list while the route behind it differs entirely from the route behind a neighbouring qualification on the same page.

That has a planning consequence that is easy to miss. If the route is a filing, the professional controls the timeline, because the work is documentary. If the route is a register, the regulator controls it, and the professional’s preparation only sets the earliest possible date. Firms that budget three weeks for everything, on the assumption that all three mechanisms behave like a filing, are the ones that miss the window.

Counting experience: where equivalence is used

The second half of the Master Plan sentence — that overseas practice experience may be treated as equivalent to mainland experience — is the part most often overlooked, because it produces no certificate. It changes arithmetic rather than permission.

The places where that arithmetic matters are familiar to anyone who has worked across the border. A title or grade review asks how many years of relevant practice the candidate has, and whether those years were spent in a recognised setting. A tender or a client mandate may specify a number of years of domestic experience in the field. An internal promotion matrix may count years in the profession, with a question about whether overseas years count the same.

Equivalence addresses that question. It does not convert a Macao qualification into a mainland one, and it does not create a right to sign. What it does is allow the years behind the qualification to be counted when someone is being assessed — which is often the difference between a candidate who looks junior on paper and one who looks experienced.

For employers, the practical implication is documentary. If equivalence is going to be relied on in a title application or a tender submission, the evidence of the overseas years should be assembled at the same time as the filing, not reconstructed years later from memory and bank records. The list tells you that equivalence is available; the file you keep is what makes it usable.

The subsidy and its four conditions

On the same day the bureau opened the 2026 round of applications for the cross-border practice subsidy for Macao professionals. The application period runs from 11 November to 31 December 2026, and applications after the close are not accepted. The grant is a one-off 30,000 yuan.

Four conditions must all be met. The applicant must hold Macao resident status. The applicant must have a labour contract of one year or more with an employer in the zone, must actually work for that employer, and must have paid individual income tax continuously for six months or more, with the employment relationship and the zone tax payments still in place on the date of application. The applicant must have actually worked in the zone for six months or more in technical or managerial work consistent with, or close to, the qualification held. And the qualification must appear on the subsidy catalogue, with a permit to provide the relevant service in the zone already obtained — by filing, registration, another form of registration, or examination.

Two details are easy to miss. The six-month tax test is counted backwards from the application period, so the qualifying months have to sit before the filing, not after it; someone who joins in the final quarter of the year is, in practice, looking at the next round. And social insurance, where it is referenced, means contributions made in the zone in the employer’s name under law — a payment made elsewhere does not substitute.

The services covered by the round are professional services in finance, accounting, taxation, law, medical care and education, consistent with the catalogue. The subsidy is administered under the zone’s measures and implementing rules for high-quality talent development, and the catalogue itself is updated with the annual application guidance rather than frozen once.

Why applications get returned

The four conditions are short, but each has a failure mode that appears repeatedly in practice. Naming them is more useful than restating the conditions.

The employment record breaks. A contract of one year or more is a condition, and so is continuous tax payment for six months with the relationship still live on the application date. A change of employer during the qualifying period, a gap between contracts, or a switch from one entity in the zone to another can break continuity even where the person never stopped working.

The scope does not match the certificate. The condition is work consistent with, or close to, the qualification held. A professional who holds a qualification in one specialism but spends the qualifying period on work in another will have difficulty showing the match, even if the work itself is senior.

The permit is incomplete. The condition requires a permit to provide the service in the zone, obtained through filing, registration, another form of registration, or examination. An application that shows the qualification but not the permit, or a permit that covers a narrower scope than the work performed, will not meet it.

The two catalogues are confused. The recognition list decides whether a qualification may be used to serve in the zone. The subsidy catalogue decides whether the holder may claim the grant. They are separate documents, and both have to line up; a qualification that appears on one but not the other produces an application that cannot succeed in its current form.

Deadlines, the platform and the two clocks

Applications are submitted through the zone’s platform for enterprises and residents, under the project application section, following the guidance published with the notice. The notice advises against filing on the closing day, when the system is at its busiest — advice that reflects experience rather than caution, since a submission that fails on 31 December has no remedy.

Two different clocks are running here, and confusing them is the most common way to miss the one that actually expires. The subsidy has a hard close on 31 December 2026 for the 2026 round; there is no late window and no discretion written into the notice. The list, by contrast, has no expiry date in the notice: it remains in force until replaced, in the same way that the 2025 edition was replaced by the 2026 edition.

The practical consequence is a planning rule. Work backwards from 31 December: six months of qualifying work and six months of tax payments have to be complete before the application, the permit has to exist before that, and the qualification has to be on both catalogues before any of it starts. For anyone not already partway through that sequence, the realistic target is the following round.

How the list is maintained year to year

The most useful thing to understand about this instrument is that it is a series, not a one-off announcement. The 2026 edition replaced the 2025 edition in full, and the notice repealed the earlier document on the same day rather than leaving both in force. That design has consequences for anyone who relies on it.

First, citations age quickly. A contract, a marketing brochure or a job description that refers to the 2025 list refers to something that no longer exists. Second, the scope can change in either direction between editions, so a qualification that was covered last year is not guaranteed to be covered this year, and one that was absent may have been added. Third, the maintenance model is consistent with how the zone handles the subsidy catalogue, which is updated with the annual application guidance rather than fixed permanently.

For a firm, the operational answer is a small one: keep a note of which edition each person’s file was checked against, and re-check at the start of each year. For an individual, the answer is simpler — treat the current edition as the only operative reference, and treat anything written before 11 November as background.

One question follows immediately, and the notice answers it only indirectly: what happens to someone who filed under the 2025 edition. Because the earlier notice is repealed rather than amended, the practical instruction is to re-check the qualification against the 2026 list and confirm the permit still matches the work being done. A person whose qualification remains listed and whose filing covers the same scope is, in substance, in the same position as before. A person whose qualification has moved, or whose work has drifted outside the filed scope, is not — and that is exactly the case a re-check is meant to catch.

The notice does not set out a separate transition procedure, so the conservative reading is the right one: no automatic carry-over should be assumed, and the file should be re-verified rather than left as it was.

What it means for small Macao firms

The measures are written for individuals, but their commercial effect lands on small professional firms — the accounting practices, tax advisory shops and law offices that serve clients on both sides of the border.

For those firms, the zone has always presented a specific problem. A partner qualified in Macao can advise a client on Macao law from Macao, but serving a client inside the zone requires a basis to do so. Filing provides that basis without requiring the partner to sit a mainland examination, and the recognition list is what makes the filing possible for a defined set of qualifications.

The economics are modest but real. A one-off grant of 30,000 yuan does not transform a practice, and it is not designed to; it offsets the cost and friction of the first year — the filing, the tax registration, the travel, the time spent building a client base that is not yet paying. For a two- or three-partner firm, that is often the difference between opening a small office in the zone and continuing to serve clients from a distance.

The constraint is the same as everywhere else in this policy: the firm still has to check that the specific qualification its partners hold is on the list, that the mechanism for that profession is one it can complete, and that the work it intends to do sits inside the scope the filing covers. A firm that assumes the list covers “accountants” in general, without checking the specific qualification, is the one most likely to be disappointed.

There is a further point about how small firms actually use the zone. A two-partner practice rarely moves. What it does is add: one partner files, spends part of each week in the zone, and builds a client group that needs a Macao-qualified adviser physically present for meetings, filings and negotiations. The grant is calibrated to that pattern — a one-off payment that offsets the first year’s friction, not a salary subsidy that supports a permanent relocation.

That also means the compliance burden is lighter than it first appears, provided the scope is respected. A firm that keeps one person’s filing narrow, and refers work outside the scope back to the Macao office, has a manageable file. A firm that treats the filing as a general permission to practise will accumulate mandates it cannot support, and the exposure surfaces later, when a client or a counterparty asks for the basis.

Two catalogues, two purposes

There are two lists in this policy area and they are routinely confused, which is why they deserve a section of their own rather than a footnote.

The recognition list — the one issued on 11 November — decides whether a qualification may be used to serve in the zone. It is the document that opens the route, whether that route is filing, registration or examination exemption. Its unit of analysis is the qualification.

The subsidy catalogue — the one referenced in the grant notice — decides whether the holder of a qualification may claim the one-off payment. Its unit of analysis is the qualification as well, but its purpose is different: it defines the population that the incentive is meant to reach, and it is updated with the annual application guidance rather than fixed once.

The two overlap substantially, which is what makes the confusion understandable, but they are not the same list and they are not maintained by the same process. A qualification that appears on the recognition list but not on the subsidy catalogue supports a practice but not a claim. A qualification on neither supports neither.

For a professional, the sequence is therefore: check the recognition list, check which mechanism applies, complete the mechanism, check the subsidy catalogue, and only then plan a claim. For a firm advising a candidate, the same sequence applies, and the most common error is to start from the subsidy — because the money is the visible part — and work backwards into a route that turns out not to exist for that qualification.

There is a practical test that resolves most of the confusion in a single step. Ask whether the person could serve a client in the zone next month without the grant. If the answer is yes, the recognition route is in place and the subsidy question is separate. If the answer is no, the subsidy is not the problem to solve first.

The test is worth applying at the level of the individual rather than the firm, because the two catalogues can produce different answers for two colleagues sitting in the same office. One may hold a qualification that is on both, another one that is on the recognition list only, and a third one that is on neither but is nevertheless the firm’s most experienced adviser. The policy addresses qualifications, not seniority, and a staffing plan built on seniority alone will not map onto it.

Who the list is not for

A recognition list invites over-reading, and it is worth being explicit about what it does not do.

It does not create a licence valid outside the zone. A professional who files in Hengqin is not thereby entitled to practise in another mainland city, and the notice makes no such claim. The geographic scope is the cooperation zone, and a client outside it is outside the scope of the filing.

It does not convert an overseas qualification into a mainland one. The holder remains a person with a Macao or other overseas qualification who is permitted to serve in the zone under defined conditions. Where a client, a court or a regulator requires a mainland credential, the list does not supply it; the licensing route does.

It does not cover every profession. The Master Plan sentence names finance, construction, planning, design and the medical field, and the list defines its own scope through an attachment. Professions outside that scope are not assisted by the instrument, however similar the work may look from the outside.

It does not remove the regulator. Filing and registration are interactions with a competent authority, not notifications to nobody. A filing can be refused or returned for incompleteness; a registration can take time. The list reduces the number of doors, not the fact that a door has to be opened.

And it does not settle the commercial question. Whether a practice in the zone makes sense depends on clients, pricing, competition and the cost of running two offices. The policy lowers the entry cost; it does not create demand. Firms that treat recognition as a business case tend to discover the difference in the second year.

What a firm should put in place before the next round

For firms that will not complete the 2026 sequence in time — which, given the six-month tests, is most of them — the useful work now is preparation for the following round. Four items cover most of it.

A per-person file. For each professional: the qualification certificate, the edition of the recognition list it was checked against, the mechanism chosen, the filing or registration record, the scope it covers, the labour contract, and the tax records. Kept together, this file answers both the subsidy question and any later question about the basis on which a service was delivered.

A calendar that starts from the end. The binding constraint is six months of qualifying work plus six months of tax payments, both completed before an application. Working backwards from a closing date gives a start date; firms that work forwards from a hiring date usually arrive late.

A scope note. For each filed professional, a short written description of what the filing covers and what it does not. This is the document that prevents a partner from accepting a mandate that sits outside the scope, which is the failure that matters most commercially.

A client-side explanation. Clients ask why a Macao-qualified adviser can act in the zone. A one-page explanation of the filing basis, the scope and the limits is more useful than a policy citation, and it is the kind of document that a client’s own compliance team will ask for.

None of this requires the grant to exist. All of it makes the grant, when it is available, a matter of filing rather than reconstruction.

How this fits the wider GBA talent picture

Hengqin’s instrument is narrower than a mutual recognition agreement and more concrete than a statement of intent. It creates no licence valid across the Greater Bay Area; it creates, for a defined set of qualifications, a documented route to serve inside one zone, plus a cash incentive for Macao residents who use it within the year.

That places it differently from the approach taken elsewhere in the region. Some platforms in the Greater Bay Area work through equivalence directories that map an overseas qualification onto a domestic vocational grade or professional title, so that the holder can apply for the corresponding domestic credential. Hengqin’s list works instead by opening a route to serve, with the qualification itself remaining the overseas one. The two designs are complementary rather than competing: one converts credentials, the other permits service.

The relationship with the region’s individual income tax preference is similar in character. The tax measure operates on the person’s liability; the recognition list operates on the person’s right to work. A professional who qualifies under both gets the benefit of both, and neither substitutes for the other — a point worth stating because the two are sometimes described as if they were a single package.

For a professional the calculation is short: check the 2026 list and the subsidy catalogue, check which mechanism the profession’s measures provide, check whether six months of qualifying work can be completed and evidenced before the window closes. For an employer the value sits in the filed record — the document that shows a client, or a regulator, that the service was delivered on a permitted basis.

The larger point is about how to read a document like this. It is not a gateway that opens once and stays open; it is an annual instrument with a defined scope, a set of profession-specific routes behind it, and a separate incentive attached to one of them. Treated that way, it is genuinely useful to a small firm deciding whether to test the zone. Treated as a general recognition of Macao qualifications, it will produce expectations the text does not support, and those expectations tend to surface at the worst moment — in front of a client who has asked for the basis of the advice.

Sources

  • 横琴粤澳深度合作区人才发展和社会保障局·《关于印发〈横琴粤澳深度合作区境外职业资格便利执业认可清单(2026 年版)〉的通知》,2026-11-11:https://www.hengqin.gov.cn/macao_zh_hans/zwgk/tzgg/gg/content/post_3959289.html
  • 横琴粤澳深度合作区人才发展和社会保障局·《关于开展横琴粤澳深度合作区 2026 年度支持澳门专业人才跨境执业补贴申报工作的通知》,2026-11-11:https://www.hengqin.gov.cn/macao_zh_hans/zwgk/tzgg/gg/content/post_3958312.html
  • 横琴粤澳深度合作区·第 3/2026 号经济发展局规范性文件《横琴粤澳深度合作区支持人才高质量发展奖补实施办法》(第十九条、第二十条为跨境执业补贴条款):https://www.hengqin.gov.cn/lab/gfxwj/all/content/post_3931634.html

横琴新版认可清单:境外资格、执业备案与年底关窗的补贴

11 月 11 日变了什么

2026 年 11 月 11 日,横琴粤澳深度合作区人才发展和社会保障局印发《横琴粤澳深度合作区境外职业资格便利执业认可清单(2026 年版)》,自发布之日起实施。原《横琴粤澳深度合作区境外职业资格便利执业认可清单(2025 年版)》,即粤澳深合经发通〔2025〕36 号,同日废止。

有两件事要先看清。第一,这是替换,不是增补:从 11 月 11 日起,生效的是 2026 年版,任何仍在引用 2025 年通知的材料,引用的都是一份已废止的文件。年内的项目建议书、合作协议、内部合规手册,只要写了旧文号,都需要更新;不少机构是在交易对手索要现行版本时才发现的。

第二,这是结构问题。清单本身以附件形式发布,适用范围要逐项核对,不能只看正文。想知道某个具体资格是否在列,得打开附件,而不是读通知。这听上去像排版细节,实际改变的是工作顺序:先查清单再查办法,和先读办法再回查清单,两条路径得出的结论可能不同。

原因在于三份文件必须对得上。正文告诉你走哪条路——备案、登记、免考;附件告诉你手里的资格在不在范围内;该专业的实施办法告诉你下一步做什么、按什么顺序做。三份缺一份,服务就落在一块没有明确依据的地带。

总体方案里的依据

通知的依据来自《横琴粤澳深度合作区建设总体方案》中的一句话:允许具有澳门等境外执业资格的金融、建筑、规划、设计等领域专业人才,在符合行业监管要求条件下,经备案后在合作区提供服务,其境外从业经历可视同境内从业经历。同一句话还支持在合作区采取便利措施,鼓励具有澳门等境外资格的医疗领域专业人才依法取得境内执业资格。

这句话里有两个概念,不能混为一谈。备案是一条”不取内地牌照也能执业”的路径:向主管部门完成备案,在核定范围内提供服务。从业经历视同,是一条关于”怎么算”的规则:它影响的是年限如何被认定,在评职称、评级,或应聘要求境内经历的岗位时起作用。

还有一处措辞差别值得注意。对金融、建筑、规划、设计,总体方案说的是”经备案后可以提供服务”;对医疗,说的是”鼓励依法取得境内执业资格”。前一句落在备案,后一句落在取证。把清单当成单一机制的读者,会高估它对第二类人的作用。

合起来看,这两条降低了把一门执业带到河对岸的成本,但并没有假装内地的牌照要求已经消失。这个区分比看上去更重要,因为在关于合作区的宣传材料里,两者经常被揉在一起,而混淆会带来两种相反的误解:以为备案等于免考,或者以为经历视同等于自动取得资格。

清单实际给了什么

通知的表述是:允许持有 2026 年版清单内境外职业资格的专业人员,按照相关实施办法在合作区提供专业服务。这句话的分量全在”按照相关实施办法”上。清单是门,各专业的实施办法是门把手。

覆盖的专业里反复出现三种机制:执业备案、登记、以及专业科目免考。具体适用哪一种,取决于专业和它的主管部门,通知本身不做选择。一项资格出现在清单上,并不自动等于可以签署意见、出庭,或担任需要持证上岗的职位;它带来的是”有资格去走那一条路”。

实务上可以把清单读成一份范围文件:它回答”持这个资格的人能不能在这里执业”,把”按什么条件执业”留给专业规则。对在合作区组建团队的雇主来说,尽职调查的问题因此不只是”资格在不在清单上”,还包括”对应办法规定的是哪种机制,候选人走完了没有”。

这也解释了一个几乎每次沟通都会出现的困惑:同一份清单上的两个资格,落地速度可能相差很远。一个只需备案,材料齐备即可;另一个要走登记,还要等主管部门的审核节奏。清单本身不体现这种差别,因为它不是为此设计的;只有翻到各专业的实施办法才看得出来。

备案与登记:两扇不同的门

通知把三种机制放在一句话里,值得拆开看,因为材料和周期都不一样。

备案产生一份记录。专业人员向主管部门提交材料,主管部门接受记录,此后可在备案写明的范围内执业。起控制作用的文件是备案本身和它列明的范围。备案通常是三者中最快的,也是对范围最敏感的一种:超出备案范围的工作,不受这份备案覆盖。

登记产生登记簿上的一条记录。专业人员被列入名册,执业权利来自”在册”这一事实。登记通常包含审核而不只是收件,所以周期取决于主管部门的节奏,而不是申请人准备得是否充分。

免考产生的是关于具体科目的决定。三者中它最窄,因为它处理的是取得资格过程中的一部分,而不是直接的服务权;在必须以内地考试作为取证路径的专业里,它最关键。

三条路里,前两条通向”不取内地牌照即可服务”,第三条是取证路径中的一步。一个下个月就要给客户出意见的人,和一个准备在合作区做十年执业的人,即便持同一项资格,需要的路径也可能不同。只看清单,看不出这一点。

