CICC: China's Investment Bank, Reinvented Three Times

I. Cold Open & Roadmap

On the morning of November 19, 2025, trading in the shares of 中国国际金融股份有限公司 China International Capital Corporation stopped. In Hong Kong the ticker 3908 went dark; in Shanghai, 601995 did the same. Halts happen. What followed did not.

CICC — the firm Chinese financial media had spent thirty years calling "China's Goldman Sachs," the house that took 中国电信 China Telecom and 中国石油 PetroChina public, the place where a generation of Beijing and Hong Kong bankers learned how a prospectus was supposed to read — announced it was absorbing two smaller state-owned rivals whole. 东兴证券 Dongxing Securities and 信达证券 Cinda Securities would be merged into CICC via an all-share swap, creating a brokerage with roughly RMB 1 trillion in assets essentially overnight.¹²

Read that again, because the direction of travel is the story. The elite advisory boutique — the one whose entire brand was built on not being a mass-market broker — was buying two mass-market brokers. The prestige firm was buying scale. And it wasn't buying them from the market. Both targets, like CICC itself, were controlled by 中央汇金投资 Central Huijin Investment, the domestic investment arm of China's sovereign wealth fund. This was less an acquisition than a reorganisation of a single owner's portfolio, executed through public markets.³

Here is the thing worth holding onto for the next three hours of reading: this is the third time CICC has been rebuilt into something structurally different from what it was, and all three times the decisive hand belonged to a shareholder rather than to management.

Version one was born in 1995 as China's first Sino-foreign joint-venture investment bank, co-founded with Morgan Stanley for the explicit purpose of teaching Chinese bankers how to take state enterprises public on Western exchanges. Version two arrived in 2016, when CICC bought a Shenzhen retail brokerage from its own controlling shareholder and quietly stopped being an advisory firm at all — today wealth management, not investment banking, is its largest revenue line by a wide margin. Version three is happening right now, in real time, as Beijing pushes to build what Chinese commentary calls 航母级券商 "aircraft-carrier-class" securities firms, and Central Huijin uses CICC as the vessel.

Along the way there is a divorce (Morgan Stanley's, expensive and slow), a scandal that started with a woman posting her husband's payslip on a lifestyle app, a compensation collapse that halved average pay across a 14,000-person firm, a regulator's fine for shoddy diligence on a fraudulent chip-company IPO, and a shareholder register that swapped 腾讯 Tencent and 阿里巴巴 Alibaba for E Fund and BlackRock.

This piece covers all of it: the founding and the Morgan Stanley exit; the pivot from prestige M&A shop to mass wealth manager; the 2022 payslip affair and the pay-cut years that followed; who runs the place now; the 2025 three-way merger and what it reveals about how China intends to restructure its financial sector; and the honest bull and bear case for owning a state-directed national champion where the controlling shareholder's agenda and the minority shareholder's returns are related but not identical.

Start where it started — with a joint venture that almost nobody thought would work.

↑ Back to Top

II. Origins: Building China's First Joint-Venture Investment Bank (1995–2010)

In 1995, China had a stock market that was five years old, a banking system carrying enormous unrecognised bad loans, and a portfolio of state-owned enterprises that Beijing had decided to restructure and list. What it did not have was anyone who knew how to do that. Nobody in China had written an international prospectus. Nobody had run a global roadshow, priced a book, or explained to a Boston fund manager why a Chinese oil company's reserve accounting could be trusted.

So Beijing did what a pragmatic state does when it needs a capability it cannot build: it imported one, on terms that guaranteed it would keep control.

China Construction Bank took 42.5% of the new venture. Morgan Stanley International took 35%. Three smaller partners — including Singapore's GIC — took 7.5% each. The result was China International Capital Corporation, the country's first Sino-foreign joint-venture securities firm, and it was explicitly designed as a technology transfer vehicle. Morgan Stanley would supply the know-how, the deal templates, the training. Beijing would supply the deal flow, which in 1990s China meant the privatisation pipeline: telecoms, oil, insurance, the banks themselves.

Why a joint venture rather than simply letting Morgan Stanley open a Beijing office? Because China's capital markets were closed to foreign control, and would remain so for another two decades. The JV was the only legal structure through which foreign capital-markets expertise could reach Chinese state issuers, and it had a feature Beijing liked a great deal: the controlling stake, the chairmanship, and the relationships stayed onshore. Morgan Stanley got a seat at the table and a share of the economics. It did not get the steering wheel.

For a while, that trade worked spectacularly for both sides. CICC became the house that landmark Chinese state listings went through, and the template it created — local relationships plus imported technical process — was copied by every foreign bank that later formed a China JV. If you were a Chinese SOE chairman in 2003 preparing for a New York or Hong Kong listing, CICC was the default first call.

It is worth being precise about what that advantage actually consisted of, because it is easy to romanticise. CICC's edge was not superior valuation technique; discounted cash flow works the same in Beijing as in Manhattan. Its edge was that it could sit in a room with a ministry, a state enterprise's party committee, and a group of London and Boston institutional investors, and act as a translator in both directions — explaining to the issuer what disclosure the market would require, and explaining to the market what a Chinese state enterprise's accounts actually meant. That is a brokerage function in the original sense of the word: standing between two parties who cannot easily transact directly. It is also, structurally, a role that shrinks as both sides learn. Every Chinese CFO who completed an international listing became slightly less dependent on a translator; every global fund that built a China team did too.