经历怎么算:视同用在哪里

总体方案那句话的后半段——境外从业经历可视同境内从业经历——最容易被忽略,因为它不产生任何证书。它改变的是算术,不是许可。

算术在哪里起作用,跨境工作过的人都熟悉。职称或等级评审会问候选人有多少年相关执业经历,以及这些年限是否在可认定的机构里度过。招标文件或客户委托可能规定若干年境内从业经验。内部晋升矩阵会统计从业年限,同时留一个问号:境外那几年算不算。

视同回答的就是这个问号。它不把澳门资格变成内地资格,也不创设签字权;它做的是在评估一个人时,允许把资格背后的年限算进去——而这往往就是”纸面上像新人”和”纸面上像老手”的差别。

对雇主来说,实务含义在证据上。如果要在职称申报或投标材料里用上视同,境外年限的证明材料应当与备案同时准备,而不是几年后凭记忆和银行流水去重建。清单告诉你视同是存在的;你自己留下的档案,才让它变得可用。

判断走哪条路,方法是向对的主管部门问对的问题。有用的问法不是”我有没有被认可”——清单回答的是这个——而是”我这个专业的实施办法,要求我先完成什么才能执业”。有些专业答案是一份写明范围的备案;有些是进入主管部门维护的名册;有些是先就免考科目作出决定,再走通常的取证步骤。同一页上相邻的两项资格,背后的路可能完全不同。

这带来一个容易漏掉的排期后果。如果路径是备案,时间表掌握在申请人手里,因为工作本质是文档;如果路径是登记,时间表掌握在主管部门手里,申请人准备得再好,也只能决定最早可能的日期。把三种机制都按备案来估算、给什么都留三周的机构,正是错过窗口的那一类。

补贴与四项条件

同一天,该局启动 2026 年度支持澳门专业人才跨境执业补贴申报。申报期为 2026 年 11 月 11 日至 12 月 31 日,逾期不予受理。补贴标准为一次性 3 万元。

四项条件须同时满足。申请人须具有澳门居民身份。须与合作区用人单位签订一年及以上劳动合同,在该单位实际工作,且连续缴纳个人所得税六个月以上;截至申请补贴之日,劳动关系仍存续,且仍在合作区缴纳个人所得税。须在合作区用人单位实际从事与其执业资格证书相一致或相近的专业技术及管理工作六个月以上。所持执业资格须在《支持澳门专业人士跨境执业补贴目录清单》之内,且已取得在合作区提供相关服务的许可——形式包括备案、注册、登记、考试。

两个细节容易漏。六个月纳税期是以申报期往前推算的,合格月份必须在申报之前,而不是之后;换句话说,年底才入职的人,这一轮基本赶不上。社会保险指的是按法律规定以用人单位名义在合作区参加的社会保险,在别处缴的不能替代。

本轮覆盖的服务领域是金融、会计、税务、法律、医疗、教育等专业服务,与目录清单相对应。补贴依据《横琴粤澳深度合作区支持人才高质量发展若干措施》及《横琴粤澳深度合作区支持人才高质量发展奖补实施办法》执行;目录清单则随年度申请指南动态更新,并非一次定死。

申报为什么会被退回

四项条件都不长,但每一条都有固定的失手方式,把失手方式讲清楚,比重复条件更有用。

劳动关系断了。 一年及以上劳动合同是一条,连续缴纳个税六个月且申请之日关系仍存续是另一条。合格期内换过雇主、两份合同之间有空档,或者从合作区一家主体转到另一家,都可能打断连续性,哪怕人一直在工作。

执业范围与证书对不上。 条件要求从事与资格”相一致或相近”的专业技术及管理工作。持某一专业的资格、却在合格期主要做另一专业的事,即便职级更高,也难以说明这种对应关系。

许可不完整。 条件要求已取得在合作区提供相关服务的许可,形式包括备案、注册、登记、考试。只提交了资格证书、没有许可,或者许可覆盖的范围比实际从事的工作窄,都不满足条件。

两份清单混了。 执业认可清单决定”这个资格能不能在合作区执业”,补贴目录清单决定”持证人能不能领钱”。这是两份文件,两份都要对上;只在其中一份里的资格,按现状申报是走不通的。

时限、平台与两个时钟

申报通过横琴粤澳深度合作区惠企利民服务平台提交,进入”项目申报”栏目,按通知发布的指引填报。通知特别提示,不要集中在申报截止当天提交,以免系统拥堵、影响申报——这条提示来自经验而非客套,因为 12 月 31 日提交失败的申请,没有补救窗口。

这里有两个不同的时钟,把它们混为一谈是最常见的错过方式。补贴有硬截止:2026 年度申报到 12 月 31 日为止,通知里没有逾期通道,也没有留下酌情空间。清单则不同,通知没有给它失效日期:在被替换之前持续有效,就像 2025 年版被 2026 年版替换那样。

由此得出一条排期规则:从 12 月 31 日往前倒推。六个月合格工作和六个月纳税要在申报前完成,许可要在那之前取得,资格要在更早的时候就在两份清单上。对尚未走到一半的人来说,现实的目标是下一轮,而不是这一轮。

清单如何逐年维护

理解这份文件最有用的一点是:它是一套序列,不是一次性公告。2026 年版整体替换了 2025 年版,通知在同一天废止旧文件,而不是让两份并行。这个设计对依赖它的人有直接影响。

第一,引用会过期。合同、宣传册、岗位说明里如果写的是 2025 年版清单,写的就是一份已不存在的东西。第二,范围可能在两个版本之间双向变化:去年在列的资格,今年未必还在;去年不在的,今年可能已增补。第三,这套维护方式与补贴目录清单一致——后者随年度申请指南更新,而不是长期固定。

对机构来说,应对办法很小:记录每个人的档案是按哪一版核对的,并在每年年初重新核对一次。对个人来说更简单:把现行版本当作唯一有效依据,11 月 11 日之前写的东西一律当背景材料看。

紧接着有一个问题,通知只间接回答:按 2025 年版完成备案的人怎么办。由于旧通知是被废止而非修改,实务上的做法是按 2026 年版重新核对资格,并确认许可仍然与正在从事的工作匹配。资格仍在清单上、备案覆盖范围没变的,实质上与此前处在同一位置;资格有变化,或实际工作已经漂出备案范围的,就不一样了——而这正是复核要抓的情况。

通知没有另设过渡程序,所以稳妥的读法是:不要假定自动延续,把档案重新核一遍,而不是原样放着。

对澳门中小专业机构意味着什么

措施是写给个人的,但商业效果落在中小专业机构身上——那些服务两地客户的会计师事务所、税务顾问和法律事务所。

对这类机构,合作区一直有个具体难题。一位在澳门取得资格的合伙人,可以在澳门为客户的澳门法律事务出意见;但要在合作区内服务客户,就需要一个依据。备案提供了这个依据,而且不要求合伙人去考内地试;执业认可清单则让备案在一份确定的资格范围内成为可能。

还有一点关于小机构实际怎么用合作区。两三位合伙人的事务所很少整体搬迁。它做的是加法:一位合伙人完成备案,每周有一部分时间在合作区,逐步积累起一批需要在现场开会、递件、谈判的、由澳门资格顾问服务的客户。补贴正是按这种模式校准的——一次性款项,抵消第一年的摩擦,而不是支撑永久迁移的薪酬补贴。

这也意味着,只要守住范围,合规负担比看上去轻。把一个人的备案做得窄、把范围外的工作退回澳门办公室的机构,档案是可控的。把备案当成通用执业许可的机构,会积下一批它支撑不了的委托,而风险会在之后浮现——当客户或交易对手追问依据的时候。

经济账不大,但真实。一次性 3 万元不会改变一家事务所的命运,也不是为此设计的;它抵消的是第一年的成本与摩擦——备案、税务登记、往返、以及建立尚未产生收入的客户基础所花的时间。对两三位合伙人的事务所,这往往就是”在合作区开个小办公室”和”继续远程服务”之间的差别。

约束与其他地方一样:机构仍要确认合伙人持有的具体资格在清单上,该专业对应的机制是它能走完的,打算做的业务落在备案覆盖的范围内。最容易失望的,正是那些以为清单覆盖了”会计师”这个大类、而没有核对具体资格的机构。

两份清单,两个用途

这个政策领域里有两份清单,经常被混为一谈,所以值得单独一节说清楚,而不是放在脚注里。

执业认可清单——11 月 11 日印发的那一份——决定某项资格能不能在合作区执业。它是打开通道的文件,无论通道是备案、登记还是免考。它的分析单位是”资格”。

补贴目录清单——补贴通知里引用的那一份——决定持证人能不能申领那笔一次性补贴。它的分析单位同样是资格,但用途不同:它界定这项激励要覆盖的人群,并随年度申请指南更新,而不是一次定死。

两份清单重合度很高,这正是混淆可以理解的原因,但它们不是同一份清单,也不由同一套流程维护。只在执业认可清单里、不在补贴目录里的资格,支撑得起执业,支撑不起申报;两份都不在的,两样都不成立。

因此,对个人的顺序是:查执业认可清单,查对应机制,走完机制,查补贴目录清单,最后才谈申报。对替候选人做判断的机构,顺序一样;最常见的错误是从补贴倒推——因为钱是最显眼的部分——结果倒推出一条对这个资格并不存在的路径。

有一个一步就能化解大部分困惑的检验:问这个人下个月能不能在没有补贴的情况下为合作区客户提供服务。能,说明执业路径已经就位,补贴是另一个问题;不能,说明补贴不是当下要解决的事。

这个检验要按人来用,而不是按机构来用,因为两份清单可能给同一间办公室里的同事不同答案。甲持有的资格两份都在,乙只在执业认可清单上,丙两份都不在,却是机构里资历最深的那位。政策针对的是资格,不是资历;只按资历来排的用人计划,和这套政策对不上号。

这份清单不是给谁用的

一份认可清单很容易被读过头,所以有必要把”它不做什么”讲明白。

它不创设在合作区之外通用的执照。在横琴完成备案的专业人士,并不因此有权在另一个内地城市执业,通知也没有这样主张。地域范围就是合作区,客户在区外,就落在备案范围之外。

它不把境外资格变成内地资格。持证人仍然是”持澳门等境外资格、在约定条件下获准在合作区服务的人”。当客户、法院或主管部门要求内地证书时,清单提供不了,得走取证路径。

它不覆盖所有专业。总体方案那句话点到金融、建筑、规划、设计以及医疗领域,清单则通过附件界定自身范围。范围之外的专业不受这份文件帮助,无论工作看起来多相似。

它不免除主管部门这一环。备案与登记是与主管部门的往来,不是向无人处通报。备案可能因材料不全被退回,登记可能要等。清单减少的是门的数量,不是”门得有人开”这个事实。

它也不解决商业问题。在合作区执业是否划算,取决于客户、定价、竞争和两地办公的成本。政策降低的是进入成本,不创造需求。把认可当成商业论证的机构,通常会在第二年发现差别。

机构在下一轮之前该准备什么

对赶不上 2026 年这一轮的机构——考虑到六个月的门槛,多数机构都属于这一类——现在该做的是为下一轮做准备。四件事基本够用。

一人一档。 每位专业人士一份:资格证书、核对时依据的是哪一版认可清单、选择了哪种机制、备案或登记记录、记录覆盖的范围、劳动合同、纳税记录。放在一起,这份档案既能回答补贴的问题,也能回答日后”这项服务凭什么可以做”的问题。

从终点倒推的日历。 硬约束是六个月合格工作加六个月纳税,都要在申报之前完成。从截止日倒推得出起始日;从入职日顺推的机构,通常会迟到。

一份范围说明。 每位已备案的专业人士,用一小段文字写清备案覆盖什么、不覆盖什么。这份文件能防止合伙人接下超出范围的委托——那是商业上最要命的一种失手。

一份面向客户的说明。 客户会问:为什么一位澳门资格的顾问可以在合作区执业。一页纸讲清依据、范围和边界,比引用政策条文有用,而且往往是客户自己的合规团队会来索要的那份材料。

这些都不依赖补贴是否存在。但它们会让补贴在可以申请时,变成一次递交,而不是一次重建。

放在大湾区人才版图里看

横琴这份文件的定位,比互认协议窄,比意向表述实。它不创设一张通行大湾区的牌照;它为一份确定的资格清单,开出一条在特定区域执业的有据可循的路径,并给在年内用上这条路径的澳门居民一笔现金激励。

这让它区别于区域内的另一种做法。大湾区有些平台走的是比照认定目录:把境外资格对应到境内职业技能等级或职称,持证人可以据此申报相应的境内证书。横琴的清单走的是另一条路——开一条执业通道,资格本身仍是境外的。两种设计互补而非竞争:一种转换证书,一种允许执业。

与区域个人所得税优惠的关系也类似。税收措施作用在个人的税负上,认可清单作用在个人的执业权上。两项都符合条件的专业人士,两项都能享受,而彼此不能替代——这一点值得说明,因为二者有时被描述成同一个政策包。

对澳门专业人士,算账很短:核对 2026 年版清单与补贴目录清单,核对对应办法规定的是哪种机制,再核对能否在窗口关闭前完成并留下六个月的合格工作记录。对雇主,价值落在备案记录上:那是能向客户或主管部门说明”这项服务是按规定提供的”的文件。

更关键的是怎么读这类文件。它不是一道开一次就一直敞着的门;它是一份按年度发布、范围明确、背后挂着一整套分专业路径、并对其中一条附有单独激励的文件。按这个方式去读,它对一家想试水合作区的小机构确实有用;把它当成”澳门资格被普遍认可”,则会生出文本并不支持的期待,而这些期待往往在最不合时宜的时候浮出水面——在客户追问意见依据的那一刻。

来源

  • 横琴粤澳深度合作区人才发展和社会保障局·《关于印发〈横琴粤澳深度合作区境外职业资格便利执业认可清单(2026 年版)〉的通知》,2026-11-11:https://www.hengqin.gov.cn/macao_zh_hans/zwgk/tzgg/gg/content/post_3959289.html
  • 横琴粤澳深度合作区人才发展和社会保障局·《关于开展横琴粤澳深度合作区 2026 年度支持澳门专业人才跨境执业补贴申报工作的通知》,2026-11-11:https://www.hengqin.gov.cn/macao_zh_hans/zwgk/tzgg/gg/content/post_3958312.html
  • 横琴粤澳深度合作区·第 3/2026 号经济发展局规范性文件《横琴粤澳深度合作区支持人才高质量发展奖补实施办法》(第十九条、第二十条为跨境执业补贴条款):https://www.hengqin.gov.cn/lab/gfxwj/all/content/post_3931634.html

Nansha’s Talent Regulation Turns an Overseas Certificate Into a Chinese Title

  • Guangzhou’s Nansha talent regulation took effect on 1 November 2026 and runs to twenty articles, the city’s second local statute written for the district alone.
    广州市为南沙人才发展制定的这部条例自 2026 年 11 月 1 日起施行,共二十条,是该市专为南沙区量身打造的第二部地方性法规。
  • Article 13 requires a mechanism that compares overseas professional qualifications with China’s vocational skill levels and professional titles, backed by a catalogue that is updated rather than fixed once.
    第十三条要求建立境外职业资格与境内职业技能等级、职称的比照认定机制,并配套一份动态管理的比照认定目录。
  • A certificate on that catalogue lets its holder apply for the corresponding skill level or title; the wording is “may apply”, not “is granted”.
    持目录内证书的人才可以申报对应的职业技能等级或专业技术职称,条文用的是”可以申报”,而不是直接授予。
  • The chain runs from a catalogue entry to a filed application, then to assessment by the human resources and social security authorities; the regulation fixes the machinery but not the timetable.
    链条是:目录在册、提交申报、由人力资源社会保障部门组织评审;条例把机制定下来,但没有规定时限。
  • Articles 14 and 15 make service officers, rental housing, talent apartments, international schools and classes for Hong Kong and Macao children a legal duty, and let the district write and review its own talent household registration rules.
    第十四条、第十五条把服务专员、租赁住房、人才公寓、国际学校与港澳子弟班写成了法定职责,并允许南沙自行制定人才入户办法、自主审核办理。
  • Article 16 commits the police and the district government to smoothing visas, work permits, residence permits and permanent residence for foreign talent, plus a green channel for Hong Kong and Macao endorsements.
    第十六条要求公安机关与南沙区政府为外籍人才办理签证、工作许可、居留许可以及申请永久居留提供便利,并设立办理往来港澳商务签注、人才签注的绿色通道。
  • State organs, public institutions, statutory bodies and state-owned enterprises in Nansha may recruit qualified Hong Kong and Macao talent, a door that was previously closed.
    南沙的国家机关、事业单位、法定机构、国有企业可以招录或招聘符合条件的港澳人才,这在此前是关着的门。
  • Nansha reports a catalogue of 137 international qualification certificates, 248 recognised qualifications across 12 priority industries, and more than 400 Hong Kong and Macao professionals practising locally.
    据权威报道,南沙已建立涵盖 137 项国际职业资格证书的职称对应关系目录,实现 12 大重点产业 248 项境外职业资格认可,超过 400 名港澳专业人士在当地便利执业。

Nansha’s Talent Regulation Turns an Overseas Certificate Into a Chinese Title | 南沙人才条例:把一张境外证书换算成中国职称

A local statute with national significance

On 1 November 2026 the Guangzhou Nansha Talent Development Promotion Regulation came into force. It runs to twenty articles and is the second local regulation Guangzhou has written specifically for Nansha District.

The first thing to notice is what the text does not contain. There is no fund, no application window, no quota and no subsidy table. Anyone who opens it looking for a number to budget against will close it disappointed. What the twenty articles contain instead is a division of labour: named government bodies are told what they must build, and the thresholds, lists and procedures are pushed into catalogues and supporting rules that will be issued later.

That structure is why the regulation matters beyond its own district. A foreign professional’s file in China is rarely decided by one document. It is decided by a chain of them: a qualification, a recognition decision, a title, an employer’s registration, a residence permit. A local regulation that takes one link in that chain and gives it a legal basis changes the terms on which the other links are negotiated. There is a second reason to read it closely, which has little to do with Nansha’s own ambitions. Local legislation is where national policy becomes administrable. The centre sets the framework, the province sets the direction, and a city regulation decides which office keeps the list, which department answers the telephone, and what happens when the two disagree. For a company planning hiring two or three years out, that last layer is the one that decides whether a promise turns into a process.

The timing says something too. Regulations are written when a practice has become stable enough to codify, not while it is still being invented. The fact that Guangzhou chose to legislate on talent development for a single district suggests that the arrangements described in the text have been running long enough for the drafters to know where they break.

What kind of document this is

A local regulation is not a departmental notice. It is adopted through a local legislative procedure and it binds the government bodies named in it, not only the enterprises it talks about. A departmental normative document can be revised by the department that issued it, sometimes quietly, sometimes within a single budget cycle. A regulation has to be amended through the legislature again, and until then it stands.

For a company planning three to five years ahead, that difference is the whole value. The promise “the district will maintain a comparison catalogue and review it periodically” is worth more than a one-off recognition list, because the obligation survives the official who signed it.

The limit is equally important. A local regulation cannot create an entitlement that national law does not allow. It cannot lower the statutory conditions for a work permit or a permanent residence application, and it cannot grant a licence to practise a regulated profession. What it can do is organise how the city uses the discretion it already has — which desk handles a file, which department maintains a list, which channel a family uses. Read within those limits, Article 13 is ambitious; A practical test of the difference is what happens when an official leaves. A departmental document tends to leave with its author: the next incumbent reviews it, and the practice softens or disappears without anyone announcing a change. A regulation stays, and so does the duty to maintain the catalogue, publish the service list and staff the service desks. That continuity is what a five-year hiring plan actually needs.

The limit is just as practical. Because the regulation cannot alter national conditions, it should be read as a map of local discretion rather than a source of new rights. Where it says the district will facilitate something, the facilitation is real; where it is silent, the national rule governs unchanged.

Article 13: from a certificate to a title

Article 13 directs the municipal and district governments to carry out the reform and piloting requirements attached to Nansha’s three designations — a national pilot zone for talent management reform, a demonstration zone for Guangdong-Hong Kong-Macao talent cooperation, and an internationalised talent special zone — and to draw on how Qianhai, Hetao and Hengqin have handled Hong Kong and Macao talent policy.

Then the operative sentence. The municipal human resources and social security department, working with the industry regulators and the district government and taking account of what local industries need, is to establish a mechanism comparing overseas professional qualifications with domestic vocational skill levels and professional technical titles, and is to draw up a comparison catalogue that is managed dynamically.

The consequence follows in a single line: a person holding an overseas professional qualification certificate listed in that catalogue may apply for the corresponding national vocational skill level or professional technical title. Two verbs carry the weight. “Establish” binds the bureaucracy to build the machinery. “May apply” tells the individual what the machinery is for. Neither promises an outcome, and The three designations in the first paragraph are not decoration. A national pilot zone for talent management reform carries authority to try institutional changes other districts cannot; a demonstration zone for Guangdong-Hong Kong-Macao talent cooperation carries a mandate to align rules across the three jurisdictions; an internationalised talent special zone carries a mandate to serve foreign nationals specifically. Article 13 stacks all three as the legal hook for the comparison mechanism.

That is why the article can use “explore” for one half of its subject and “establish” for the other. Exploration covers what depends on agreement with Hong Kong and Macao. Establishment covers what lies inside the city’s own competence. One half waits for counterparts; the other half does not.

The chain from certificate to title

Read as a process rather than a principle, the article describes four steps, and each has its own owner.

The first is the catalogue. Nothing starts until a certificate appears on it, and the list is maintained by the human resources and social security department together with industry regulators. For an applicant, this is a screening question that can be answered before any paperwork is prepared: either the certificate is listed, or the conversation ends there for now.

The second is the application itself. The regulation uses the language of filing — the holder applies for the corresponding grade — which means the initiative sits with the individual and, in practice, with the employer’s human resources team that assembles the evidence.

The third is assessment. The article assigns the mechanism, not the decision, to the human resources and social security authorities. What the assessors weigh, which documents they accept for a qualification earned outside mainland China, and how long the review takes are matters for the catalogue and the implementing rules. The regulation does not fix a deadline, and no deadline should be inferred from it.

The fourth is the outcome: a vocational skill level or a professional technical title. These are two different ladders. Skill levels are graded through assessment arrangements organised around employers and evaluation bodies; titles run through the professional title evaluation system, which is where public institutions, state-owned employers and many talent programmes look when they need a grade. The catalogue has to say which ladder a given certificate maps to, The employer’s role deserves its own note. In most cases the individual holds the certificate, but the employer holds the evidence: the contract, the job description, the record of what the person actually does. Assessors read a file as a story about a job, not as a certificate in isolation, and an application that describes the role vaguely tends to produce a grade that undersells it. Human resources teams that have been through the process once usually rewrite the job description before the second attempt.

What happens when a certificate is not on the list is worth knowing too. The answer is not an appeal against the catalogue; it is either a different route to a grade, or a wait for the next revision. A company with several staff in the same position should check the list once for the group rather than once per person.

What equivalence does not cover

Three boundaries keep expectations calibrated, and all three are visible in the text.

Application is not award. The regulation opens a filing route; assessment and decision stay with the authorities under rules the article does not rewrite. A certificate on the list converts an eligibility question into a procedure, which is a real gain but a procedural one.

A title is not a licence. A professional title says where a person sits in an evaluation hierarchy. Permission to practise a regulated profession — medicine, law, architecture — is a separate authorisation with its own examination and registration route. Article 13 places mutual recognition of practice qualifications alongside mutual recognition of evaluation standards and calls it something to be explored, which is the careful wording of a subject still under negotiation among the three jurisdictions.

And the catalogue moves. Managed dynamically and tied to industry demand, the list that governed a decision two years ago may not be the list that governs the next one. The date to check is the filing date, A further distinction catches people out: recognition and examination are different acts. Recognition asks whether a qualification earned elsewhere is equivalent to something here; examination asks whether a person meets a standard set here. Article 13 is about the first. A regulated profession normally requires the second as well, and no amount of catalogue work substitutes for it.

That is why the article is more useful to some professions than others. For an engineer, an accountant, or a project manager in a sector where titles drive grading and programme eligibility, the mechanism does real work. For a physician or a lawyer, it may smooth one step while leaving the licensing step untouched. Reading the article against your own profession, rather than against its headline, is the only way to know which case you are in.

Borrowed ground: Qianhai, Hetao and Hengqin

Article 13 tells Nansha to draw on the practice of three other platforms in the Greater Bay Area, and the instruction is more than courtesy. It signals that the comparison mechanism is not a Nansha experiment starting from zero; it is a codification of a route that has already been walked elsewhere in the region.

The practical consequence is that the questions have answers. A professional who has been through a similar process in Qianhai, Hetao or Hengqin knows what kind of evidence tends to be accepted, how a title evaluation body reads a foreign certificate, and where a file usually stalls. That knowledge transfers, at least in outline.

What Nansha adds is a statutory basis and a delivery structure that reaches down to the district. The mechanism is assigned jointly to the municipal department, the industry regulators and the district government, which is a different arrangement from a pilot run by a single office. It also ties the work to three designations that carry their own reform mandates, For a professional who has already been through a similar process elsewhere in the region, the value is a shorter learning curve. The questions are the ones that mattered in Qianhai or Hengqin: which body holds the catalogue, whether a translation and notarisation are expected, whether the assessors want the syllabus of the overseas course or only the certificate, and how a grade awarded by an employer compares with one awarded through the evaluation system.

The difference in Nansha is the delivery layer. Because the mechanism is assigned jointly to the municipal department, the industry regulators and the district government, a company with a specific problem has more than one place to take it — and the industry regulator is often the most useful of the three, since it knows what the sector needs, which is the reason the article gives for having a catalogue at all.

Articles 14 and 15: services, housing and schooling

Article 14 is administrative but practical: a published list of talent introduction service items, an online platform alongside physical service stations, and a service officer for anyone entered in the high-level talent database. None of that changes a legal status, but it changes how quickly a newcomer can find out what applies to them.

Article 15 addresses what families ask first. The district may adopt its own rules on household registration for talent and handle applications itself. It is to build international, Hong Kong and Macao, and youth communities; provide rental housing, talent apartments and purchase subsidies; strengthen international schools and schools and classes for Hong Kong and Macao children; and push hospitals toward cross-border medical insurance settlement.

The order of that list is not accidental. Housing and schooling are the two items on which a relocation decision usually turns, and they are the two items a company can least credibly promise on its own. When they appear in a local regulation rather than in an investment agreement, they become duties that can be reported on, The service officer is the least glamorous item in the article and, for a newly arrived family, often the most useful. A named person who can say which window handles a residence permit renewal, which school has a place in the international stream, and which housing scheme the employer’s registration qualifies for is worth more than a portal with fifty links. The regulation pairs the two: a platform for information, stations for presence, and a person for the cases that do not fit a form.

The community provisions work the same way. International, Hong Kong and Macao, and youth communities are not just addresses; they are where a trailing spouse finds a professional network and Schooling deserves one more paragraph, because it is where timetables collide. International schools and schools for Hong Kong and Macao children keep their own admission cycles, language streams and waiting lists, and a family that arrives in March for a September place is already late in a way no residence permit can fix. The regulation does not promise a place. It puts the duty to strengthen those schools into a text the district must answer for, which is the difference between a brochure and a plan.

Continuity is the second point. Children who move between systems lose time when the curricula do not line up, and classes for Hong Kong and Macao children exist to soften that transition. An employer negotiating a package should ask which of those options has capacity in the relevant year group, rather than which ones exist in principle.

The household registration power, read closely

Buried in Article 15 is a competence that deserves separate attention: the district may formulate differentiated regional rules on talent household registration and review and process the applications itself.

Household registration is the hinge on which much of the rest turns. It is commonly the precondition for school admission, for purchase eligibility in a controlled housing market, and for the long-term planning that a family does before it commits to a city. A district that can set its own criteria and process its own files can respond to a labour shortage faster than a city-wide queue allows.

The caveat is that this is a power to make rules, not a promise of a quota. The district still has to write the implementing measures, and the criteria will still be published. For an employer, the useful question is not whether the power exists but For a family, the practical sequence usually runs like this: a job offer, a decision about schooling, a decision about housing, and only then a decision about how long to stay. Household registration sits underneath all three. Where a district controls both the rules and the review, it can shorten the interval between the offer and the enrolment — which is often the interval in which a candidate accepts a competing offer instead.

Two cautions are worth keeping. The power is to make differentiated rules, so the criteria will be specific and will change; and the review is autonomous, which means the district’s own reading governs. Employers should ask for the current measures in writing rather than rely on what a colleague was told last year.

Article 16: entry, residence and permanent residence

Article 16 is the one foreign employees will read first. The municipal public security authorities and the district government are to improve consultation on entry and residence policy and language training, and to ease access to visas, work permits, residence permits and permanent residence. The district is also to open a green channel for Hong Kong and Macao business and talent endorsements.

The honest reading is that this creates a service obligation, not a new eligibility standard. The conditions for a work permit, a residence permit or permanent residence remain where they were, in national law and the implementing rules of the immigration authorities. What the article does is connect those conditions to a local delivery system: named desks, a green channel, interpretation and language support, and staff whose job is to explain the rules before a file is submitted rather than after it is rejected.

That connection matters more than it sounds. The work permit system already sorts foreign employees into categories that determine what an employer must show and how long a permit can run. Article 16 does not touch those categories. The work permit system is the piece most often misunderstood at this point. Foreign employees are already sorted into categories that determine what an employer must show and how long a permit may run, and those categories are set nationally. Article 16 does not create a new one and does not promise a faster one. It promises that the local handling of an existing category will be explained, staffed and, where the file is straightforward, moved through a dedicated channel.

The endorsement green channel is a separate item that is easy to overlook. It concerns Hong Kong and Macao business and talent endorsements, which mainland staff need for short trips south. Placing it in the same article as visas and residence permits says something about how the district sees mobility: Permanent residence is the item most often misread. The article speaks of facilitating an application, not of changing the criteria; those are set nationally and are demanding. What a district can add is advice before the file is built, an officer who stays with it, and a realistic account of the residence history the application will need. Companies that read the article as a reason to start earlier, rather than as a reason to expect a different answer, get the most from it.

Public-sector recruitment and its knock-on effects

Separately, the regulation states that state organs, public institutions, statutory bodies and state-owned enterprises in Nansha may recruit or employ qualified Hong Kong and Macao talent.

For Hong Kong and Macao residents this widens the set of employers worth approaching, including institutions that previously had no route to hire them at all. Public-sector posts on the mainland have generally been closed to non-mainland residents, and removing that bar in one district is a structural change rather than a symbolic one.

For firms, the effect is indirect but worth tracking. Where tender documents score the credentials of named staff, or where a client’s project team must include people with public-sector experience, a wider pool of qualified professionals changes what a bidder can credibly offer. The regulation does not mention procurement; it does not need to. It changes who can be on the team.

Two effects follow. The first is on the market for experienced Hong Kong and Macao professionals: a public-sector employer that can now hire them becomes a competitor for the same people that consultancies and corporate functions recruit. The second is on how credentials are read. A curriculum vitae that includes a stint in a mainland public institution reads differently in a tender than one that does not, and the regulation quietly widens the pool of people who can acquire that line.

Neither effect arrives on 1 November. Both depend on recruitment rules, job design, and how personnel systems handle Hong Kong and Macao identity documents. The regulation removes a bar; the operating details decide how wide the opening becomes.

A practical note for firms: the effect is easiest to see in the composition of a bid team. Where a tender asks for named staff with experience of public projects, a candidate pool that now includes people who have worked inside a mainland public institution is worth mapping before the next submission, even if nobody on the current team is affected.

The mapping already in place

According to authoritative reports, Nansha has built a catalogue mapping 137 international qualification certificates to Chinese titles, extended recognition to 248 overseas qualifications across 12 priority industries, and seen more than 400 Hong Kong and Macao professionals take up practice under facilitated arrangements.

Those figures describe the pilot. What the statute adds is durability: a legal basis, a catalogue with a maintenance duty, and obligations on named departments rather than a programme that ends with its funding cycle.

The useful follow-up questions are administrative. Which department maintains the list, how often it is revised, and what changed between the last two versions. Answers to those tell a professional more about whether a certificate is worth pursuing It is also worth noticing what the regulation does not do with these numbers. It does not freeze them, endorse them or turn them into a target. They remain a description of how the pilot has behaved, and the statute’s job is to keep the machinery that produced them running under a legal duty. For a reader, that is a more useful relationship than a promise of a particular total. The second is to distinguish a certificate recognised for practice from one that maps to a title; the reported totals usually count both, and the two do different things for a career.

The regulation’s contribution is that these questions now have an address. A catalogue maintained by named departments, under a duty to keep it current, is a different object from a list published once at the launch of a pilot. It can be queried, compared across versions, and cited in an internal business case — It is worth noticing what the regulation does not do with these numbers. It does not freeze them, endorse them or turn them into a target. They remain a description of how the pilot has behaved, while the statute’s task is to keep the machinery that produced them running under a legal duty.

What a foreign professional should prepare

None of this amounts to a filing checklist, because the checklist has not been published. What can be settled now is where you stand: whether the certificate is listed, which ladder it maps to, and whether the employer’s job description supports the grade being sought. The rest follows from those three answers, and each of them can be obtained in a single conversation with the district bureau rather than through a consultant.

  • Confirm whether your qualification is on the current comparison catalogue, and note the date of the version you checked.
  • Separate two questions: whether you may practise, and what grade you will be placed at. They run on different tracks.
  • Establish which ladder your certificate maps to — vocational skill level or professional title — because the downstream uses differ.
  • Ask the district human resources and social security bureau which body assesses your file and what evidence it accepts for a qualification earned abroad.
  • Involve your employer early; titles are read by employers and by the talent programmes they apply for.
  • If you are a Hong Kong or Macao resident, look at the public-sector employers the regulation now opens, not only the private market.
  • Diarise the catalogue review cycle. A dynamic list invites re-checking, not a one-time answer.
  • Keep a copy of the version you relied on; a later revision will not retroactively explain an earlier decision.
  • If your employer is preparing a title application, check that the job description matches the grade you are asking for.

Sources

  • 广州市南沙区人民政府·《广州市南沙人才发展促进条例》中英双语全文:https://www.gzns.gov.cn/zwgk/rdzt/hgazczl/cyzc/hgazxzc/rc/content/post_11083998.html
  • 广东省人民政府·11 月起施行法规解读(含南沙境外职业资格比照认定机制):https://www.gd.gov.cn/zwgk/zcjd/snzcsd/content/post_4973852.html
  • 广东省人民政府新闻办公室·南沙人才政策报道:https://gdio.southcn.com/node_5201f00af5/8623acaa6c.shtml

南沙人才条例:把一张境外证书换算成中国职称

一部有全国意义的地方性法规

2026 年 11 月 1 日,《广州市南沙人才发展促进条例》施行。全文共二十条,是广州市专为南沙区量身打造的第二部地方性法规。

首先要看的是它没有写什么。没有资金、没有申报窗口、没有名额,也没有补贴表。抱着找一个可以写进预算的数字的期待翻开它,多半会失望地合上。二十条里写的是分工:点名要求哪些政府部门去建什么,把门槛、目录和程序推到之后出台的比照认定目录与配套细则里。

也正因如此,这部条例的意义不止于一个区。一位外籍专业人士在中国的材料,很少由一份文件决定,而是由一串文件决定:一张资格证书、一次认可决定、一项职称、一份用人单位登记、一张居留许可。地方性法规的作用,是从这条链上取下一环,给它一个法律依据,从而改变其余各环的谈判条件。还有一层理由与南沙自身的抱负无关。国家层面的政策要落地,靠的是地方立法:中央定框架,省里定方向,而一座城市的法规决定由哪个部门保管目录、由谁接电话、两边口径不一致时听谁的。对需要提前两三年规划用工的企业来说,正是这最后一层决定了承诺能不能变成流程。

时间点也说明了问题。法规通常在做法足够稳定、可以成文的时候才写,而不是在试验刚起步的时候写。广州选择为一个区的人才发展专门立法,说明条文描述的那些安排已经运行了足够久,起草者知道它们会在哪里卡住。

这是一份什么性质的文件

地方性法规不是部门通知。它经过地方立法程序通过,约束的是条文里点名的政府机关,而不只是它谈论的企业。部门规范性文件可以由发文部门自行修改,有时悄无声息,有时就在一个预算周期之内;法规要改,得再走一次立法程序,在那之前,它一直有效。

对做三到五年规划的企业来说,这个差别就是全部价值所在。“本区将维护一份比照认定目录并定期复审”这句承诺,比一份一次性的认可名单更有分量,因为它所附带的义务会一直存在,不会随着签字的官员离任而消失。

边界同样重要。地方性法规不能创设国家法律不允许的权利:它不能降低工作许可或永久居留的法定条件,也不能授予受管制职业的执业资格。它能做的是安排本市如何行使既有的裁量权——哪个窗口受理、哪个部门维护目录、家庭走哪条通道。在这个边界内看,第十三条是进取的;一个实用的检验办法是看官员离任之后会怎样。部门文件往往随作者一起走:继任者重新审一遍,做法慢慢松掉甚至消失,也没有人正式宣布改变。法规留了下来,维护目录、公布服务清单、配备服务专员的义务也留了下来。五年期的用工规划真正需要的,正是这种连续性。

边界同样务实。既然法规改不动国家层面的条件,就应当把它读成一张地方裁量权的地图,而不是新权利的来源。它说本区将提供便利,那个便利就是真的;它没有说的地方,仍然按国家规定办。

第十三条:从一张证书到一项职称

第十三条要求市人民政府、南沙区人民政府在南沙落实三项定位所附带的改革创新、先行先试要求——全国人才管理改革试验区、粤港澳人才合作示范区、国际化人才特区——并借鉴前海深港现代服务业合作区、河套深港科技创新合作区、横琴粤澳深度合作区在港澳人才政策上的做法。

接下来是真正起作用的那句条文。市人力资源和社会保障部门会同市相关行业主管部门、南沙区政府,结合产业需求,在南沙建立境外职业资格与境内职业技能等级、职称的比照认定机制,并制定一份动态管理的比照认定目录。