CICC's internal culture through this era acquired a distinctive reputation — technically rigorous, hierarchical, and politically well-connected, with senior leadership drawn from families deep inside the Chinese establishment. Zhu Levin, son of former Premier 朱镕基 Zhu Rongji, ran the firm as chief executive for roughly a decade into the mid-2010s and is inseparable from the era in which CICC's league-table dominance in state-enterprise listings was near-absolute. The details of that period matter less to a 2026 investor than the structural point it illustrates: CICC's franchise was built on access, and access in China is a function of who is standing where, which makes it powerful and impermanent in equal measure.

But joint ventures are structurally unstable when one partner is learning and the other is teaching, because the moment the student stops needing the teacher, the arrangement's logic dissolves. Through the 2000s, the partnership frayed over exactly what you would expect: who ran the firm, whose strategy it followed, and whether Morgan Stanley — which by then wanted its own wholly-controlled China platform — could keep a minority position in a competitor while building a rival.

The divorce took years and required regulatory blessing at every step. In December 2010, Morgan Stanley agreed to sell its 34.3% stake to a consortium of TPG Capital, KKR, GIC and Great Eastern Life, a deal that closed after Chinese regulators signed off and booked Morgan Stanley a pre-tax gain of roughly US$700 million.⁴ Fifteen years of patient capital in a business it never controlled, exited at a very large profit — and, crucially, exited into the arms of financial sponsors rather than a strategic rival.

That detail is not incidental. It is the pattern. Note who bought: private equity firms and sovereign funds, all of them financial investors with finite holding periods and no interest in operating control. Both PE sponsors eventually monetised well — KKR took roughly US$195 million off the table in a partial 2017 exit, and TPG realised about US$303 million in 2019, a return in the region of 40%. Financial owners came, held, and left. Through every one of those transitions, the Chinese state's position at the centre of the register never moved.

That is the throughline worth carrying into everything that follows. CICC has always had a rotating cast of financial shareholders orbiting a permanent state anchor. The names in orbit change — Morgan Stanley, then TPG and KKR, then Tencent and Alibaba, and by 2026 E Fund and BlackRock. The anchor does not. And when the anchor decides the firm should be something different, the firm becomes something different.

By 2010, with Morgan Stanley heading for the exit, the question was who that anchor would be — and the answer arrived in the form of an entity most Western investors had barely heard of.

↑ Back to Top

III. Central Huijin Takes the Wheel & the Dual Listing Era (2010s–2020)

Central Huijin is one of the more remarkable institutions in global finance, and one of the least discussed. It is a subsidiary of China Investment Corporation, the sovereign wealth fund, and its domestic mandate is to hold and exercise the state's equity in China's major financial institutions — the big commercial banks, insurers, and a stable of securities firms. It does not run them day to day. It appoints the people who do, and it decides what the portfolio should look like.

As Morgan Stanley's stake dispersed, Central Huijin consolidated its position at the top of CICC's register. Today it holds just over 40% of total shares directly — 40.11% as of the 2025 annual report — a stake that has been notably stable while everything around it churned.⁵ That is the anchor, quantified.

What followed through the 2010s was CICC's transformation from a privately-held JV into a dual-listed public company, and both listings are case studies in something CICC does unusually well: timing a capital raise to a policy inflection.

The Hong Kong listing came first. In early November 2015, CICC priced 1.42 billion H-shares at the top of its HK$9.12–10.28 range, raising roughly US$811 million, and the stock jumped nearly 10% on debut to HK$11.38.⁶ The context is what makes it interesting. Days earlier, China's securities regulator had ended a four-month freeze on mainland IPOs imposed during that summer's market crash. Investors bidding for CICC shares were not primarily buying a bank; they were buying a call option on the reopening of the Chinese IPO machine, on the assumption that CICC would underwrite a large share of the backlog. The listing arrived at precisely the moment that option was worth the most.

The A-share listing came five years later and was larger in every dimension. In early November 2020, CICC listed on the Shanghai Stock Exchange main board under 601995, raising roughly RMB 13.2 billion — about US$2 billion — and becoming one of the first joint-venture-heritage brokers to trade as a genuine A+H dual listing.⁷ The strategic-investor book was the tell: 17 strategic investors subscribed for roughly RMB 3.96 billion of the offering, about 30% of the deal, and the list read like a map of who Beijing wanted on the register. China Structural Reform Fund and China Reform Holdings each took the largest single allocations. The National Social Security Fund participated. So did Abu Dhabi Investment Authority and the Brunei Investment Agency — foreign sovereign capital, invited in on the same terms as the domestic "national team."

Read as capital allocation, the two listings tell a consistent story: CICC raises when the policy window is open and the narrative is favourable, not when it is desperate. That is a mark of discipline. Read as governance, they tell a different story: an A+H structure means two regulators, two disclosure regimes, two shareholder classes with different information sets and different exit options — a complexity that becomes material later, when a merger requires separate class meetings to approve it.

The most colourful chapter of the dual-listing era, though, involved neither sovereign funds nor regulators. In September 2017, Tencent bought roughly 5% of CICC's total shares for about US$372 million, a stake worth around HK$2.9 billion.⁸ In early 2019, Alibaba followed, acquiring about 4.84%.⁹ For a brief moment, China's two dominant internet platforms were both shareholders in the country's most prestigious investment bank, and the strategic logic seemed obvious to everyone: platform-era finance, where hundreds of millions of retail users flow through a super-app into wealth products distributed by a trusted institutional brand.

Hold that thought. The unwinding of both positions in 2025 and 2026 is one of the more instructive parts of this story, and we will come back to it.

Because in between Tencent's arrival and its departure, CICC did something that changed what the company actually is — and it did it almost without the outside world noticing.

↑ Back to Top

IV. The Pivot Nobody Expected: From Elite M&A Shop to Mass Wealth Manager (2016–today)

Here is the single most under-appreciated fact about CICC, and it takes some getting used to if your mental model was formed by the "China's Goldman Sachs" framing.

Investment banking is not CICC's main business. It has not been for years, and it is not close.

Look at the segment disclosure. In FY2024, on the mainland reporting basis, CICC's largest revenue segment was wealth management at roughly 32.7% of the RMB 21.33 billion top line. Equities came second at about 20.8%, fixed income at about 17.4%, and investment banking — the flagship, the brand, the reason anyone has heard of the firm — was fourth, at roughly 12.1%. Asset management contributed about 5.1% and private equity about 3.6%.¹⁰

A note on a genuine data trap here, because it matters for anyone comparing sources. Depending on which revenue base you use, investment banking's share of CICC looks either like 12% or like something closer to 6–7%. The lower figure appears when banking fee income is measured against the broader HKFRS "revenue and other income" line, which grosses up interest and certain investment income and therefore inflates the denominator. The segment figures above use the mainland-GAAP segment presentation, which sums cleanly to 100% and is the cleaner comparison across business lines. Both bases tell the same directional story — the flagship is a minority of the business — but the specific percentage is not interchangeable between them. Cite carefully.

How did a firm built to advise state enterprises end up as a wealth manager? One transaction.

In November 2016, CICC announced it would acquire 中国投资证券 China Investment Securities, a Shenzhen-based retail brokerage, for RMB 16.7 billion.¹¹ The seller was Central Huijin, which owned CISC outright. The consideration was paid in CICC shares, which is why Central Huijin's economic interest in CICC — held directly and through affiliates — rose sharply on completion.

Contemporaneous English coverage of the price was inconsistent, and it is worth flagging: some reports converted RMB 16.7 billion to roughly US$247 million, an obvious order-of-magnitude error, while Bloomberg and Caixin both reported the deal at approximately US$2.5 billion, which is the arithmetic that actually works at 2016 exchange rates. Anyone modelling this deal from secondary sources should use the RMB figure.

What CICC bought was not intellectual capital. It was pipes. CISC came with a branch network across mainland China and a retail client base numbering in the millions — the kind of distribution CICC had spent two decades deliberately not building, because its entire positioning was that it served institutions and the very rich, not the mass affluent. The acquired business was subsequently rebranded 中金财富 CICC Wealth, and it became the vehicle through which CICC's product manufacturing could reach an ordinary Chinese saver.

Now assess this as capital allocation, because the structure demands scrutiny. Central Huijin controlled both the buyer and the seller. It set the price, approved the price, and received the consideration in shares of the company it already controlled. There was no competitive auction, no arm's-length negotiation between adverse principals, and the effect of the transaction was to increase the controlling shareholder's stake in the listed entity. In any Western market this would be a related-party transaction requiring independent valuation, a fairness opinion, and a minority shareholder vote — and it would still attract activist attention.

Was it a good deal? The honest answer, ten years on, is that the strategic logic proved sound and the price is unverifiable against a market benchmark, because no market benchmark was ever created. That is precisely the problem with related-party deals: even when they work, minority shareholders cannot know whether they got a fair share of the value. What can be said is that Chinese retail brokerages in 2016 were not scarce assets, and a buyer with a captive seller and no competing bidders has structurally weak grounds for claiming it drove a hard bargain.

It helps to be concrete about why these two businesses behave so differently, because the distinction drives everything that follows. An investment bank earns a fee when a transaction happens — a percentage of an IPO, a fee on an M&A deal. No transaction, no revenue. It is closer to a construction contractor than to a landlord: the pipeline can be full or empty, and when it is empty the bankers still need paying. A wealth manager earns a fee on assets it holds and advises on, charged continuously. It is the landlord. Rent falls when markets fall and rises when they rise, but the tenants do not all move out at once. Bolting a landlord onto a contractor is a deliberate reduction in earnings volatility, purchased at the cost of a lower peak.

The strategic case, though, has held up. Investment banking revenue is a function of the IPO approval calendar, and the IPO approval calendar in China is a policy instrument. That makes advisory revenue not merely cyclical but politically cyclical — it can go to near-zero for reasons that have nothing to do with the quality of the franchise. Wealth management, by contrast, is an asset-gathering business: clients deposit assets, assets generate fees, and the fee base moves with markets but does not switch off. Grafting an annuity stream onto a lumpy advisory business was a genuine risk-management decision, and when the downturn came in 2022–2024, it was the reason CICC's revenue fell rather than collapsed.