再往下是后果,只有一句话:持有目录内境外职业资格证书的人才,可以申报对应的国家职业技能等级或者专业技术职称。两个动词承重。“建立”把机制建设变成对政府部门的硬约束;“可以申报”告诉个人这套机制是给谁用的。两者都不承诺结果,而这两者之间的距离,第一段里的三项定位不是装饰。全国人才管理改革试验区带有尝试制度改革的授权,别的区做不了的事这里可以试;粤港澳人才合作示范区带有推动三地规则衔接的任务;国际化人才特区则明确指向服务外籍人才。第十三条把三者叠加起来,作为比照认定机制的法律抓手。

这也是同一条条文里,一半用“探索”、另一半用“建立”的原因。“探索”覆盖的是需要与港澳达成一致的部分;“建立”覆盖的是本市权限之内的部分。一半在等对手方,另一半不用等。

从证书到职称的四个环节

把这一条当成流程而不是原则来读,它描述的是四个环节,每个环节都有各自的负责人。

第一个环节是目录。证书没有进入目录,后面的事都不会发生;目录由人力资源社会保障部门会同行业主管部门维护。对申请人来说,这是一道可以在准备任何材料之前就先问清的问题:要么在册,要么暂时谈不下去。

第二个环节是申报。条文用的是“申报”这个词,意思是主动权在个人手里,实务中则往往由用人单位的人事团队来组织材料。

第三个环节是评审。条文把机制交给人力资源社会保障部门,而不是把决定权写死。评审看什么、境外取得的资格接受哪些证明、审核要多久,属于目录与配套细则的范畴。条例没有规定时限,也不应当从条文中推断出时限。

第四个环节是结果:一项职业技能等级,或者一项专业技术职称。这是两把不同的尺子。职业技能等级围绕用人单位与评价机构组织的认定展开;职称走的是专业技术职称评审体系,事业单位、国有用人单位和不少人才计划看的正是这一级。因此,目录必须写清一张证书对应的是哪一把尺子——对应错了,用人单位的角色值得单独说一句。多数情况下证书在个人手里,但证明材料在单位手里:劳动合同、岗位说明、这个人实际做什么的记录。评审机构读的是一份关于岗位的叙述,而不是一张孤立的证书;岗位描述写得含糊,评出来的等级往往偏低。走过一次流程的人事团队,通常会在第二次申报之前先把岗位说明重写一遍。

证书不在目录上时该怎么办,同样值得知道。答案不是去申诉目录,而是换一条通往等级的路径,或者等下一次修订。同一岗位有多名员工的企业,应当按群体查一次目录,而不是一个人查一次。

比照认定不覆盖什么

三条边界让预期停在合适的位置,而且三条都能在条文里看到。

第一,申报不等于授予。条例打开的是一条申报通道,评审与决定权仍在主管部门手里,依据的是这一条并未改写的规则。证书进入目录,把“够不够资格”的问题变成了一道程序,这是实实在在的收获,但收获是程序性的。

第二,职称不是执业许可。职称说明一个人在评价序列里处在什么位置;从事受管制职业的许可——医师、律师、建筑师——是另一套授权,有自己的考试与注册路径。第十三条把执业资格互认与评价标准互认并列,用的词是“探索”,这是三地之间仍在协商的话题所特有的谨慎措辞。

第三,目录会动。动态管理、与产业需求挂钩,意味着两年前决定过一次的那一版,未必就是下一次适用的那一版。要核对的是申报当日有效的那一版,还有一处区分常常让人误判:认可与考试是两件事。认可问的是在别处取得的资格与这里的某种资格是否等价;考试问的是一个人是否达到了这里设定的标准。第十三条讲的是前者。受管制职业通常还需要后者,目录做得再好也替代不了它。

这也是为什么这一条对某些职业更有用。对工程师、会计师,或者在职称直接影响定级与项目申报资格的行业里做项目管理的人,这套机制是实打实起作用的。对医师或律师,它可能只抹平了一步,而许可那一步原地不动。

借来的地基:前海、河套、横琴

第十三条要求南沙借鉴另外三个平台的做法,这句要求不只是客套。它说明比照认定不是从零开始的南沙试验,而是把一条已经在区域内走通的路径写成地方法规。

实际影响是:问题已经有了答案。在前海、河套或者横琴走过类似流程的专业人士,大致知道什么样的证明材料容易被接受、职称评审机构怎么读一张境外证书、材料通常卡在哪一环。这些经验至少在大框架上可以平移。

南沙增加的是两样东西:一个法律依据,以及一套下沉到区的执行结构。机制由市人社部门、行业主管部门与区政府共同承担,这与单一办公室主导的试点不同;同时它挂靠在三项本身带有改革授权的定位上,对已经在区域内走过类似流程的专业人士来说,实际价值是学习曲线变短。该问的问题与前海、横琴并无不同:目录由哪个机构保管、是否要做翻译与公证、评审方要看课程大纲还是只要证书、用人单位评的等级与评审体系评的等级如何对应。

南沙的不同在于执行层。机制由市人社部门、行业主管部门与区政府共同承担,遇到具体问题的企业因此有不止一个地方可以问;三者之中,行业主管部门往往最有用——它清楚本行业真正需要什么,而“结合产业需求”正是条文为目录给出的理由。

第十四条与第十五条:服务、安居与入学

第十四条偏行政,但很实用。市、区两级要编制并公布人才引进服务事项清单,建设线上人才服务平台与线下人才驿站,并为纳入高层次人才数据库的人才配备服务专员。这些都不改变任何法律状态,但会改变一个新来的人多快能弄清哪些规定适用于自己。

第十五条回答的是家庭最先问的那几个问题。南沙可以制定差异化区域性人才入户实施办法,并自主审核办理。条例要求建设国际人才社区、港澳社区、青年社区;提供租赁住房、人才公寓、购房补贴;加强国际学校、港澳子弟学校、港澳子弟班建设;推动医疗卫生机构实行跨境医疗保险直接结算。

这份清单的顺序不是随手排的。住房与入学,是举家迁移决策通常绕不开的两件事,也是企业最难单方面作出承诺的两件事。当它们出现在地方性法规里而不是投资协议里,就变成了可以被追问、被报告的法定职责,服务专员是条文里最不显眼的一项,对一个刚落地的家庭却往往最有用。一个能说清哪扇窗口办居留许可续期、哪所学校国际部还有名额、以本单位登记情况能申请哪一类住房的人,比一个挂着五十个链接的网站管用。条例把两者配在一起:线上平台负责信息,线下驿站负责在场,专员负责那些填不进表格的情形。

社区条款是同一个逻辑。国际人才社区、港澳社区、青年社区不只是地址,它们是随迁配偶找到职业圈子、入学值得再写一段,因为时间表在这里最容易撞车。国际学校与港澳子弟学校有自己的招生周期、语言分轨与等候名单;三月才到、九月要入学的家庭,往往在居留许可还没办完的时候就已经晚了。条例不承诺学位,它做的是把“加强建设”写进一份区里必须回应的文本——宣传册与工作计划之间的差别,就在这里。

另一个实务点是衔接。在不同体系之间转学的孩子,课程对不上就会损失时间,港澳子弟班存在的意义正是缓冲这一段。企业在谈薪酬方案时,应当问的是相关年级今年有没有名额,而不是原则上有没有这类学校。

人才入户这项权限,值得单独看

第十五条里藏着一条值得单独拎出来看的权限:南沙可以制定差异化区域性人才入户实施办法,并自主审核办理人才入户。

户籍是许多事情的门轴。它常常是入学的先决条件,是限购市场里购房资格的前提,也是一个家庭在决定扎根一座城市之前必须先算清的一项。一个可以自定标准、自己审自己批的区,回应用人缺口的速度,会比全市统一排队快得多。

需要提醒的是,这是一项制定规则的权力,不是名额承诺。实施细则仍要出台,标准仍要公布。对用人单位来说,有用的问题不是“这项权力有没有”,对一个家庭来说,实际顺序通常是这样:拿到录用通知、决定孩子上学、决定住哪里,最后才决定待多久。户籍压在这三件事下面。一个既掌握规则又掌握审核权的区,可以把从录用到入学之间的间隔压短,而这段间隔,往往正是候选人转投另一份录用通知的窗口。

两点提醒要留住。这项权力的内容是制定差异化规则,所以标准会很具体,也会变;审核是自主的,因此区里的解释就是准绳。用人单位应当索取现行办法的书面文本,而不是听同事转述去年听到的说法。

第十六条:签证、居留与永久居留

第十六条是外籍雇员会先翻到的那一条。市公安机关、南沙区政府要完善出境入境和居留政策咨询、语言文化培训等服务,为外籍人才办理签证、工作许可、居留许可以及申请在中国永久居留提供便利。南沙区政府还要设立绿色通道,为本行政区域内单位和人才办理往来港澳商务签注、人才签注提供便捷服务。

诚实的读法是:这一条创设的是服务义务,不是新的资格标准。工作许可、居留许可与永久居留的条件仍在原处,写在国家法律与移民管理部门的实施规定里。这一条做的是把这些条件接到一套本地的服务体系上:有指定的窗口、有绿色通道、有解释与语言支持,有专人在材料递交之前而不是被退回之后讲解规则。

这个衔接比听起来更重要。工作许可制度本来就把外籍雇员分成不同类别,类别决定用人单位要证明什么、许可可以批多久。第十六条没有改动这些类别,它是把它们当作前提,在这个节点上,最容易被误解的是工作许可制度。外籍雇员本来就被分成不同类别,类别决定用人单位要证明什么、许可可以批多久,而这些类别由国家统一设定。第十六条没有新设类别,也没有承诺更快的类别。它承诺的是:对既有类别的本地经办会有解释、有人负责,材料清楚的情形可以走专门通道。

绿色通道是另一项容易被忽略的内容。它针对的是往来港澳商务签注与人才签注——内地员工南下短期出差要用到的证件。把这件事与签证、居留许可写进同一条,其实是在表明本区如何看待人员流动:永久居留是最常被误读的一项。条文说的是为申请提供便利,不是改动条件;条件由国家设定,而且不低。一个区能增加的是材料成形之前的指导、一位跟到底的经办人,以及对居住年限证据的务实说明。把这一条读成“应当更早开始准备”的理由,而不是“结果会不一样”的理由,收获最大。

这也解释了为什么同一份条文对不同企业的价值差别很大:常年有外籍高管轮换的公司,受益于流程的确定性;偶尔办一次的公司,更需要的是一份写得清楚的清单。

公共部门招录,以及它的连带影响

在上述条款之外,条例还明确:南沙的国家机关、事业单位、法定机构、国有企业可以招录或招聘符合条件的港澳人才。

对港澳居民来说,值得接触的雇主范围因此变宽,其中还包括过去根本没有渠道招聘他们的机构。内地公共部门的岗位此前基本不向非内地居民开放,在一个区里移除这道门槛,是结构性的改变,而不只是象征性的姿态。

对企业来说,影响是间接的,但值得跟踪。在那些按拟派人员资历打分的招标文件里,或者在客户要求项目团队必须具备公共部门经历的场合,可选的专业人士池子变大,会改变一家投标方能够拿出什么样的组合。条例没有提到采购,也不需要提——由此产生两个效果。第一,对有经验的港澳专业人士来说,劳动力市场变了:一个现在可以招聘他们的公共部门用人单位,会成为咨询机构与企业职能部门争夺同一批人的竞争者。第二,资历的读法变了。简历上有一段内地公共机构经历的人,在投标材料里的分量与没有的人不同,而条例悄然扩大了能够获得这段经历的人群。

两个效果都不会在 11 月 1 日当天出现。它们取决于招录规则、岗位设置,以及人事系统如何处理港澳居民的身份证件。条例移除的是一道门槛,开口最终有多宽,由操作细节决定。

已经运行的那张对应表

据权威报道,南沙此前已建立涵盖 137 项国际职业资格证书的职称对应关系目录,实现 12 大重点产业 248 项境外职业资格认可,并有超过 400 名港澳专业人士在当地按便利化安排执业。

这些数字描述的是试点。法规增加的是耐久性:一个有法律依据的机制、一份负有维护义务的目录、以及落在具体部门头上的职责,而不是随经费周期结束就停下来的项目。

接下来更有用的问题偏行政:这张目录由哪个部门维护、多久修订一次、上一版与这一版之间改了什么。对这些问题的回答,要让这些数字真正可用,有两个习惯。第一是索取目录本身而不是目录摘要,并保留你依据的那一版。第二是区分“可用于执业的认可”与“可对应职称的认可”;公布的总数通常把两者合在一起算,而它们对一个人的职业路径作用并不相同。

条例的贡献在于,这些问题现在有了收件地址。由点名的部门维护、负有保持更新义务的目录,与试点启动时公布过一次的名单,是两种不同的东西:前者可以被查询、可以跨版本比较、可以写进内部立项说明。

外籍专业人士该准备什么

下面这几条不构成一份申报清单,因为清单本身还没有公布。现在能确定的是自己所处的位置:证书在不在目录上、对应哪一把尺子、单位的岗位说明是否撑得住你想申报的等级。其余问题都由这三个答案推出来,而且每一个都可以在区人社部门的一次沟通里问到,不必经由中介。

  • 先确认你的资格是否在现行比照认定目录内,并记下你核对的是哪一版、哪一天。
  • 把两个问题分开:你能否执业,以及你会被评到哪一级。这两件事走的是两条线。
  • 弄清你的证书对应哪把尺子——职业技能等级还是专业技术职称——因为两者的下游用途不同。
  • 向区人力资源社会保障部门问清楚:你的材料由哪个机构评审,境外取得的资格接受哪些证明。
  • 尽早让雇主参与。职称既被雇主读,也被雇主申报的人才计划读,这类申报很少是一个人的事。
  • 如果你是港澳居民,除了民营市场,也看看条例新打开的那批公共部门用人单位。
  • 把目录的复审周期记进日历。动态目录意味着要反复核对,而不是一次问清、长期有效。
  • 保留你依据的那一版目录;后来的修订不会自动解释此前的决定。
  • 如果单位正在准备职称申报,先核对岗位说明与你要申报的等级是否对得上。

来源

  • 广州市南沙区人民政府·《广州市南沙人才发展促进条例》中英双语全文:https://www.gzns.gov.cn/zwgk/rdzt/hgazczl/cyzc/hgazxzc/rc/content/post_11083998.html
  • 广东省人民政府·11 月起施行法规解读(含南沙境外职业资格比照认定机制):https://www.gd.gov.cn/zwgk/zcjd/snzcsd/content/post_4973852.html
  • 广东省人民政府新闻办公室·南沙人才政策报道:https://gdio.southcn.com/node_5201f00af5/8623acaa6c.shtml

China’s National Trustworthiness Incentive List: What the 2026 Edition Offers Compliant Foreign Firms

  • The National Development and Reform Commission and the People’s Bank of China issued the 2026 edition of the national trustworthiness incentive list on 9 November 2026 and published it the following day.
    国家发展改革委与中国人民银行于 2026 年 11 月 9 日联合印发《全国守信激励措施基础清单(2026 年版)》,次日公布。
  • The list is an incentive catalogue, not an entry threshold: every item names the trusted entities it covers, the department that recognises them and the body that delivers the benefit.
    这是一份激励目录,不是准入门槛:清单逐项写明激励对象、认定部门与实施主体。
  • Twenty-five categories of trusted entities are covered, spanning natural persons, enterprises, individual businesses and social organisations across market regulation, transport, tax, customs, industry and ecology.
    清单覆盖二十五类守信主体,包括自然人、企事业单位、个体工商户和社会组织,涉及市场监管、交通运输、税务、海关、工业、生态环境等领域。
  • More than two hundred measures fall into two families: administrative incentives such as green channels, notification and commitment, and lower inspection frequency, and social incentives such as faster airport screening, easier financial services and platform membership benefits.
    两百余项措施分为两类:一类是绿色通道、告知承诺制、降低抽查检查频次等行政化激励,一类是便捷乘机出行、便利金融服务、平台会员权益等社会化激励。
  • Item 3 covers trusted operators in the information and communications sector, is recognised by the Ministry of Industry and Information Technology, and opens the pilot notification and commitment route for Category II value-added telecom services in free trade zones.
    清单第 3 项针对信息通信行业守信经营主体,由工业和信息化部认定,其中包含自贸试验区第二类增值电信业务告知承诺试点这一通道。
  • The same item offers one-to-one credit cultivation for companies applying for AEO status with customs, with priority certification where other conditions are equal.
    同一项还为申请海关 AEO 资质的企业提供”一对一”信用培育支持,同等条件下优先认证。
  • Recognition in one system can count in another: an A-grade tax payment credit rating earns a credit bonus when a company is assessed for A-grade labour compliance.
    一个体系内的认定可以在另一个体系里加分:纳税缴费信用 A 级经营主体,在认定劳动保障守法诚信 A 级企业时可获信用加分。

China’s National Trustworthiness Incentive List: What the 2026 Edition Offers Compliant Foreign Firms | 全国守信激励措施基础清单:合规外资企业能从中得到什么

What the list is and is not

On 9 November 2026 the National Development and Reform Commission and the People’s Bank of China jointly issued the National Baseline List of Trustworthiness Incentive Measures (2026 Edition), numbered Fa Gai Cai Jin [2026] No. 1593. It was published the following day, on 10 November. The document is a catalogue, and the most useful thing to establish at the outset is what kind of catalogue it is, because the natural mistake is to read it as a list of permissions.

It is not a market access list. Nothing in it opens a sector, removes a licensing requirement or creates an entitlement to operate. Nor is it a whitelist in the sense that inclusion automatically produces a benefit. Each entry in the list identifies three things: the class of trusted entity the measure is aimed at, the department that recognises an entity as belonging to that class, and the body that actually delivers the measure. A company that reads its own business into an entry has learned that a route may exist. It has not travelled it. The route begins with recognition by the named department.

Nor is the list a statement of new policy in most cases. Much of what it contains describes treatment that already existed somewhere in the system — a green channel here, a reduced inspection frequency there — and the list’s contribution is to write it down in one place, at national level, in a comparable form. That is a modest kind of reform, and an easy one to underestimate. A benefit that exists in one province under one name and in another province under a different name is difficult for a company to plan around. A baseline list makes the benefit legible.

The reason a foreign-invested company should read it, despite its domestic framing, is that the underlying credit systems it draws on are ones foreign firms already inhabit. Customs recognises authorised economic operators. Tax authorities issue credit ratings. Market regulators classify entities by their compliance record. A foreign-invested manufacturer with an AEO certificate and an A-grade tax rating is already inside several of the recognition systems from which the list’s measures flow. What the list adds is a map of what those recognitions can be exchanged for.

The word “baseline” in the title does real work. It signals that the national list is a floor rather than a ceiling — the treatment that should be available to a trusted entity wherever it operates, with provinces and municipalities free to go further. For a company with operations in several places, that is the more useful construction. A single provincial programme tells you what one place offers; a baseline tells you what you can reasonably expect not to lose when you move.

It is also worth being precise about the word “incentive”. In ordinary commercial usage an incentive is something of value offered in exchange for behaviour. Here the measures are largely procedural: a faster channel, a lighter touch, a recognised status. They do not put money in a company’s account and they do not reduce a tax bill. What they do is reduce the cost of compliance — the hours, the waiting, the repeated demonstration of things already demonstrated. For a company whose Chinese operations are compliance-heavy, that is where the value sits, and it is easy to miss if you go looking for a financial benefit that is not there.