The operating evidence supports that the pipes have been used. Through 2024, CICC opened more than 1.7 million new client accounts and grew product holdings under its asset-allocation platform to roughly RMB 370 billion.¹⁰ By the end of 2025, total client account assets reached about RMB 4.28 trillion, and wealth management revenue rose 35.9% to roughly RMB 9.49 billion, comfortably the largest segment.¹² Within that, the buy-side advisory book — where CICC charges on assets advised rather than on transactions, the Chinese analogue of a fee-based RIA model — passed RMB 130 billion.¹³

That last number deserves the emphasis rather than the headline. Buy-side advisory is the part of wealth management that is genuinely sticky and genuinely aligned; commission-driven product pushing is neither. RMB 130 billion out of RMB 4.28 trillion in client assets is roughly 3% — real progress from a standing start, and nowhere near a transformation. The honest read is that CICC has built distribution scale and is in the early innings of converting it into recurring, advice-based economics. Whether it finishes that conversion is one of the two or three things actually worth tracking.

Around the core, two smaller businesses deserve a mention rather than a section. CICC Capital, the private equity arm, managed roughly RMB 524 billion by end-2025, with an investment footprint stretching from South Korea to the Gulf. Asset management ran about RMB 597 billion.¹³ Both are real and both are growing; neither is what determines the company's earnings in a given year.

Which raises the obvious question. If investment banking is a tenth of the business, why does everyone — including, frequently, CICC — still talk about it as though it were the whole company?

↑ Back to Top

V. What Investment Banking Still Means for CICC — and the Competitive Landscape

Because the brand is the business, even when the revenue is not.

In 2025, Hong Kong's IPO market came roaring back. The exchange hosted 117 new listings. CICC participated in 53 of them, sponsored 42, and underwrote more than US$10 billion — ranking first in Hong Kong IPO sponsorship, ahead of 中信证券 CITIC Securities and 华泰证券 Huatai.¹⁴ Across Chinese issuers globally, CICC ranked first as sponsor, arranging roughly US$26.8 billion of IPO financing.¹³ It topped both the China M&A and China domestic M&A league tables, advising on 86 transactions worth more than US$111 billion — including 宁德时代 CATL's US$5.25 billion Hong Kong listing, the largest IPO anywhere in the world since 2023, and 海尔 Haier's US$1.8 billion acquisition of 汽车之家 Autohome.¹⁴

Segment revenue followed: investment banking grew nearly 78% in 2025 to about RMB 4.6 billion, at a segment operating margin north of 42%.¹² Equities did better still, up more than 65% to roughly RMB 7.3 billion at an 81% segment margin — a reminder that in a good year, the trading and prime businesses out-earn the advisory franchise several times over.

A word on what "sponsor" means, because the term does a lot of work in Hong Kong and is not intuitive. Underwriting a listing means helping sell the shares. Sponsoring one means signing your name to the statement that you have investigated the company and that its prospectus is not misleading — a legal responsibility, enforceable against the sponsor firm and, increasingly, against the individuals who signed. Sponsorship is the highest-liability, highest-status role in a listing, and exchanges track who takes it because it is a reasonable proxy for which firms issuers actually trust with the hard part. Ranking first in Hong Kong sponsorship is therefore a stronger signal than ranking first in underwriting volume, which can be bought by joining large syndicates cheaply.

So the "elite dealmaker" claim survives contact with evidence. League tables are audited, public, and hard to fake. What the claim does not survive is the assumption that league-table leadership translates into durable economics. Consider the swing: mainland equity financing arranged by CICC grew more than 580% year-on-year in 2025.¹⁴ A business that can grow 580% in a year is a business that can shrink 80% in a year, and it has.

Now map the battlefield, because CICC's competitive position looks very different depending on whether you measure prestige or scale.

中信证券 CITIC Securities is the incumbent heavyweight — China's largest brokerage by balance sheet for most of the past decade, with domestic distribution CICC has never matched and a consistent number-one position in A-share underwriting. When a large mainland issuer wants a domestic bookrunner with maximum retail placement power, CITIC is the default. CICC competes on cross-border sophistication, not on domestic reach.

国泰海通证券 Guotai Haitong Securities is the newcomer that changed the arithmetic. Formed by the merger of 国泰君安 Guotai Junan and 海通证券 Haitong Securities, completed in March 2025 with the combined entity debuting on the Shanghai exchange the following month, it instantly became China's largest securities firm by assets, overtaking CITIC.¹⁵ This matters for two reasons. First, it demonstrated that a mega-merger of two listed Chinese brokers could actually be executed — regulatory approvals, share swap, integration, relisting — in roughly a year. Second, it created the valuation and structural template against which CICC's own 2025 deal would be judged. Guotai Haitong was the proof of concept; CICC is running the same play.

Behind them sits the next tier: 华泰证券 Huatai Securities, 中国银河证券 China Galaxy Securities, 广发证券 GF Securities, 光大证券 Everbright Securities, 申万宏源 Shenwan Hongyuan. Several of these — Galaxy, Shenwan Hongyuan, Everbright, and CITIC itself, plus 长城国瑞证券 Great Wall Guorui — sit inside the same Central Huijin portfolio as CICC. That fact deserves a moment. CICC is not a state-backed champion competing against private rivals. It is one of several state-backed brokers competing against each other, all of them owned by the same shareholder, which is simultaneously deciding which of them should get bigger and which should be absorbed.

Then there is the door that has been slowly opening. UBS became the first foreign firm to take majority control of a mainland securities JV in 2018. Goldman Sachs reached full ownership of its China securities business in 2021, and Morgan Stanley — CICC's founding partner, returning as a competitor — took its stake to 94% in 2022. Citigroup has been reported as moving toward a wholly-owned China securities licence. None of these has yet built domestic scale that threatens CICC's league-table position. But the direction is one-way, and it introduces a competitor set CICC never had to face in its first thirty years: firms with global balance sheets, global research distribution, and no political constraint on what they pay their bankers.

Where CICC genuinely wins is the seam between onshore and offshore. Its Hong Kong arm has become disproportionately important to group profitability: in 2025, overseas business revenue grew 58% to make up nearly 30% of the group total, and CICC International alone earned HK$5.1 billion in net profit, up 78%.¹⁶¹² Against a group net profit of RMB 9.79 billion, an offshore subsidiary earning HK$5.1 billion is contributing something close to half the group's earnings from under a third of its revenue. That is a real and specific edge: when a Chinese company wants to list in Hong Kong, restructure a VIE, buy a European asset, or place a block offshore, CICC is one of a very small number of firms that can do the onshore relationship work and the offshore execution without handing the client to a foreign bank.

Where it is structurally behind is equally specific: retail distribution scale and balance sheet size. CITIC and Guotai Haitong can commit more capital and place more paper domestically. In a market where regulators have signalled they will grant larger, higher-quality firms more leverage headroom and easier access to new business lines, being sixth by assets is not a neutral position — it is a compounding disadvantage.

Which is exactly the gap the Dongxing and Cinda merger was designed to close. But before getting there, we need to understand why CICC entered 2025 as a firm that had spent three years cutting rather than growing — and that story begins with a social media post.

↑ Back to Top

VI. Boom, Bust, and the Viral Payslip: 2021–2024

2021 was the best year CICC ever had. Revenue reached RMB 30.1 billion, net profit crossed RMB 10 billion for the first time at RMB 10.8 billion, and return on equity hit 14.64%.[^33] Chinese IPO issuance was running hot, the Hong Kong pipeline was full, and average compensation across the firm hit roughly RMB 1.24 million per employee — the highest of any listed Chinese broker, ahead of CITIC's RMB 946,000 and Huatai's RMB 886,000.¹⁷

Then, on July 28, 2022, a woman posted on 小红书 Xiaohongshu — the lifestyle and shopping app, sometimes called Little Red Book — a photograph of her husband's income verification letter. The document showed an average monthly income of RMB 82,500. She paired it with couple photos taken at a Hangzhou shopping mall and a caption about being spoiled by her trader husband.¹⁷

The arithmetic is almost comic. Annualised, RMB 82,500 a month is roughly RMB 990,000 a year — below CICC's own firm-wide average for the prior year. This was not an outlier compensation package. It was a mid-level trader having an ordinary year at a Chinese investment bank.

It did not matter. The screenshots outran the original post, which vanished along with the account. The number trended across Weibo. And it landed in the middle of Xi Jinping's 共同富裕 common prosperity campaign, at a moment when China's economy was slowing, youth unemployment was rising, and public tolerance for displays of financial-sector wealth had gone to approximately zero. A single payslip became a national referendum on whether Chinese bankers were paid too much.

CICC's response was immediate and defensive: the employee was suspended pending investigation, for disclosing internal compensation information.¹⁷ The firm treated it as a confidentiality breach. The state treated it as a signal.

What followed over the next two years reshaped CICC's cost structure, its promotion system, and arguably its culture.

The Securities Association of China issued formal guidance directing securities firms to establish sound compensation systems — deferred payment, clawbacks, alignment between pay and long-term risk, explicit discouragement of short-term incentive excess.¹⁸ This was not a suggestion. In May 2023, Bloomberg reported CICC had cut senior bankers' 2022 pay by more than 40%.¹⁹ In April 2024 came another round: base pay for onshore investment bankers cut by up to 25%.²⁰ Days later, reporting surfaced that CICC was demoting senior bankers and stripping managing director titles under a new performance framework, with plans to shrink onshore investment banking headcount by as much as a third through 2026.²¹

The cumulative effect on the disclosed numbers is stark. Firm-wide average compensation fell every single year from the 2020 peak of roughly RMB 1.24 million, reaching RMB 642,600 in 2024 — a decline of 8.3% in that year alone and close to a halving over four years.²² Senior management took it harder still: total pre-tax compensation across the board and management committee fell from roughly RMB 168 million in 2020 and RMB 94 million in 2021 to a fraction of those levels.²³ Chinese financial media described the industry's mood shift as bankers going from 炫富 flaunting wealth to 哭穷 crying poverty.

And the timing could hardly have been worse, because the revenue side reversed simultaneously.