Who compiles and who delivers

The list was compiled by twenty departments and units working together. That number is worth pausing on, because it explains the document’s shape. A single ministry writing about its own incentives produces a short, deep document about one sector. Twenty bodies writing together produce a broad, shallow document that touches many sectors and goes into detail only where the compilers chose to. The breadth is the achievement; the shallowness is the price.

Delivery is a separate matter from compilation, and the list distinguishes them. Ten enterprises in aviation, finance and the internet sector participate in implementation. That is a departure from the purely administrative model in which a government body both promises and delivers. Some of the measures in the list are commercial in nature — faster screening at an airport, convenience in financial services, membership benefits on an internet platform — and only the firms that run those operations can actually provide them.

The practical consequence is that the list should be read as a description of what participating bodies have agreed to do, not as a guarantee that any particular counter will recognise it. An airline that participates in the arrangement delivers fast-track treatment within its own operations; it does not speak for every airport. A platform that participates delivers membership benefits on its own service; it does not bind its competitors. When a company plans around a social incentive, the question to ask is which participating entity actually touches its customers or employees.

There is a second consequence, less obvious but more important for compliance planning. Because compilation and delivery sit in different hands, an entity may satisfy the recognition test set by a department and still find that the delivering body has its own operational conditions. The recognition establishes eligibility; the delivering body establishes the terms. Both have to be satisfied, and they are not always documented in the same place.

The fact that twenty bodies had to agree is also, in a quiet way, the list’s main achievement. Each department guards its own assessment criteria and its own discretion, and coordination across that many of them means a series of compromises about how a category is described and how wide a measure is drawn. A reader who finds a provision vague should read the vagueness as the residue of that negotiation rather than as an oversight. It also means the list will move at the speed of its slowest member when revised, which is an argument for checking the annual edition rather than assuming continuity.

Twenty-five categories of trusted entities

The list covers twenty-five categories of trusted entities. The categories are not limited to companies: natural persons, enterprises, individual businesses and social organisations all appear. Nor are they confined to one regulatory domain. The coverage runs across market regulation, transport, tax, customs, industry and ecology, among other fields.

The breadth reflects how trustworthiness is actually assessed in China’s administrative system: sector by sector, by the regulator that holds the relevant file. There is no single national trust score that a company either has or lacks. There are records — inspection histories, filing accuracy, payment punctuality, licensing compliance — held by different bodies, each of which draws its own conclusions. The list works across those separate assessments rather than replacing them with one.

For a foreign-invested enterprise, the practical task is to identify which of the twenty-five categories its own operations place it in. That is not a question about ownership. The categories are defined by conduct and by sector, not by the nationality of the investor. A wholly foreign-owned manufacturer is assessed as a manufacturer; a foreign-invested logistics operator is assessed as a transport or logistics entity. What changes with foreign ownership is the surrounding licensing picture, not the trustworthiness test itself.

The more useful framing is to work backwards from the licences a company holds. Each licence implies a regulator, and each regulator maintains a compliance record and a set of classifications. Those classifications are the raw material from which recognition under the list is built. A company that maps its licences to regulators, and its regulators to entries in the list, will find that the exercise produces a short list of realistic opportunities rather than a long list of theoretical ones.

The presence of natural persons among the twenty-five categories is worth a moment, because it is the part of the list a corporate reader is most likely to skim. Some measures in the list are aimed at individuals rather than at entities, and those can matter for a foreign-invested employer in a mundane, practical way: the employees who travel most, and whose time is most expensive, are often the ones whose personal compliance record determines how smoothly they move through an airport. A group that manages mobility as a benefit rather than as a compliance question may be leaving something on the table.

The inclusion of individual businesses is a reminder of the list’s domestic base. A very large share of China’s service economy operates through individual businesses rather than companies, and a national baseline that ignored them would describe only part of the economy. For a foreign group, the relevance is indirect but real: franchise networks, service partners and last-mile operators in its supply chain may sit in that category, and their access to incentives affects the reliability of the network the group depends on.

Industry and ecology appear among the covered fields, which suggests that environmental compliance records feed into at least some of the categories. That is consistent with the wider direction of Chinese regulation, where environmental performance has become a mainstream input to credit assessment rather than a specialist concern. A manufacturer that has invested in its environmental record may find that the investment is legible in more places than it expected.

Two kinds of incentives

The measures in the list — more than two hundred of them — fall into two families. The distinction matters because the two are obtained and used in different ways.

Administrative incentives are delivered by government bodies and take the form of changed treatment in administrative processes. The list names several: honours and commendation, publicity and promotion, a bonus or an upgrade in public credit evaluation, availability of the notification and commitment model, a green channel for administrative matters, priority consideration in evaluations and selections, and reduced frequency of spot checks and inspections. These are reductions in friction. They do not remove an obligation; they change how often and how intrusively the obligation is verified.

Social incentives are delivered by the participating enterprises and take the form of commercial convenience. The list names convenient air travel, easier access to financial services, and membership benefits on internet platforms. These are consumer-facing and employee-facing benefits rather than regulatory ones, and their value depends entirely on which participating entity is involved and what it has agreed to offer.

Two cautions follow from the structure. The first is that reduced inspection frequency is not exemption from inspection. A company that reads a lower frequency as a licence to let standards slip has misread the measure; the underlying obligation is unchanged, and the consequence of a finding is unchanged. The second is that public credit evaluation bonuses are inputs to other assessments, not outputs in themselves. Their value is realised when the company applies for something else that uses the credit score as a criterion.

For a foreign-invested company, the administrative family is where the operational value sits. Green channels and notification and commitment shorten timelines on applications the company has to make anyway. Lower inspection frequency reduces disruption to production. Neither changes what the company is allowed to do, and both change how much of its week is spent demonstrating that it is allowed to do it.

To capture that value, the measures have to be attached to a calendar. A green channel is only useful to a company that knows which of its recurring applications can use it, and when those applications fall due. The exercise is to list the filings, registrations, renewals and inspections that recur over a year, and then to check each against the current edition of the list. What emerges is usually a small number of touchpoints where a recognised status converts into days saved, and a larger number where it does not apply at all.

There is a communications risk in the other direction. A company that announces that it is “on the national trustworthiness list” has said something the document does not support. The accurate statement is narrower and more useful: recognised by a named department as belonging to a named category of trusted entity, with access to named measures. Precision here is not pedantry. Buyers, banks and counterparties who check the claim will find the narrower version, and a company whose public statements match the document is easier to do business with than one whose statements have to be walked back.

The two families also differ in who inside the company owns them. Administrative incentives belong to the regulatory affairs or government relations function, which knows the filing calendar and the departmental contacts. Social incentives belong to whoever manages travel, treasury and customer relationships. A group that assigns the whole list to one team will usually find that half of it goes unused, not because the measures are unavailable but because the wrong people were looking at them.

The telecom and AEO provisions foreign firms should read

Item 3 of the list concerns trusted operators in the information and communications sector. The recognising department is the Ministry of Industry and Information Technology. For a foreign-invested company in telecoms, cloud or adjacent digital services, this is the entry that deserves the closest reading, because it touches two things such companies tend to care about: how quickly a licence application moves, and how the company is treated once it holds one.

The measure contains four elements. First, it applies to the pilot notification and commitment route for Category II value-added telecom services in free trade zones. Notification and commitment is a model in which the applicant files a statement and undertakes to comply, and the authority accepts the filing rather than conducting a full prior examination. Applied to Category II value-added services in the free trade zones, it shortens the front end of the process considerably — the applicant’s own undertaking does part of the work that a prior review would otherwise do.

Second, the measure provides for lower frequency of routine supervision and spot checks, or exemption from on-site inspection as circumstances warrant. The phrase “as circumstances warrant” carries the discretion; it is not an automatic entitlement, and a company should read it as a direction of travel rather than a fixed rule. Third, under equal conditions, priority is given in the allocation of communications resources, network access, telecom equipment network entry and tendering.

Fourth, and most cross-cutting, the measure provides one-to-one credit cultivation support for companies applying for AEO status with customs, with priority certification under equal conditions. This is the element that reaches beyond the telecoms sector in its logic: a recognition granted by the industry regulator becomes an input into a customs assessment. For a foreign-invested group that already holds AEO status in another jurisdiction and is building a Chinese application, the availability of structured cultivation support is a practical accelerant.

The qualifier “under equal conditions” appears twice, and it should not be read past. It means the measure breaks ties; it does not create a preference that overrides substantive criteria. A company that does not meet the underlying standard for AEO certification will not be certified because it is trusted in the telecoms sense. What it may get is a faster, better-supported route to demonstrating that it does meet the standard.

It is not an accident that telecoms is the sector where the list goes into this level of detail. Value-added telecom services have been among the areas where foreign investment has been permitted within the free trade zones, subject to licensing and to the conditions attached to the pilot. A sector with a live foreign-invested population, a licensing bottleneck at the front end and a supervisory relationship that continues after licensing is exactly where a trustworthiness incentive has something to bite on. The measure is written for companies that are already in the queue.

The order of operations matters. Recognition comes first, from the industry regulator; the measures follow. A company cannot invoke the notification and commitment route on the strength of having read the list. It becomes available once the company is recognised as a trusted operator in the sector, and that recognition depends on the company’s own compliance record with the regulator. The practical implication is that the work of qualifying is done long before an application is filed, in the ordinary course of keeping a clean record.

The AEO element deserves a note on sequencing too. Customs certification is a demanding process with its own criteria, its own documentation and its own assessment. The value of “one-to-one” cultivation support is that it shortens the period between deciding to apply and being ready to be assessed — it does not shorten the assessment itself, and it does not lower the bar. A group planning a Chinese AEO application should treat the support as a project accelerant rather than as a substitute for preparation.

Cross-departmental recognition in practice

The list’s most interesting structural feature is that recognition in one system can be used in another. The example the document gives is specific: an entity with an A-grade tax payment credit rating receives a credit bonus when it is assessed for recognition as an A-grade enterprise for labour compliance.

Read narrowly, that is one line item. Read structurally, it is a statement about how the credit architecture is being knitted together. Historically, a company’s tax record was relevant to the tax authority, its labour record to the human resources authority, its customs record to customs. Each body assessed on its own information. Cross-recognition means that a strong record in one place becomes visible, and valuable, in another.

For a group with multiple entities in China, this changes the calculus of compliance investment. Improving the tax credit rating of one subsidiary may produce benefits in assessments conducted by a different regulator entirely. The return on a compliance improvement is no longer confined to the relationship with the regulator that demanded it. That is an argument for treating credit management as a group-level function rather than a matter for each entity’s finance team.

The example also shows the limits of the mechanism. The tax rating produces a bonus in the labour assessment; it does not produce the labour recognition. The substantive conditions for A-grade labour compliance still have to be met, and the bonus operates at the margin — it can tip a close case, not carry a weak one. Companies that treat cross-recognition as a substitute for meeting the underlying standard will be disappointed, and may waste effort pursuing recognitions they cannot convert.

There is a further practical point about evidence. A cross-recognition works only if the second body can see the first body’s conclusion. That requires the company to be able to produce current evidence of its standing in the originating system, in a form the receiving body accepts. A credit rating that exists in a database the company cannot query is of limited use in an application elsewhere. Groups should be able to document their own standing, on demand, for each recognition they intend to rely on.

Group structure complicates this in a way that is worth planning for. Recognition attaches to a legal entity, not to a group. In a structure with a holding company, a manufacturing subsidiary, a trading company and a services company, the entity that holds the customs certification may not be the entity that holds the tax rating, and neither may be the entity that applies for the licence the group actually wants. Cross-recognition only helps where the recognitions and the application sit in the same entity, or where the receiving body accepts group-level evidence. Mapping which entity holds what is therefore the first step, and it is often the step that reveals the gap.

Annual review and what replaced the old memoranda

The list is updated annually. The compilers have committed to adjusting both the categories of trusted entities and the measures attached to them as laws and policies change. That is a meaningful design choice: a static list would drift out of alignment with the underlying regulations within a few years, and a list that is revised without a fixed rhythm is difficult for companies to track.

The 2026 edition also repealed five joint memoranda on trustworthiness incentives concluded between 2016 and 2018. Those memoranda were bilateral arrangements between pairs of departments, each covering a specific combination of recognition and reward. Replacing them with a single baseline list consolidates the landscape. Instead of tracking a handful of separate agreements, a company tracks one document, revised on a known cycle.

The repeal has a practical implication that is easy to miss. Guidance written around the old memoranda — internal compliance manuals, consultant notes, articles of the kind this one is — may still describe arrangements that no longer exist in that form. A company relying on a memorandum it remembers from several years ago should check whether the measure it describes now appears in the current list, and in what terms. The substance may have survived the consolidation; the instrument has not.

For planning purposes, the annual cycle suggests a simple discipline. Once a year, review which of the twenty-five categories the company plausibly falls into, which recognitions it currently holds, and which measures in the current edition it could actually use. The exercise is short if the company’s licence portfolio is stable, and revealing if it is not. It also produces the evidence file that cross-recognition depends on, as a by-product rather than as a separate project.

Two framing points are worth keeping in view throughout. First, the list rewards conduct, and conduct is something a company controls. Unlike a sectoral opening, which arrives on a schedule set by negotiators, a recognition under the list is earned through the company’s own record. Second, the list is a baseline, not a ceiling. Provincial and municipal programmes may offer more, and often do; the national list establishes the floor of comparable treatment rather than the full extent of what is available.

Where the list sits beside market access rules

The single most important thing to avoid is confusing this document with the rules that govern whether a foreign investor may enter a sector at all. Those rules live in separate instruments and are administered by separate processes. They determine the permitted scope of foreign ownership, the licensing requirements that attach to a given activity, and the conditions under which an application will be considered. The trustworthiness list does none of that.

The relationship between the two is sequential rather than substitutive. Access rules decide whether a company may hold a licence. The trustworthiness list decides how that company is treated once it holds one, and how quickly a further application is processed. A company that cannot obtain the licence will not obtain it by being trustworthy; a company that can obtain it may obtain it faster, and may be supervised more lightly afterwards.

Keeping the two apart matters for planning, because the two carry different kinds of uncertainty. Access rules change through published amendments, and their direction is visible in advance. Trustworthiness recognition depends on the company’s own record and on a departmental assessment, which means it is within the company’s influence but not within its control. A group that builds a China plan on the assumption that recognition will arrive on schedule is taking a risk it does not need to take; a group that builds the plan on the assumption that recognition may never arrive, and treats it as upside, is not.

There is one place where the two do intersect, and it is the reason the telecom item matters. Where a sector’s access route includes a notification and commitment model, the trustworthiness recognition determines whether a company can use that route. In that case the list is not merely about post-licensing treatment; it is about the shape of the application itself. That is the exception, and it is worth identifying clearly rather than generalising from.

The distinction also matters in transactions. A counterparty conducting diligence on a Chinese target will sometimes present a trustworthiness recognition as evidence that the business is properly licensed. It is not. A recognition speaks to conduct within a regulatory relationship; it says nothing about whether the underlying licence exists, covers the activity in question, or remains valid. Buyers should verify the licence independently and treat the recognition as a separate, and lesser, data point.

What a foreign group should review this year

The list is annual, so the review should be too. A short agenda, run once a year, is enough to keep a group from missing what the current edition offers.

Start with the licence portfolio. List every licence, registration and filing the Chinese entities hold, and the regulator behind each. This is the raw material for everything else, and it changes more often than groups expect — a new business line, a new site, an acquisition, and the portfolio has moved.

Then map the portfolio against the twenty-five categories. The aim is not to claim every category that might apply, but to identify the two or three where the company’s record is genuinely strong and where a recognition would be worth having. Chasing recognitions that lead nowhere is a common and expensive mistake.

Then check the current edition for measures attached to those categories, and test each against the company’s actual calendar. A measure that never coincides with a filing the company makes is of no practical value, however attractive it reads.

Then establish the evidence position. For each recognition the company holds, confirm that current proof can be produced on demand, in a form a receiving body will accept. For each recognition it wants, identify what the recognising department would need to see, and whether the company’s record already shows it.

Finally, note the date. The list is revised annually, and the edition that governs this year’s planning is not necessarily the edition that will govern next year’s. A review that runs on a fixed date each year, shortly after the new edition appears, is the simplest way to stay current without tracking every departmental announcement.