China implemented its registration-based IPO system market-wide in February 2023 — a reform that was supposed to move listing approval from bureaucratic discretion toward disclosure-based standards, and which the industry expected to expand issuance volumes. Instead, from late 2023, regulators deliberately throttled the pace of new listings to stop supply from pressuring an already-weak secondary market. CICC-underwritten mainland IPO proceeds fell 31% in 2023. Net profit dropped roughly 19–20% that year, taking ROE down to 6.43%. In 2024 it fell another 7.5% to RMB 5.69 billion on revenue of RMB 21.33 billion — a trough more than 45% below the 2021 peak.¹⁰²²

Now the credibility question, which is the reason this section exists.

Was the pay response proportionate crisis management or reactive damage control? The evidence points toward the latter, with a caveat. CICC did not get ahead of the compensation issue; it responded to a viral post, then responded again to regulatory guidance, then responded again to a third year of falling earnings. Each round was reported by external media before or alongside company disclosure. There is no evidence of a pre-articulated compensation philosophy that the firm then executed against — which is what proportionate crisis management looks like.

The caveat is that CICC arguably had no alternative. When the controlling shareholder is the state and the political environment demands compensation restraint, a listed subsidiary does not get to run its own pay policy. The more useful analytical framing is not "did management handle this well" but "what does this reveal about who sets CICC's cost structure" — and the answer is that it is not CICC.

The retention consequence showed up on schedule. By September 2025, Bloomberg reported CICC was constructing new senior job-title tiers specifically to give bankers advancement and status without raising headline pay — because raising pay back toward pre-2022 levels risked triggering another public controversy.²⁴ That is a company solving for a political constraint using titles as currency. It may work. It is not a durable substitute for compensation in a market where reopening foreign banks face no such constraint.

One more item from these years belongs in the record, because it complicates the prestige narrative. In October 2024, CICC disclosed that China's securities regulator had penalised it for failing to exercise due diligence as sponsor of chip design company S2C's 2021 STAR Market IPO application — a listing attempt in which the issuer had inflated profits substantially. CICC was fined RMB 6 million, had RMB 2 million of sponsorship income confiscated, and two sponsor representatives were personally fined RMB 1.5 million each.²⁵²⁶

The financial magnitude is trivial against a multi-billion-renminbi P&L. The signal is not. Sponsorship is the one product where a bank's entire value proposition is that its name on a prospectus means the numbers were checked. A firm that markets itself on institutional quality, then gets fined for not checking, has damaged the only asset that justifies its pricing premium.

Against that backdrop — cost base cut, brand dinged, earnings at a four-year low — the question of who was running the firm became more than a governance footnote.

↑ Back to Top

VII. Current Management: Who's Running CICC Now

For eighteen months, one person held both top jobs at CICC. That is not how the firm is meant to work, and the way it was resolved says a lot about what Central Huijin wants from this company.

Chairman 陈亮 Chen Liang took the chair on November 10, 2023, arriving as an executive director, chairman, and legal representative on the same day.²⁷ He is not a CICC lifer, and the distinction matters. Chen built his career in the domestic brokerage system, starting at 宏源证券 Hongyuan Securities in 1994 — where he rose from running the computer department to heading the brokerage business and eventually to deputy general manager. He carried through the merger that created 申万宏源 Shenwan Hongyuan, serving as a party committee member, group president, and executive director of its western subsidiary. From 2019 to 2023 he was at 中国银河证券 China Galaxy Securities as president, then vice chairman, then chairman — before being moved laterally to CICC.²⁷

Read that career path carefully. Hongyuan, Shenwan Hongyuan, Galaxy, CICC: four institutions, all within the state financial system, all ultimately in the Central Huijin orbit. Chen was not promoted at CICC. He was assigned to CICC. This is a textbook state-financial-system reshuffle, and the person it produces is someone whose institutional loyalty runs to the system rather than to any single firm within it.

That has an obvious upside for executing a state-directed consolidation and an obvious downside for anyone hoping for management continuity. It also produced an awkward episode: through 2024 and into 2025, persistent market speculation held that CICC would merge with Galaxy — Chen's immediate former employer. Both companies denied it. The actual deal, when it came, involved Dongxing and Cinda instead. The rumour was wrong about the counterparty and right about the direction, which is the most common way market speculation about state-directed deals goes.

In November 2025, Chen also took the chairmanship of CICC Wealth, the subsidiary that houses the retail franchise built out of the 2016 acquisition — a business with roughly RMB 193 billion in total assets and RMB 20.2 billion in net assets at mid-2025.²⁷ Concentrating group and wealth strategy under one person is defensible when wealth management is the largest segment. It also removes a layer of independent oversight from the fastest-growing part of the business, at a moment when it is about to absorb two retail-heavy brokerages.

President 王曙光 Wang Shuguang is the mirror image. Appointed president and nominated as an executive director on August 29, 2025, Wang was born in November 1974 and is the closest thing CICC has to a homegrown chief executive.²⁸ He took dual bachelor's degrees at 清华大学 Tsinghua University in physics and economics in 1996, added a master's in engineering in 1998, and joined CICC's investment banking department that same year. He never left. Managing director in January 2010, head of the growth-enterprise investment banking division, co-head of CICC Capital Management, head of the investment banking department from 2022, party committee member from December 2022, party committee vice secretary from August 2025.²⁸

Twenty-seven years, one firm, one core discipline. Wang also carries the finance portfolio. His appointment ended Chen Liang's eighteen-month period of holding both chairman and president roles, which began when former president 吴波 Wu Bo departed in April 2024.²⁸

Now the compensation read, which is where governance analysis gets uncomfortable for anyone applying Western frameworks.