Sources

  • 国家发展改革委,《国家发展改革委、中国人民银行联合印发〈全国守信激励措施基础清单(2026 年版)〉》(发改财金〔2026〕1593 号),2026-11-09。https://www.ndrc.gov.cn/xwdt/tzgg/202611/t20261109_1408743.html
  • 国家发展改革委,《关于〈全国守信激励措施基础清单(2026 年版)〉的说明》,2026-11-10。https://www.ndrc.gov.cn/xxgk/zcfb/tz/202611/P020261110521301048830.pdf
  • 国家发展改革委,《全国守信激励措施基础清单(2026 年版)》正文,2026-11-10。https://www.ndrc.gov.cn/xwdt/tzgg/202611/P020261110521429669759.pdf
  • 央视网,《国家发展改革委负责同志就〈全国守信激励措施基础清单(2026 年版)〉答记者问》,2026-11-11。https://gongyi.cctv.com/2026/11/11/ARTIg5m10IKVdLbkUjy3BR7n261111.shtml

全国守信激励措施基础清单:合规外资企业能从中得到什么

这份清单是什么,不是什么

2026 年 11 月 9 日,国家发展改革委与中国人民银行联合印发《全国守信激励措施基础清单(2026 年版)》,文号为发改财金〔2026〕1593 号,次日即 11 月 10 日公布。这是一份目录,而读它之前最该弄清楚的,是它属于哪一类目录——因为最容易犯的错,是把它当成一份许可清单。

它不是市场准入清单。里面没有任何一条打开某个行业、取消某项许可要求,或者创设一项经营权利。它也不是那种”进了名单就自动享受待遇”的白名单。清单的每一条都写明三件事:措施针对的是哪一类守信主体、由哪个部门认定主体属于该类、由哪个主体实际实施这项措施。一家公司把自己的业务读进某一条,只是知道可能存在一条路,还没有走上去。路从认定部门作出认定那一刻才开始。

多数情况下,这份清单也不是新政策的宣告。它记载的许多内容,是系统里本来就存在的待遇——这里一条绿色通道,那里一次减少的检查——清单的贡献在于把它们写进同一份文件,放在国家层面,用可比较的形态呈现。这是一种温和的改革,也很容易被低估。同一项便利,在一个省叫一个名字、在另一个省叫另一个名字,企业就无法据此做计划;一份基础清单让这项便利变得可读。

之所以外资企业也该读它,是因为清单所依托的那些信用体系,外资企业本来就在其中。海关认定高级认证企业,税务部门给出纳税信用级别,市场监管部门按合规记录给主体分类。一家持 AEO 证书、纳税信用 A 级的外资制造企业,早已身处清单多项措施所流出的那几个认定体系之内。清单补上的,是一张地图:这些认定可以换到什么。

标题里的”基础”二字是有分量的。它表明国家清单是地板而不是天花板——守信主体无论在哪里经营都应当享有的待遇,省市两级可以在此基础上加码。对在多地布局的企业,这个构造更有用:一份省级方案告诉你一个地方给什么,一份基础清单告诉你,换一个地方时哪些东西不至于丢掉。

“激励”一词也值得说准。在通常的商业语境里,激励是指为换取某种行为而给出的有价值之物。而这里的措施大多是程序性的:更快的通道、更轻的监管、一个被承认的身份。它们不会往企业账上打钱,也不减免税款。它们做的是降低合规成本——工时、等待、以及把已经证明过的事情再证明一遍。对合规负担较重的在华业务而言,价值就在这里;如果一心想找一项财务上的好处,反而会错过它。

谁来编制,谁来实施

清单由二十个部门和单位共同编制。这个数字值得停一下,因为它解释了文件的形状。一个部委写自己的激励措施,会写出一份范围窄、挖得深的文件;二十个主体一起写,写出的是一份覆盖面宽、每处只到必要深度的文件。宽是成就,浅是代价。

实施与编制是两件事,清单把二者分开。航空、金融、互联网领域有十家企业参与实施。这与此前那种”政府既承诺又兑现”的纯行政模式不同。清单里有些措施本身是商业性的——机场快速安检、金融服务的便利、互联网平台的会员权益——只有经营这些业务的企业才真正提供得了。

由此产生一个实务上的后果:清单应当被读成”参与各方同意做什么”的说明,而不是”任何一个柜台都会认”的保证。参与安排的航空公司,在自己的运营范围内提供快捷通道,它代表不了所有机场;参与的平台,在自己服务内提供会员权益,它约束不了竞争对手。企业围绕一项社会化激励做安排时,该问的问题是:究竟是哪一个参与主体,真正接触到自己的客户或员工。

还有第二个后果,不那么显眼,但对合规规划更重要。编制与实施分处两手,意味着一家企业可能满足了部门设定的认定标准,却发现实施主体另有自己的操作条件。认定解决的是”有没有资格”,实施主体解决的是”按什么条款”。两者都要满足,而它们并不总写在同一个地方。

二十个主体必须达成一致,这件事本身也是清单的主要成就,只是不事声张。每个部门都守着自己的评价标准和裁量空间,这么多部门之间的协调,意味着一连串关于类别怎么描述、措施画多宽的妥协。读到某条写得笼统的读者,应当把这种笼统读成谈判的余留,而不是起草的疏漏。它也意味着清单修订时的推进速度取决于最慢的那个成员——这正是每年核对新版本、而不是假定内容延续的理由。

二十五类守信主体

清单覆盖二十五类守信主体。这些类别不限于企业:自然人、企事业单位、个体工商户、社会组织都在其中。它们也不局限于一个监管领域,而是横跨市场监管、交通运输、税务、海关、工业、生态环境等方面。

覆盖面之宽,反映的是中国行政体系认定守信的实际方式:一个领域一个领域地认定,由掌握相应档案的监管部门作出。并不存在一个统一的”全国信用分”,企业要么有、要么没有。存在的是分散在不同机构手里的记录——检查历史、备案准确度、缴款及时性、许可合规情况——各机构各自得出自己的结论。清单是在这些分散评价之间做工作,而不是用一套评价取代它们。

对外资企业来说,实务任务是判断自己的经营把自己放进了二十五类中的哪几类。这不是所有权问题。类别按行为与行业界定,不按投资者的国籍界定。外商独资的制造企业按制造企业评,外资参股的物流企业按交通运输或物流主体评。外资身份改变的是周边的许可图景,不是守信评价本身。

更有用的思路,是从企业持有的许可证件倒推。每一张许可证件对应一个监管部门,每个监管部门维护一份合规记录和一套分类。这些分类,就是清单认定所依据的原料。把许可证件对到部门、把部门对到清单条目,做完之后,企业得到的会是一份不长的、现实可行的机会清单,而不是一份很长的、理论上的可能性清单。

二十五类中有自然人,这一项值得停一下,因为企业读者最容易一眼扫过去。清单里有些措施是针对个人而不是针对单位的,对外资雇主来说,它的意义很朴素:出差最频繁、时间最贵的那些员工,往往正是其个人信用记录决定过关是否顺畅的那批人。把人员流动当福利来管、而不当合规事项来管的集团,可能正把一部分好处留在桌上。

个体工商户被纳入,是清单本土底色的一个提示。中国的服务业有相当大的比例通过个体工商户而非公司运营,一份忽略它们的国家级基础清单,只能描述经济的一部分。对外资集团而言,相关性是间接的,但确实存在:加盟网络、服务商、供应链末端的配送主体可能就落在这一类,它们能否获得激励,会影响集团所依赖的那张网络的可靠程度。

工业和生态环境被列入覆盖领域,说明环境合规记录至少进入了某些类别的认定。这与中国监管的整体走向一致:环境表现已经从专门事项变成了信用评价的常规输入。在环境记录上真正投入过的制造企业,可能会发现这项投入在比预想更多的地方被看见。

两类激励

清单里的措施超过两百项,分为两族。这个区分重要,因为两者的取得方式和使用方式不同。

行政化激励由政府机关实施,表现为行政流程中待遇的改变。清单点到的有:荣誉表彰、宣传推广、公共信用评价加分或提升等级、适用告知承诺制、行政事项办理绿色通道、评先评优优先考虑、降低抽查检查频次。这些是摩擦的减少,不是义务的免除;它们改变的是义务被核验的频率与强度。

社会化激励由参与实施的企业提供,表现为商业上的便利。清单点到的是便捷乘机出行、便利金融服务、互联网平台会员权益。这些面向消费者与员工,而非面向监管,其价值完全取决于涉及哪个参与主体、它承诺了什么。

由这个结构可以引出两点提醒。第一,降低抽查检查频次不等于免于检查。把”频次降低”读成”标准可以松”的企业读错了;底层义务没变,被查出问题后的后果也没变。第二,公共信用评价加分是其他评价的输入,不是结果本身。它的价值,要等到企业去申请另一件以信用分为条件的事情时才兑现。

对外资企业而言,操作价值主要落在行政化那一族。绿色通道与告知承诺制,缩短的是企业本来就要办的申请的时间;检查频次下降,减少的是对生产的打扰。两者都不改变企业可以做什么,改变的是它每周要花多少时间去证明自己可以做。

要把这层价值拿到手,措施必须挂到日历上。绿色通道只对”知道自己哪些常规申请用得上、什么时候到期”的企业有用。做法是把一年里重复发生的备案、登记、续期、检查列出来,再逐项对照现行清单。结果通常是这样:少数几个节点上,一个被承认的身份能换成几天时间;更多的节点上,它完全不适用。

反方向还有一个传播上的风险。宣称自己”进入了全国守信激励清单”的企业,说出了一句文件并不支持的话。准确的说法更窄,也更有用:经某个部门认定为某一类守信主体,可适用某几项措施。这里的精确不是咬文嚼字。会去核对的买方、银行、交易对手,找到的是窄的那个版本;对外表述与文件对得上的企业,比需要不断往回圆的企业更好打交道。

两族措施在企业内部归属的部门也不同。行政化激励属于政府事务或合规事务团队,他们知道申报日历,也知道对口的人是谁。社会化激励属于管差旅、管资金、管客户关系的人。把整份清单交给一个团队去管的集团,通常会有一半用不上——不是因为措施不存在,而是因为看的人不对。

外资企业该细读的电信与 AEO 条款

清单第 3 项针对信息通信行业的守信经营主体,认定部门是工业和信息化部。对身处电信、云服务或相邻数字业务的外资企业来说,这一项最值得细读,因为它触及这类企业通常最在意的两件事:许可申请走得有多快,拿到许可之后被如何对待。

这项措施包含四个要素。其一,适用于自贸试验区第二类增值电信业务的告知承诺试点。告知承诺是一种申请人作出声明并承诺合规、机关接受备案而非进行全面事前审查的模式。用在自贸试验区的第二类增值业务上,它把流程的前端大幅压缩——申请人自己的承诺,承担了一部分本应由事前审查完成的工作。

其二,降低日常监管抽查检查频次,或者视情况可以免予实地检查。”视情况”三个字承载着裁量空间,它不是自动取得的权利;企业宜把它读成一个方向,而不是一条定规。其三,同等条件下,在通信资源、网络接入、电信设备进网以及招投标中优先。

其四,也是最具有穿透力的一条:为申请海关 AEO 资质的企业提供”一对一”信用培育支持,同等条件下优先认证。这一条的思路超出了电信行业本身——行业监管部门作出的认定,成了海关评估的一项输入。对一个已在他国取得 AEO 资质、正在搭建中国申请的外资集团来说,有结构化的培育支持可用,是实打实的加速器。

“同等条件”这个限定出现了两次,不该被读过去。它的意思是:这项措施在条件相当时打破平局,而不是创设一项可以压过实体标准的偏好。达不到 AEO 认证底层标准的企业,不会因为它在电信意义上守信就被认证。它能得到的,是一条更快、更有支撑的路径,去证明自己确实达到标准。

清单恰恰在电信这一行写到这个细度,并非偶然。增值电信是自贸试验区内允许外资进入的领域之一,进入须经许可,并受试点附加条件的约束。一个既有现成外资群体、前端又有许可瓶颈、发证之后还持续监管的行业,正是守信激励能咬得住的地方。这项措施是写给已经在队伍里的企业看的。

顺序很重要。先有认定,来自行业监管部门;措施随之而来。企业不能凭读过清单就去用告知承诺这条通道。它要在被认定为该行业守信经营主体之后才打开,而这项认定取决于企业在监管部门的合规记录。实务含义是:取得资格的功夫,早在提交申请之前就下完了,就在日常把记录做干净这件事里。

AEO 那一部分也有先后次序值得说明。海关认证是一套有自身标准、自身材料、自身评审的过程。”一对一”培育支持的价值,在于缩短从决定申请到具备受评条件之间的那段时间——它不缩短评审本身,也不降低门槛。计划在中国申请 AEO 的集团,宜把这项支持当作项目的加速器,而不是准备的替代品。

跨部门互认在实务中的样子

清单最有意思的结构特征,是一个体系里的认定可以在另一个体系里使用。文件给出的例子很具体:纳税缴费信用 A 级的经营主体,在认定劳动保障守法诚信 A 级企业时可以获得信用加分。

窄读,这只是一个条目;结构性地读,它是一句关于信用架构正在被编织起来的陈述。过去,企业的纳税记录对税务部门有意义,劳动记录对人力资源社会保障部门有意义,海关记录对海关有意义,各管各的信息。跨部门互认意味着,一个地方的好记录,会在另一个地方被看见,并且值钱。

对在华拥有多个实体的集团来说,这改变了合规投入的算法。把某一家子公司的纳税信用级别提上去,可能在完全另一个监管部门主持的评价里产生好处。一项合规改善的回报,不再局限于提出要求的那家监管机构。这本身就是一条论据:信用管理应当作为集团层面的职能,而不是各实体财务团队各自的事。

这个例子也显示了机制的边界。纳税信用带来的是劳动保障评价中的加分,不是劳动保障认定本身。A 级劳动保障守法诚信的实体条件仍须满足,加分只在边缘起作用——它能让一个接近的案子过关,撑不起一个薄弱的案子。把互认当成满足底层标准的替代品的企业,会失望,也可能把力气花在争取一些自己转换不了的认定上。

还有一点关于证据。互认要成立,第二个部门必须看得见第一个部门的结论。这要求企业能够拿出原体系内当前状态的证明,并且是接收方接受的形式。一项只存在于企业查不到的数据库里的信用级别,在别处申请时用处有限。集团应当对打算依赖的每一项认定,都能随时出具有据可查的自身状态说明。

集团架构会让这件事复杂一层,值得提前安排。认定跟着法人实体走,不跟着集团走。在一个有控股公司、制造子公司、贸易公司、服务公司的架构里,持海关认证的实体未必是持纳税信用的那个,而两者又未必是去申请集团真正想要的那张许可证的实体。互认只在两种情况下有用:认定与申请落在同一个实体上,或者接收方接受集团层面的证据。因此第一步是把”哪个实体持有什么”画出来,而这一步往往正是暴露出缺口的那一步。

年度更新,以及被取代的那批备忘录

清单按年度更新。编制方已承诺,随法律法规和政策文件调整,及时调整守信主体的类别和所附的措施。这是一个有意义的设计选择:静态清单过不了几年就会与底层法规脱节,而没有固定节奏的修订又让企业难以跟踪。

2026 年版同时废止了 2016 年至 2018 年间订立的五份守信联合激励合作备忘录。那些备忘录是部门两两之间的安排,每一份覆盖一组特定的”认定加奖励”组合。用一份基础清单取代它们,等于把这张网收拢:企业不必再追几份彼此独立的协议,只需跟踪一份文件,按已知周期修订。

废止有一个容易漏掉的实务含义。围绕旧备忘录写成的指引——内部合规手册、顾问笔记、以及像本文这样的文章——可能仍在描述一种已不以那种形式存在的安排。企业若依赖若干年前记得的一份备忘录,应当核对该措施如今是否出现在现行清单里、以什么措辞出现。实质内容也许在归并中保留了下来,载体没有。

从规划角度看,年度周期提示了一条简单的纪律。每年一次,回顾企业大致落在二十五类中的哪几类、目前持有哪些认定、现行版本里哪些措施是自己真用得上的。许可证件组合稳定的企业,这件事很短;不稳定的企业,会发现它很有启发。它还会顺带产出互认所依赖的那份证据档案,而不是另立一个项目去做。

有两点框架性的判断值得始终放在视野里。第一,清单奖励的是行为,而行为是企业自己可控的。行业开放按谈判方的日程到来,清单下的认定则靠企业自己的记录挣得。第二,清单是基础,不是上限。省市两级的项目往往提供更多;国家清单确立的是可比待遇的地板,而不是可及范围的屋顶。

清单与准入规则各管一段

最要紧的一件事,是不要把这份文件与”外资能不能进入某个行业”的规则混为一谈。那些规则写在另外的文件里,由另外的程序执行。它们决定外资持股的允许范围、某项活动所附的许可要求、以及一份申请在什么条件下会被受理。守信清单一样都不管。

两者的关系是先后关系,不是替代关系。准入规则决定企业能不能拿到许可证件;守信清单决定拿到之后被如何对待,以及下一次申请走得多快。拿不到许可证件的企业,不会因为守信而拿到;能拿到的企业,可能拿得更快,事后也可能被监管得更轻。

把两者分开,对规划有实际意义,因为两者带来的不确定性性质不同。准入规则通过公开修订变化,方向提前可见。守信认定取决于企业自身记录和部门评价,意味着它受企业影响,但不由企业掌握。把中国计划建立在”认定会按时到来”这一假设上的集团,承担了本不必承担的风险;把计划建立在”认定可能一直不来”、并将其视为上行空间的假设上,则不必。

两者有一个交汇点,这也正是电信那一项重要的原因。当某个行业的准入路径包含告知承诺模式时,守信认定就决定了企业能否走这条路。此时清单不只是发证之后的待遇问题,而是申请本身的形态问题。这是例外,值得清楚地点出来,而不是据此推广到所有行业。

这个区分在交易里同样要紧。对一家中国标的做尽职调查的交易对手,有时会把某项守信认定当作”这项业务已经持证”的证据。它不是。认定说的是某段监管关系中的行为表现,与许可证件是否存在、是否覆盖所涉活动、是否仍然有效,一概无关。买方应当独立核验证件,把认定当作另一条、且分量更轻的数据。

今年该复盘什么

清单按年出,复盘也应当按年做。一份简短的议程,每年跑一次,就足以让集团不漏掉现行版本给的东西。

先从许可证件组合开始。把中国各实体持有的许可、备案、登记逐项列出,并标明背后的监管部门。这是其余一切工作的原料,而它变化得比集团预想的更频繁——新增一条业务线、新增一个场地、完成一次收购,组合就变了。

再把组合对到二十五类上去。目的不是把所有可能沾边的类别都申报一遍,而是找出两三类企业记录确实扎实、且认定值得去争取的。追逐一些通向虚无的认定,是常见且昂贵的错误。

然后查现行版本中这些类别所附的措施,逐项对照企业真实的日历。一项永远不会与企业的申报节奏重合的措施,读起来再漂亮,也没有实务价值。

接着确立证据状态。对已经持有的每项认定,确认能否按要求随时出具当前证明;对想要取得的认定,弄清认定部门需要看到什么,以及企业现有记录是否已经呈现了这一点。

最后记下日期。清单按年修订,管今年规划的版本未必管明年的。每年固定一天、在新版本面世后不久做一次复盘,是不必跟踪每一次部门公告、又能保持同步的最简办法。

来源

  • 国家发展改革委,《国家发展改革委、中国人民银行联合印发〈全国守信激励措施基础清单(2026 年版)〉》(发改财金〔2026〕1593 号),2026-11-09。https://www.ndrc.gov.cn/xwdt/tzgg/202611/t20261109_1408743.html
  • 国家发展改革委,《关于〈全国守信激励措施基础清单(2026 年版)〉的说明》,2026-11-10。https://www.ndrc.gov.cn/xxgk/zcfb/tz/202611/P020261110521301048830.pdf
  • 国家发展改革委,《全国守信激励措施基础清单(2026 年版)》正文,2026-11-10。https://www.ndrc.gov.cn/xwdt/tzgg/202611/P020261110521429669759.pdf
  • 央视网,《国家发展改革委负责同志就〈全国守信激励措施基础清单(2026 年版)〉答记者问》,2026-11-11。https://gongyi.cctv.com/2026/11/11/ARTIg5m10IKVdLbkUjy3BR7n261111.shtml

Beijing’s Second Five-Year Business Environment Plan: Twenty-Nine Reform Areas and Five Clauses Foreign Investors Will Feel

  • Beijing has published its second consecutive five-year business environment plan, built around six environments and 29 reform areas and framed by the development and reform commission as a “partner-style” approach to service.
    北京发布连续第二部营商环境五年专项规划,围绕六大环境提出 29 个方面改革举措,市发展改革委将其概括为”伙伴式”服务理念。
  • Registered capital subscription is to be reinforced by a “point-to-point” reminder mechanism that prompts eligible companies to adjust their contribution schedules and pay in as the law requires.
    注册资本认缴登记制度改革继续深化,将建立完善”点对点”提醒服务机制,提示、督促符合条件的公司及时调整出资期限、依法履行出资义务。
  • Standardised domicile registration requires both the address information and the property owner’s consent to be genuine, and lets users of standardised addresses avoid submitting paper supporting documents.
    经营场所(住所)标准化登记要求”地址信息”和”产权人意愿”双真实,使用标准化地址的经营主体可免于提交纸质证明材料。
  • Bilingual application forms are to be promoted further, extending the bilingual registration documents Beijing already provides for foreign-invested entities.
    中英文对照版申请文书将进一步推广,这是在既有外资登记双语文书基础上的延伸。
  • Jing-Jin-Ji licence recognition is to reach the point where a business licence can be handled, issued and collected anywhere across the three places, on top of 254 “same matter, same standard” items already in place.
    京津冀将加快实现全域营业执照异地”办、发、领”,此前已累计推出 254 项”同事同标”事项。
  • Beijing will pursue local fair competition legislation and sample 500 policy measures a year, alongside a campaign against the abuse of administrative power to exclude or restrict competition.
    北京将推动公平竞争地方立法,每年抽查政策措施 500 件,并开展整治滥用行政权力排除、限制竞争专项行动。
  • Supervision is shifting from penalty after the fact to a check-up beforehand, supported by full coverage of scanned inspections, wider use of artificial intelligence in off-site supervision and an appointment-based service model.
    监管正从”事后处罚”转向”事前体检”,配套措施包括”扫码检查”全覆盖、人工智能大模型在非现场监管中的应用,以及”预约式”服务。
  • The 137 “one thing at a time” scenarios already cut processing time by 57 per cent, materials by 50 per cent and trips by 87 per cent.
    已建成的 137 个”高效办成一件事”场景,整体压缩时限 57%、精简材料 50%、减少跑动 87%。

Beijing’s Second Five-Year Business Environment Plan: Twenty-Nine Reform Areas and Five Clauses Foreign Investors Will Feel | 北京第二部营商环境五年规划:29 个方面改革,与外资最直接相关的五条

Why a second five-year plan matters

Beijing has released the business environment plan for its fifteenth five-year period. It is the city’s second consecutive five-year special plan on the subject, and that continuity matters: the earlier plan built the machinery — one-stop registration, online filing, a single window — and this one assumes the machinery exists and starts adjusting what runs through it.