Chen Liang's disclosed pre-tax pay for 2025 was RMB 1.428 million — down marginally from RMB 1.438 million in 2024, in a year when group net profit rose 72%.²³ That is roughly US$200,000 to run a firm with close to RMB 800 billion in assets. Total pre-tax compensation for the entire board and management committee was RMB 21.6 million in 2025, against RMB 94.4 million in 2021 and RMB 168.3 million in 2020.²³ Several subordinate executives out-earned both the chairman and the president.

There is no significant public disclosure of personal share ownership by either Chen or Wang. That is normal for executives of Chinese state-linked financial institutions and abnormal by the standards of global peers, where equity ownership is the primary claimed alignment mechanism between management and shareholders.

So what actually aligns CICC's leadership with minority shareholders? Not pay, which is politically capped and disconnected from results. Not equity, which is not disclosed. What remains is career progression within the state financial system — and that system's evaluation criteria include policy execution, stability, and scale, which overlap with shareholder returns but are not identical to them.

The pairing itself is legible as design: a state-appointed chairman whose background is running consolidations inside the system, and a career banker president whose background is running the franchise. One executes Beijing's agenda; one keeps the machine operating. Whether that division of labour is a stable governance model or a source of strategic drift is an open question — and 2026 is the year it gets tested, because the two men now have to merge three companies.

↑ Back to Top

VIII. The 2025 Merger: Building an "Aircraft Carrier" Investment Bank

The policy came first. The deals followed.

In 2024, China's State Council promulgated a document setting out the ambition to build several world-class investment banks, and the China Securities Regulatory Commission issued guidelines to restructure the securities industry toward that goal by 2035. At least six merger plans were rolled out in the following period. The regulator also signalled it would ease leverage and capital constraints for large, high-quality firms while applying differentiated, tighter criteria to smaller and foreign-invested ones.²⁹

That last clause is the one that made consolidation inevitable rather than optional. If regulatory privilege scales with size, then size stops being a strategic choice and becomes a survival requirement. Every mid-sized Chinese broker read that document and understood it was either going to acquire or be acquired.

Guotai Haitong went first and proved the mechanics worked. CICC went second, with a structure that was more complex — two targets, not one — and a rationale that was more explicit.

The announced terms: CICC absorbs Dongxing Securities and Cinda Securities through simultaneous share-swap mergers, with CICC surviving as the listed entity. Dongxing shareholders receive 0.4376 CICC A-shares for each Dongxing share; Cinda shareholders receive 0.5210. The implied swap prices were RMB 16.05 for Dongxing and RMB 19.11 for Cinda, against a CICC issue price of RMB 36.68, with CICC issuing over 3.1 billion new A-shares.³⁰ Total implied transaction value came to approximately RMB 114.3 billion — roughly RMB 52.2 billion for Dongxing and RMB 62.1 billion for Cinda.³¹

The scale arithmetic is what Beijing cares about. Combined total assets move to roughly RMB 1 trillion — around US$140–148 billion — lifting CICC from sixth-largest Chinese broker by assets to fourth.³⁰ The branch network expands by roughly 80%, to about 436 outlets, almost entirely through the retail-heavy acquired businesses.³¹ Both targets, like CICC, sit under Central Huijin.

Take the strategic logic on its own terms first, because it is coherent. CICC's identified structural weakness is retail distribution and balance sheet. Dongxing and Cinda are retail-branch businesses with capital. Bolting them on addresses the exact gap identified in the competitive analysis — more branches to distribute wealth products through, more capital to commit to underwriting and financing, and a scale ranking that qualifies for the regulatory privileges reserved for large firms. If you accept the premise that Chinese brokerage is consolidating toward a handful of survivors, being an acquirer is unambiguously better than being a target.

Now the sceptical read, because the coherence of the strategic logic does not settle the question of price.

The consideration is CICC equity, which means existing shareholders are paying for scale by diluting themselves. Independent analysis has flagged the transaction as dilutive to earnings per share and to return on equity for CICC in the near term.³¹ Mechanically, that is what you would expect: CICC's franchise earns higher margins than a regional retail brokerage's, so blending in two lower-ROE businesses drags the average down until synergies arrive — and integration synergies in brokerage consolidations arrive late, if at all, because the cost base is people and branches, and closing branches in China is politically constrained in ways it is not elsewhere.

The harder question — did CICC overpay — cannot be answered by asserting that state-directed deals are cheap. It requires comparing the implied swap valuations against Dongxing's and Cinda's standalone trading multiples before the announcement, and against the exchange ratio Guotai Junan struck with Haitong. Two structural features should temper any confidence in the outcome. First, all three parties share a controlling shareholder, so the negotiation was not adversarial in the way an arm's-length merger is. Second, in a share-swap deal, both the numerator and the denominator are set by the same set of prices, which means the fairness of the ratio depends entirely on whether the market prices of all three stocks were themselves fair at the reference date — a strong assumption in a market where merger speculation had been running for over a year.