It is a planning document, not a menu. Nothing in it opens an application channel on its own; the provisions become operable when supporting rules and departmental implementation schedules follow. Read it as a statement of direction; the detail comes with the operating rules.

What a five-year plan does that an annual action plan cannot is set the terms on which the annual reports will be written. A measure that appears here as a direction — say, local legislation on fair competition — will be reported against for the next five years, which is a different kind of commitment from a pilot announced for one season.

The base it starts from is measurable. Registered market entities in the city rose from 2.115 million at the end of 2020 to nearly 2.87 million at the end of 2025. A plan written against that base is dealing with a stock of companies that already exist, most of which will not be registering for the first time; that is why so much of the content is about what happens after establishment rather than at the counter.

Six environments, twenty-nine areas

The plan is organised around six environments — market, rule of law, government services, investment and trade, humanistic, and the Jing-Jin-Ji region — and sets out 29 reform areas across them, an average of roughly five per environment. The development and reform commission’s summary of the intent is a “partner-style” service concept, aimed at an internationally competitive benchmark.

The distribution is not accidental. Each environment answers a different question a company asks.

The market environment answers whether the market is contestable: curbing involutionary competition, supporting small and medium-sized enterprises in bidding on fair terms, smoothing entry and exit, promoting credit commitments and smart monitoring of government contract performance, deepening the reform that lets a licence follow market entry, and rolling out application coaching across professionally demanding sectors.

Government services answer what a filing costs in time. The chapter carries numbers already in service: 137 “one thing at a time” scenarios that cut processing time by 57 per cent, materials by 50 per cent and trips by 87 per cent, and more than 1,400 services on the Jingtong app.

The rule of law environment answers what happens when something goes wrong: online handling of litigation, rules for commercial mediation organisations and model mediation rules, and a handbook on bankruptcy formalities issued with a pocket edition.

Investment and trade, humanistic and Jing-Jin-Ji answer the remaining questions — whether the city is open, whether people will come, and whether the region behaves as one market. For a foreign investor, the useful reading is that only a minority of the 29 areas touch the establishment file directly; the rest shape what the company meets in its second and third year.

Two market-environment items deserve a closer look, because they change the arithmetic of dealing with the administration rather than the paperwork.

Credit commitments replace a prior check with a promise and a later verification. For a company that can keep the promise, the trade is good: less waiting, fewer documents, faster decisions. For one that cannot, the same mechanism converts a documentary problem into a credit problem, which is a heavier thing to carry. The commitment is worth making only where the underlying facts are stable.

Smart monitoring of government contract performance runs in the other direction. It applies where the administration is the counterparty, and its subject is whether a contract is performed as written. For a supplier on a public contract, the practical implication is record-keeping: milestones, deliverables and acceptances documented as they happen, in a form that a system can read.

What “partner-style” commits the city to

The phrase the development and reform commission uses to summarise the plan is “partner-style” service. It is a framing rather than a rule, and framings are worth testing against the text.

Four mechanisms in the plan give the phrase content, and they have one thing in common: in each, the administration moves first.

The capital reminder reaches a company before the company has noticed a problem. Application coaching in professionally demanding sectors puts an official in the room before an application is filed, rather than returning it afterwards. The appointment-based service lets a company ask for an examination instead of waiting for a cycle. The list of companies left undisturbed is a commitment not to visit, published in advance.

That is a different posture from the one most compliance systems are built on, where the administration waits, the company files, the administration checks and, if something is wrong, penalises. The change of posture does not change any substantive rule, and it does not create an entitlement: nothing in the plan gives a company a right to advance guidance, and a planning document cannot create one.

What it does change is who is expected to speak first. A company that waits to be told will still be told — but the mechanisms are designed for the company that asks. In practice, that means knowing which channel exists for which question: coaching for an application, appointment for a borderline issue, the reminder for a capital schedule.

Five provisions foreign investors will feel

Most of the 29 areas are domestic housekeeping. Five touch what a foreign investor actually does: how much capital it must commit and by when, what it must prove about its address, what language it files in, where in the region it can collect a licence, and whether the market it enters is genuinely open to it.

The five share a property that makes them worth separating from the rest. Each one changes something at the boundary between the company and the administration — a deadline, a piece of evidence, a language, a location, a procurement rule. The others mostly change how the administration organises itself.

They also differ in maturity. Bilingual forms exist and are being extended. Licence handling across the region rests on 254 “same matter, same standard” items already published. The capital reminder and the domicile standardisation are directions with operating rules still to come. Local fair competition legislation is the least advanced of the five. Reading them as one package overstates how much is already in force.

The first year, provision by provision

Laid along a timeline, the five provisions land in a sequence that is easier to plan against than a list.

Before establishment, two of them apply. The domicile decision comes first, because it determines whether the filing is a data match or a paper file: a standardised address brings the benefit of no supporting documents, a bespoke one brings the folder. The language decision comes second, and it is simpler — the forms are bilingual, and the question is only how much of the surrounding submission is not.

At establishment, the relevant direction is the reform that lets a licence follow market entry, together with application coaching in sectors where the technical review is demanding. For a foreign investor in food, medical devices or another licensed sector, the coaching channel is the one worth locating before filing, because a returned application costs a cycle.

During the first year, the capital reminder becomes relevant. The schedule filed at establishment is the thing that will later be checked, and the check is easier to pass if the schedule and the payments have been kept consistent in the meantime.

On expansion, licence handling across the three places applies. It is a convenience for a group that already has entities in more than one of them, and it is not a substitute for the sector approval that sits behind the licence.

On market access, the fair competition track applies — not at establishment, but at the first tender, the first subsidy application, the first licensing condition that looks tailored.

Three of the five can be acted on now. Two await operating rules. Separating them that way is more useful than treating the plan as a single package to be implemented at some point in the next five years.

The sequence also explains why the five feel unequal in weight. A company meets the domicile question once, the capital question once a year at most, and the fair competition question every time it bids. Frequency, rather than novelty, is the better filter for deciding where to spend attention first.

Registered capital: what a “point-to-point” reminder is for

The plan deepens reform of the registered capital subscription system and adds a “point-to-point” reminder service: eligible companies are to be prompted and urged to adjust their contribution schedules in good time and to perform their contribution obligations according to law.

The subscription system lets a company set its capital and a payment schedule at establishment rather than paying in full up front. That flexibility produces a predictable administrative problem: a schedule set in year one can become inconsistent with the rules as they develop, and the company that set it may not notice.

A reminder mechanism is the administrative answer to a compliance gap that is nobody’s fault in particular. It assumes the administration can see, from registration and contribution data, which companies are drifting out of alignment, and it reaches them individually rather than through a general notice.

What the plan does not say is as important as what it does. “Eligible companies” is not defined. The trigger — a schedule that exceeds a permitted period, a paid-in amount that falls short of a filed figure, or some other mismatch — is not stated. Nor is the consequence of an unanswered reminder. Those three gaps are where the operating rules will do their work, and they are the reason this provision should be diarised rather than acted on immediately.

Two readings of the reminder, and why they differ in practice

There are two ways to read a mechanism described as “prompting and urging”, and a company’s response should depend on which one turns out to be right.

The first reading is housekeeping. The administration has noticed that a large number of companies set contribution schedules that no longer fit the rules, and prefers a nudge to a wave of enforcement. On this reading, a reminder is an invitation to file an adjustment, and a company that responds with a revised schedule has closed the matter.

The second reading is early intervention. A reminder is the first step in a sequence that ends in enforcement, and the company’s response — or silence — becomes part of the record. On this reading, the reminder is not advice but a deadline in disguise.

The plan’s own language supports the first reading: the verbs are “prompt” and “urge”, not “require” or “order”. But the practical difference between the two readings is large enough that a company should not wait for clarification before checking its own position. A foreign-invested company with a long-dated contribution schedule, a schedule set under earlier rules, or a share transfer that changed the shareholders without changing the schedule, is exactly the population such a mechanism is designed to find.

The sensible preparation is unglamorous: confirm what the filed schedule says, confirm what has actually been paid, and confirm whether the two are still consistent with the rules that now apply. That is a two-hour exercise that removes the question of which reading is correct.

Three documents answer most of it. The filed schedule, which is the reference point the administration will use. The payment records, which show what has actually happened rather than what was planned. And the record of any share transfer, because a change of shareholder does not automatically rewrite the contribution schedule, and a schedule inherited from a previous owner is still the company’s obligation.

A fourth item is worth adding for a group: the internal approval that authorised the schedule in the first place. If the schedule has to be adjusted, the same authority usually has to approve the adjustment, and finding that out after a reminder arrives is a slower path than finding it out now.

Domicile registration: why owner consent is listed separately

The plan promotes standardised registration of business premises, with a requirement that the “address information” and the “property owner’s intention” both be genuine, and a benefit for those who use a standardised address: no paper supporting documents.

The interesting part is the second element. If the purpose were only to confirm that an address exists, a registry extract would do. Listing the owner’s intention separately says something about where the problem actually lies.

An address can be entirely real and still be wrong as a registered domicile. Premises are sublet by a tenant who has no right to sublet; an agent signs a lease without the owner’s knowledge; a building is real, the suite number is not; a landlord has agreed orally and will deny it later. In each case the address information checks out and the consent does not. That is why the two are tested separately, and why a data-matching regime that verifies only the first would let the same failures through.

The benefit follows from the same logic. An applicant who uses a standardised address is spared the folder of leases, title certificates and consent letters because the two facts have already been verified in the system that issued the address. The trade is explicit: less paper in exchange for facts that hold up against data the administration can already see.

For a foreign-invested entity, the practical consequence depends on the building. Leasing in a standardised property usually means a faster filing. Using a bespoke address — a factory, a warehouse, a serviced office outside the library — does not make the paper route disappear, and the owner’s consent still has to be produced in a form the registry accepts.

Two questions are worth asking before a lease is signed, and neither is about rent.

The first is whether the building is in the standardised library at all. The benefit attaches to the address, not to the company, so a perfectly ordinary office in a building that has not been standardised gets none of it.

The second is how consent is recorded. In a standardised building the answer is usually systematic — the owner has agreed in advance, once, for the properties it registers. Outside that, consent is a document, and the document has to come from someone with the right to give it: not a subtenant, not an agent acting without written authority, and not a landlord whose title is itself disputed.

Bilingual forms: one line of continuity

The plan also promotes bilingual application forms.

That is a short sentence with a long history behind it. Beijing issued the country’s first set of bilingual registration documents for foreign-invested entities, covering companies and partnerships across establishment, change, deregistration and filing, with bilingual prompts on the online portal and the mini-programme. The plan’s contribution is to move that from a working arrangement into a stated direction for the next five years.

The limit is worth noting. Registration documents are one layer; sector licences are another, and they are administered by different departments with their own forms. A company that files its establishment in two languages may still translate a fire-safety submission or a drug licence application itself.

Jing-Jin-Ji: cause or result?

The plan sets out to make business licences handleable, issuable and collectable anywhere across the Beijing-Tianjin-Hebei region, building on 254 “same matter, same standard” items already released.

The obvious question is which of the two comes first. The “same matter, same standard” list is the substrate: it establishes that a given registration or filing is the same transaction in all three places, with the same materials and the same criteria. Only once that is true does it make sense for a licence to be handled in one city for an entity registered in another.

So licence recognition is best read as the visible layer of a deeper convergence, not as a standalone convenience. That reading has a practical consequence: the pace at which a licence can actually be collected elsewhere depends less on the licence system than on how completely the underlying items have been standardised, and on whether the registration data behind them is shared.

For a group with entities in more than one of the three places, the value is administrative: a filing that no longer requires a trip to the registration authority where the entity sits. The limit is clear: recognition covers the licence, not the sector approval behind it.

Fair competition legislation and the 500-document sample

Beijing will push for local legislation on fair competition, carry out sampling of policy measures at a rate of 500 documents a year, and run a campaign against the abuse of administrative power to exclude or restrict competition.

The three measures operate at different depths. The campaign addresses specific conduct. The sampling is a statistical instrument: 500 documents a year is a volume that only makes sense if review is routine rather than exceptional, and the city reports having sampled 440 policy documents since the start of 2026. Local legislation is the deepest, because a statute would give the review process a procedure and a footing that an annual programme cannot.

There is a national layer underneath. The amended implementation measures of the Fair Competition Review Regulation take effect on 1 December 2026, and a local statute would sit on top of that framework rather than replace it.

For a foreign company, the documents that matter are rarely the ones about foreign investment. They are the procurement rules, subsidy notices and licensing conditions that determine whether a bid is winnable, and those are precisely the documents a sampling programme is designed to reach.

What to do on finding one is less obvious than it sounds. The plan describes a review system and a campaign, not a complaints procedure; it does not say where a company takes a clause it believes is exclusionary. A local statute might add one — a procedure, a standing, a time limit — and that is one of the things to watch. Until then, the practical move is documentary: keep the notice, keep the scoring table, keep the timetable, and keep the record of what the company did in response.

From penalty after the fact to a check-up before

The supervision chapter reads as a deliberate change of posture. Scanned inspections are to reach full coverage, frontier technology including large artificial intelligence models is to be applied to off-site supervision, the list of companies left undisturbed is to be kept current, and an appointment-based service is to be introduced so that a problem is examined before it becomes a penalty.

The direction already shows in outcomes. In the first nine months of 2026, more than ninety per cent of inspections of foreign-invested enterprises in the city were conducted off site.

Two things follow for a company under this regime. The first is that the evidence base moves from the visit to the record: an off-site inspection reads filings, reports and data rather than walking a factory floor, so a company whose records are complete and consistent meets the inspection on its own terms.

The second is that “appointment-based” is an option, not an obligation, and options are worth using. A company that knows it has a borderline issue — a labelling variant, a filing that has drifted from practice — can invite the conversation before a routine cycle finds it. What the check-up produces is a finding; what the finding becomes depends on how it is handled, and handling it early is the whole point of the change of posture.

Two preparations make the difference between a check-up and an audit.

The first is that the record has to agree with itself. An off-site inspection compares filings against reports against data; the failures it finds are usually inconsistencies rather than violations — a figure that moved in one document and not in another, a change that was filed late, a capacity that was declared and never amended.

The second is knowing whether the company is on the undisturbed list. The list is a published commitment, and being on it is worth confirming rather than assuming, because the assumption is what a company discovers to be wrong at the worst moment.

The 137 scenarios: stock versus new commitments

The 137 “one thing at a time” scenarios are the plan’s most quotable numbers, and they are also the easiest to misread. They are stock, not new promises: the scenarios have been built, and the 57 per cent cut in processing time, 50 per cent cut in materials and 87 per cent cut in trips are the measured results of what already exists. The Jingtong app’s more than 1,400 services are in the same category.

What the plan adds is direction rather than a list. It does not publish which new scenarios will be built, and it does not say which of the existing ones will be extended to foreign-invested entities specifically.

That matters for planning. A company should treat the 137 as a description of the baseline it can already use, and treat the plan as a signal that the baseline will keep improving in the same direction — more matters handled as a single transaction, fewer materials, fewer trips. The practical question for a given company is narrow: which of its recurring filings is already a scenario, and which still requires assembling a file from scratch.

Answering it does not require the plan. It requires asking at the counter, or in the portal, whether a given combination of filings is handled as one matter — and then comparing that answer with what the company actually does each year. Most groups discover that two or three recurring procedures account for most of their administrative time, and that one of them is already covered.

Exit: the handbook and the mediation rules

The rule of law environment includes items that are easy to skip and expensive to need: a handbook on bankruptcy formalities, issued with a condensed pocket edition, and rules on the management of commercial mediation organisations and model mediation rules.

Exit is part of the business environment in the same way entry is. A market in which a company can be established quickly but cannot be wound up predictably is not a cheap market to enter, because the cost of failure is priced into every decision to try. A handbook does not change the law, but it changes how much of the law a finance director has to discover alone.

Mediation plays a similar role on the dispute side. It is cheaper and faster than litigation when it works, and its usefulness depends on the quality and predictability of the mediators — which is why rules on organisation management and model rules matter more than they sound.

There is a timing point that applies to both. An exit route is used at the worst moment, by people who are already under pressure, which is exactly when nobody has time to learn a procedure. The value of a handbook is that it can be read in a calm week and put aside; the value of a pocket edition is that it can be found again in a bad one.

What to watch over the next five years

  • The operating rules for the capital reminder: who is selected, on what data, and what follows an unanswered reminder.
  • Whether the standardised address library covers the building you intend to lease, and how owner consent is recorded in it.
  • Whether bilingual forms extend from registration into sector licences, where the translation burden actually sits.
  • Whether local fair competition legislation adds a route for complaints, or only a duty for drafters.
  • Which new “one thing at a time” scenarios are built, and whether any covers the filing you make most often.

None of these can be answered from the plan itself, and that is the correct expectation for a five-year document. The plan states a direction; the operating rules state what a company must do. The gap between the two is where the work of the next five years will happen — and where a foreign investor’s attention is best spent.

A practical way to hold the list is to assign each line to a function rather than to a file. Capital belongs with finance, domicile with administration, bilingual forms with legal, regional licensing with the company secretary, fair competition with whoever owns public bidding. The plan is written for a city; the list is written for the people who will have to act on it.

Sources

  • 北京市科学技术委员会、中关村科技园区管理委员会·《北京市”十五五”时期优化营商环境规划》政策解读:https://www.ncsti.gov.cn/zcfg/zcjd/202611/t20261109_262113.html
  • 北京市人民政府·《北京市”十五五”时期优化营商环境规划》发布报道:https://www.beijing.gov.cn/ywdt/gzdt/202611/t20261113_4944796.html
  • 辽宁省司法厅·外资企业办事改革实例(转载):https://sft.ln.gov.cn/sft/qmyfzs/tszs/2026110216071533603/index.shtml
  • 北京市人民政府·顺义区”十五五”高水平开放型经济发展规划:https://www.beijing.gov.cn/zhengce/zhengcefagui/202611/t20261111_4942194.html

北京第二部营商环境五年规划:29 个方面改革,与外资最直接相关的五条

为什么第二部五年规划值得看

北京发布了”十五五”时期优化营商环境规划。这是北京连续第二部营商环境五年专项规划,而”连续”本身就值得注意:上一部把机器装了起来——一窗办理、网上申报、单一窗口;这一部假定机器已经在那儿,开始调机器里跑的东西。

它是一份规划文件,不是一张清单。其中没有任何一项靠自己就能开通申请通道,条款要等配套细则和部门实施方案出来才具备可操作性。合理的读法是:把它当作方向陈述,细节随操作规则一起来。

五年规划能做而年度行动方案做不到的事,是设定年度报告将来按什么口径来写。一项在这里只以方向出现的措施——比如公平竞争地方立法——未来五年都会被拿来对照汇报,这和只宣布一个季度的试点,是两种不同性质的承诺。

它起步的基数是可量的。全市经营主体总量从 2020 年底的 211.5 万户增至 2025 年底的近 287 万户。对着这样一个基数写出来的规划,面对的是已经存在的存量企业,其中大多数不会再去办第一次登记;这也是为什么大量内容讲的是设立之后发生的事,而不是柜台上的事。

六大环境与 29 个方面

规划围绕市场、法治、政务、投资贸易、人文、京津冀六大环境展开,提出 29 个方面改革举措,平均每个环境约五项。市发展改革委对意图的概括是”伙伴式”服务理念,目标是打造国际一流营商环境标杆。

这个分布不是随意的。每一个环境,回答的是企业提出的一个不同问题。

市场环境回答的是市场是否可竞争:整治和防范”内卷式”竞争,加大支持中小企业公平参与招投标,畅通市场准入退出,健全社会信用体系,加强政府合同履约智慧监管,推广信用承诺制,深化”准入即准营”改革,对专业性较强领域全面推行申报辅导服务。

政务环境回答的是办一件事要花多少时间。这一章拿得出已经在跑的数字:137 个”高效办成一件事”场景,整体压缩时限 57%、精简材料 50%、减少跑动 87%;京通 App 上线 1400 余项服务。

法治环境回答的是出了事怎么办:提高涉诉网办水平,建立健全商事调解组织管理、示范规则等制度规范,同步发布《北京市企业破产事务办理一本通》及浓缩版口袋书。

投资贸易、人文、京津冀三章,回答的是剩下的问题——这座城市是否开放、人愿不愿意来、区域能不能当成一个市场来用。对外资而言,有用的读法是:29 个方面里只有少数直接碰到设立材料,其余塑造的是公司在第二、第三年会遇到什么。

市场环境里有两条值得单独看一眼,因为它们改变的是与行政机关打交道的算术,而不是材料本身。

信用承诺制用”先承诺、后核验”替代”先核验”。对能守住承诺的公司,这笔交易划算:少等、少交、决定更快。守不住的公司,同一套机制会把一个材料问题变成一个信用问题,而后者更重。只有在底层事实稳定的时候,承诺才值得作。

政府合同履约智慧监管的方向相反。它适用于行政机关作为合同一方的情形,管的是合同有没有按约履行。对承接政府合同的供应商,实务含义落在记录上:节点、交付物、验收,发生时就要留下可被系统读取的痕迹。

“伙伴式”这三个字承诺了什么

市发展改革委用来概括这份规划的说法是”伙伴式”服务理念。这是一种框定,不是一条规则,而框定值得拿文本去检验。

规划里有四项机制给这三个字填了内容,它们的共同点是:都由行政机关先动。

注册资本提醒,在公司自己发现问题之前先找到公司。专业性较强领域的申报辅导,是在材料递上去之前把官方的意见放到桌面上,而不是事后退回。预约式服务,让公司主动要求检查,而不是等一个周期轮到自己。无事不扰企业清单,是一份提前公布的不上门承诺。

这与多数合规体系赖以建立的那种姿态不同——机关等、企业报、机关查、有问题就罚。姿态变了,实体规则一条没变,也没有创设权利:规划没有给企业”事先获得指导”的权利,一份规划文件也创不出来。

真正变的是”该谁先开口”。等着被通知的公司仍然会被通知,但这几项机制是为主动发问的公司设计的。落到操作上,就是知道哪个问题走哪条通道:申报走辅导,边界问题走预约,缴款安排走提醒。

与外资直接相关的五条

29 个方面里,大部分是内部的自我整理。真正落到外资企业手上的有五条:要承诺多少注册资本、什么时候缴;要证明什么样的地址;用什么语言递材料;在区域里哪儿能领执照;以及它要进入的市场是否真的对它开放。

这五条有一个共同点,使它们值得从其余条款里单拎出来:每一条改变的都是企业与行政机关交界处的东西——一个期限、一份证据、一种语言、一个地点、一条采购规则。其余的,大多改变的是行政机关怎么组织自己。

它们的成熟度也不一样。双语文书已经存在,正在扩展;京津冀执照互办依托的是已经公布的 254 项”同事同标”事项;注册资本提醒和住所标准化是方向,操作规则还没出来;公平竞争地方立法是五条里推进最靠后的一条。把它们当成一个已经落地的整体来读,会高估当下已经生效的部分。

把五条摊在时间轴上

五条按时间摊开,会呈现出一个比清单更好用的顺序。

设立之前,用得上其中两条。先是住所的取舍,因为它决定这次备案是一次数据比对还是一摞纸:标准化地址带来免交证明的好处,非标准地址带来一个文件夹。然后是语言的取舍,这一条简单——表单是双语的,问题只在于周边材料里还有多少不是。

设立当时,起作用的是”准入即准营”这项改革,以及技术审查要求较高的那些领域的申报辅导。对食品、医疗器械等许可类行业的外资企业,值得在递材料之前先找到辅导通道,因为一次退回就是一个周期。

设立后第一年,注册资本提醒开始相关。设立时填报的那份缴款安排,就是日后会被核对的东西,而只要安排与实缴在这期间保持一致,核对就容易通过。

扩张时,用得上京津冀执照异地办发领。这对在三地已有主体的集团是便利,但它不能替代执照背后的行业许可。

市场准入这一侧,公平竞争那条线在第一次投标、第一次申报补贴、第一次遇到像是量身定制的许可条件时才真正用得上——不在设立阶段。

五条里现在就能动手的有三条,等操作规则的有一条半。这样分开看,比把整份规划当成一个”五年后统一落地”的包裹要有用。

这个顺序也解释了为什么五条的分量并不相等。住所只遇到一次,注册资本一年最多遇到一次,公平竞争则是每次投标都会遇到。决定先花力气在哪一条,该看频次,而不是看新鲜度。

注册资本:这条”点对点”提醒是干什么的

规划提出深化注册资本认缴登记制度改革,建立完善”点对点”提醒服务机制,提示、督促符合条件的公司及时调整出资期限、依法履行出资义务。

认缴制允许公司在设立时自主约定资本与缴款安排,而不是一次缴足。这种灵活会带来一个可预期的管理问题:第一年定下的缴款安排,可能随着规则演变而不再吻合,而定下它的公司未必会注意到。

提醒机制是对一个”谁也不算特别有错”的合规缺口的行政回应。它假定登记机关能从登记与实缴数据里看出哪些公司正在偏离,并且逐个触达,而不是发一纸通告了事。

规划没说的部分和说了的部分同样重要。”符合条件的公司”没有定义;触发条件——是缴款期限超过了允许的区间,还是实缴金额与填报数字不符,或是别的错配——没有写明;提醒之后仍不处理会怎样,也没有写。这三处空白正是操作规则要填的,也是这一条应当先记进日历、而不必立刻行动的原因。

两种读法,实务差别很大

对一个被描述为”提示、督促”的机制,有两种读法,而企业的应对应当取决于最后哪一种成立。

第一种读法是内务整理。登记机关发现大量公司的缴款安排已经不合规则,宁愿先提醒,而不是一次性掀起执法潮。按这种读法,提醒是请你来改一份安排,公司交一份修订后的缴款计划,事情就结了。

第二种读法是提前介入。提醒是一连串动作的第一步,终点是执法,而公司的回应——或者沉默——都会成为记录的一部分。按这种读法,提醒不是建议,而是伪装成建议的期限。

规划自己的用词支持第一种读法:”提示””督促”,而不是”要求””责令”。但两种读法在实务上的差别足够大,企业不该等到口径明确再来核对自己的位置。缴款期限排得很远的外资公司、按旧规则定下的安排、发生过股权转让而缴款安排没跟着改的公司,恰恰是这类机制设计出来要找的那批人。

稳妥的准备并不体面:确认备案的缴款安排写的是什么,确认实际缴了多少,再确认这两者与现行规则是否仍然吻合。这是一件两小时能做完的事,做完就不必再猜哪种读法才对。

三份文件能回答其中大部分。备案的缴款安排,是登记机关会拿来对照的基准。缴款记录,说明实际发生了什么,而不是当初计划了什么。股权转让的记录也要备着,因为股东变了,缴款安排不会自动跟着重写,而从前手继承下来的安排,仍然是这家公司的义务。

对集团还要加第四项:当初批准这套安排的内部决议。安排要调整时,通常还是同一个层级来批;等提醒到了才发现这一点,比现在就发现要慢一步。

住所登记:产权人意愿为什么要单列

规划推广经营场所(住所)标准化登记模式,要求”地址信息”和”产权人意愿”双真实;对使用标准化地址的经营主体,支持免于提交纸质证明材料。

有意思的是第二项。如果目的只是确认地址存在,一份产权摘录就够了。把产权人意愿单独列出来,说明问题真正出在哪里。

一个地址可以完全真实,却仍然不适合作为登记住所。房子被没有转租权的承租人转租;中介在产权人不知情的情况下签了租约;楼是真的,房号不是;房东口头答应了,事后可以不认。这几种情形里,地址信息都过得去,意愿过不去。这就是两者要分开验的原因,也是只验第一项的数据比对会漏掉同一批问题的原因。

便利来自同一个逻辑。使用标准化地址的申请人省下一叠租赁合同、产权证明和同意书,是因为这两件事已经在发出该地址的系统里被验过。交换说得很明白:少交纸,换来的是那两件事必须经得起登记机关已有数据的比对。

对外资企业,实际后果取决于楼。在标准化楼宇里租房,通常意味着备案更快;使用非标准地址——厂房、仓库、不在库里的服务式办公室——纸质路径并不会消失,产权人意愿仍要以登记机关接受的形式拿出来。

有两个问题值得在签租约之前问,且都跟租金无关。

第一个是这栋楼在不在标准化地址库里。便利附着在地址上,不附着在公司上,所以一间再普通不过的办公室,只要所在的楼没被标准化,就一点也享受不到。

第二个是意愿怎么记录。在标准化楼宇里,答案通常是系统性的——产权人对它登记在册的房产一次性预先同意过。出了这个范围,意愿就是一份文件,而这份文件必须由有权作出的人出具:不是转租人,不是没有书面授权的代理人,也不是产权本身还有争议的房东。

双语文书:一句话的延续

规划同时提出推广中英文对照版申请文书。

这句话很短,背后是一条很长的线。北京已推出全国首套中英文对照版登记申请文书,覆盖公司、合伙企业等各类外资主体,适用设立、变更、注销、备案场景,网上平台与小程序均提供双语操作提示。规划的作用,是把这件事从一项工作安排,变成未来五年的明确方向。

边界也要说清。登记文书只是一层,行业许可另有一层,由不同部门主管、用各自的表单。以双语完成设立登记的公司,可能仍然要自己翻译消防申报材料或药品许可申请。

京津冀:执照互办是结果还是前提

规划提出加快实现京津冀全域营业执照异地”办、发、领”,此前已累计推出 254 项京津冀”同事同标”事项。

显而易见的问题是:这两件事谁在先。”同事同标”清单是底子——它确认某一项登记或备案在三地是同一件事,材料一样、标准一样。只有这件事成立之后,一张执照在甲地为乙地登记的主体办理,才说得通。

所以执照互办更适合读成更深一层趋同的可见表层,而不是一项孤立的便利。这个读法有实务后果:一张执照究竟能在异地多顺畅地领到,取决于底层事项标准化的完成度,也取决于其背后的登记数据是否共享,而不是取决于执照系统本身。

对在三地都有主体的集团,这项便利是行政性的:一次备案不再需要跑到主体所在地的登记机关。边界同样清楚——互认覆盖的是执照本身,不是执照背后的行业许可。

公平竞争立法与每年 500 件的抽查

北京将积极推动公平竞争地方立法,严格开展公平竞争审查抽查,每年抽查政策措施 500 件,并开展整治滥用行政权力排除、限制竞争专项行动。

三项措施的作用深度不同。专项行动针对具体行为;抽查是统计工具——每年 500 件,只有审查是常态而非例外才需要这样的量,而北京披露 2026 年以来已抽查政策文件 440 件;地方立法最深,因为一部法规能给审查程序一套程序和一个落脚点,而年度计划给不了。

下面还有一层国家层面的框架。修改后的《公平竞争审查条例实施办法》自 2026 年 12 月 1 日起施行,地方立法会是叠加在这套框架之上,而不是替代它。

对外资企业来说,真正相关的文件很少是”关于外商投资”的那一类,而是采购规则、补贴通知和许可条件——它们决定一单投标能不能赢,也恰恰是抽查这类机制想要触及的文件。

发现问题之后怎么做,比想象中含糊。规划描述的是审查体系与专项行动,不是投诉程序;它没有说企业认为某条条款具有排除性时该往哪里送。地方立法有可能补上这一块——一套程序、一种主体资格、一个时限——那也是值得盯的地方之一。在那之前,稳妥的做法是留证据:留下公告、留下评分表、留下时间安排,也留下公司当时是怎么应对的记录。

从”事后处罚”到”事前体检”

监管这一章读起来像一次姿态调整。”扫码检查”要全覆盖,人工智能大模型等前沿技术要用于非现场监管,”无事不扰”企业清单要持续完善,并推行”预约式”服务,把问题在变成处罚之前先查一遍。

方向在结果上已经看得出来。2026 年 1—9 月,全市外资企业非现场检查占比超过九成。

在这套监管之下,企业要跟上两件事。第一,证据的基础从”来人”转到了”记录”:非现场检查读的是备案、报告和数据,而不是走一遍车间,所以记录完整、前后一致的公司,是在用自己的方式应对检查。

第二,”预约式”是一个选项,不是义务,而选项值得用。知道自己有一处边界问题的公司——标签的某个变体、与实操已经脱节的某项备案——可以在例行周期找上门之前主动发起对话。体检产出的是一个发现;这个发现会变成什么,取决于怎么处理,而提前处理正是这次姿态调整的全部意义。

两项准备,决定了来的是体检还是审计。

第一项是记录要自洽。非现场检查做的是备案、报告与数据之间的比对,它找出来的问题通常不是违规,而是不一致——一个数字在甲文件里改了、在乙文件里没改,一次变更报晚了,一项产能申报过却从未修正。

第二项是知道自己在不在无事不扰清单上。清单是一份公开承诺,在不在上面值得去确认,而不是想当然,因为想当然这件事,往往在最糟的时候被证明想错了。

137 个场景:存量成绩与新增承诺

137 个”高效办成一件事”场景是规划里最好引用的数字,也最容易被误读。它们是存量,不是新承诺:场景已经建成,时限压缩 57%、材料精简 50%、跑动减少 87% 是既有成果的实测值。京通 App 上线 1400 余项服务,属于同一类。

规划增加的是方向,而不是一张清单。它没有公布接下来要建哪些新场景,也没有说现有场景中哪些会专门向外资企业延伸。

这对规划安排有影响。企业应当把 137 个场景当成”现在就能用的基线”,把规划当成”基线会沿同一方向继续改善”的信号——更多事项按一件事办、更少材料、更少跑动。对某一家具体公司,问题很窄:自己反复要办的那些事项里,哪些已经是场景,哪些还要从头攒材料。

回答这个问题不需要读规划,需要在窗口或系统里问一句:某几项申报能不能按一件事办。然后把答案和公司每年实际要做的事对一遍。多数集团会发现,两三道反复办理的程序占掉了大部分行政时间,而其中一道已经在场景里。

退出:一本通与调解规则

法治环境里有几项容易被跳过、一旦需要就很贵的安排:《北京市企业破产事务办理一本通》及浓缩版口袋书,以及商事调解组织管理、示范规则等制度规范。

退出和进入一样,是营商环境的一部分。一个能快速设立、却不能可预期地清算退出的市场,进入成本并不低,因为失败的代价会被计入每一次尝试的决定里。一本通不改变法律,但它改变一位财务负责人要独自摸索多少法律。

调解在争议一侧扮演类似角色。它管用的时候比诉讼更便宜、更快,而它管不管用,取决于调解员队伍的质量与可预期性——这也是为什么组织管理与示范规则这类听起来枯燥的东西,比字面上更重要。

两者还有一个共同的时点问题。退出通道是在最糟糕的时刻被用上的,用的人已经承压,而那恰恰是没人有时间学一套程序的时候。一本通的价值在于它可以在一个平静的星期里读完、然后搁在一边;口袋书的价值在于到了不平静的时候,还能把它翻出来。

未来五年该盯的五件事

  • 注册资本提醒的操作规则:谁被选中、依据什么数据、提醒之后不处理会怎样。
  • 标准化地址库是否覆盖你打算租的楼宇,产权人意愿如何录入其中。
  • 双语文书是否从登记延伸到行业许可——翻译负担真正压在那里。
  • 公平竞争地方立法是否带来投诉通道,还是只给起草方增加义务。
  • 新增的”高效办成一件事”场景有哪些,其中是否有你办得最频繁的那一项。

这些问题都无法从规划本身得到答案,而对一份五年文件来说,这本就是正确的预期。规划说的是方向,操作规则说的是企业必须做什么。两者之间的空隙,就是未来五年工作会发生的地方,也是外资注意力最该放的地方。

一个可用的办法是:把这几条按职能分下去,而不是按文件归档。注册资本归财务,住所归行政,双语文书归法务,区域执照归公司秘书,公平竞争归负责公开投标的那个人。规划是写给一座城市的,这份清单是写给要照着做事的人的。

来源

  • 北京市科学技术委员会、中关村科技园区管理委员会·《北京市”十五五”时期优化营商环境规划》政策解读:https://www.ncsti.gov.cn/zcfg/zcjd/202611/t20261109_262113.html
  • 北京市人民政府·《北京市”十五五”时期优化营商环境规划》发布报道:https://www.beijing.gov.cn/ywdt/gzdt/202611/t20261113_4944796.html
  • 辽宁省司法厅·外资企业办事改革实例(转载):https://sft.ln.gov.cn/sft/qmyfzs/tszs/2026110216071533603/index.shtml
  • 北京市人民政府·顺义区”十五五”高水平开放型经济发展规划:https://www.beijing.gov.cn/zhengce/zhengcefagui/202611/t20261111_4942194.html

Professional Business Consulting for International Companies in China. | ylinsight | YINGLONG INSIGHT