The process moved faster than sceptics expected. CICC published the full merger circular through the Hong Kong exchange on December 17, 2025, setting out the terms, the conditions precedent, and the dissenting-shareholder mechanics.³² Six months later it put the transaction to a vote. In June 2026, the company's first extraordinary general meeting of the year, together with separate A-share and H-share class meetings held in Beijing, passed every resolution — including the authorisation to issue the new A-shares.³⁴

One procedural detail in that vote is worth pausing on, because it is the closest thing to a minority-protection mechanism in the whole transaction: Central Huijin abstained on the merger-related resolutions.³⁴ As the controlling shareholder of all three parties, it was a related party to itself on both sides, and abstention is the standard remedy. It is a real safeguard in form. In substance, a vote in which the 40% holder abstains is still a vote among shareholders who know precisely what the 40% holder wants, and who bought into a state-controlled entity with open eyes. Nobody should mistake the abstention for a contested outcome — but its absence would have been a genuine red flag, and it was not absent.

On June 16, 2026, the Shanghai Stock Exchange formally accepted CICC's application, confirming the materials were complete and compliant, and the transaction moved into regulatory review.³⁰

As of today, August 10, 2026, the mergers have shareholder approval and sit with the regulator. Completion remains conditional on those approvals and on other conditions being satisfied or waived, and the company's own language continues to state that there is no assurance the transaction will be completed.³⁰ The right posture for an investor is to treat closing as likely but not done — and to remember that in a share-swap deal with fixed exchange ratios agreed in late 2025, a long approval period transfers real economic risk to whichever side's shares have moved most in the interim.

Step back and the deal reads as the culmination of the ownership story. Central Huijin holds stakes in CICC, Galaxy, CITIC, Shenwan Hongyuan, Everbright, and Great Wall Guorui. It is not a passive anchor. It is running a portfolio restructuring, deciding which of its brokerages become platforms and which become inputs. CICC has been designated a platform — which is the good outcome, and also a reminder that the designation was not CICC's to make.

Which brings us to everyone else on the share register, and what they have been doing about it.

↑ Back to Top

IX. Who Actually Owns CICC Now

On July 25, 2025, two things happened on the same day. Tencent, through Tencent Mobility Limited, began selling CICC H-shares. And E Fund, one of China's largest domestic asset managers, bought 7.14 million of them at HK$21.44.⁵

That is a clean picture of a handoff.

Tencent's exit has been methodical rather than panicked. Having held 216 million H-shares at the end of 2024 — 11.36% of the H-share class — it had cut its position to 112 million shares by the end of 2025, representing 5.92% of H-shares and 2.33% of total share capital.⁵ Nearly eight years after the 2017 investment, the strategic partnership was being wound down in the open market.

Alibaba went further and faster. Having held 203 million H-shares through Des Voeux Investment Company and roughly 13.76 million A-shares through a Hangzhou vehicle at end-2024 — enough to rank ninth among A-share holders — it disappeared from CICC's top-ten shareholder lists entirely during 2025.⁵

Coming the other way: E Fund built to 199 million H-shares by year-end 2025, 10.46% of the H-share class and 4.13% of total shares. BlackRock added 42.9 million H-shares in June 2025 at HK$16.23 and finished the year holding 121 million H-shares, 6.34% of the class and 2.5% of total shares. Brunei Investment Authority appeared newly in the A-share top ten with roughly 10.32 million shares.⁵ Through all of it, Central Huijin's 40.11% did not move.

Two readings are available, and the honest answer is that both are partly right.

The benign reading is maturation. Platform-era strategic stakes in financial institutions were a product of a specific moment — 2017 to 2019, when Chinese internet companies were building financial ecosystems and buying options on distribution. That moment ended, decisively, with the regulatory reset of Chinese fintech from late 2020 onward. Tencent and Alibaba unwinding financial-sector equity stakes is consistent with a broader retreat from balance-sheet-heavy financial adjacencies, and replacing them with a large domestic fund manager and the world's biggest asset manager arguably improves the quality of the float. Index and long-only money is stickier and less strategically conflicted than a competitor-adjacent platform.

The less comfortable reading is what was lost. Tencent and Alibaba between them touch essentially every Chinese consumer with a smartphone. For a company whose largest business is now gathering retail and mass-affluent wealth assets, having both platforms on the register was at least a theoretical distribution and data option. That option is now gone — sold, not exercised — at precisely the point in CICC's history when retail wealth-gathering became the core of the business model. Neither partnership ever produced a disclosed, material distribution arrangement, which suggests the option was always more theoretical than real. But the sequencing is unflattering: the strategic shareholders left before the strategy that would have used them fully matured.

For a minority investor, the practical implication is narrower and clearer. With a 40% state anchor and the remaining float increasingly held by index-aware institutions, CICC's shareholder base has almost no capacity to influence outcomes. There is no plausible activist campaign, no realistic proxy contest, no scenario in which minority holders block a transaction the controlling shareholder wants. You are a passenger. The question is whether the driver is going somewhere you want to go.

↑ Back to Top

X. Strategic Position: Why CICC Wins From Here, and Why It May Not

Start with what actually happened in the numbers, because 2025 was a genuine inflection and the temptation to over- or under-read it is strong in both directions.

CICC's FY2025 revenue reached RMB 28.481 billion, up 33.5%. Net profit attributable to shareholders was RMB 9.791 billion, up 71.93%, with earnings per share of RMB 1.876. Return on equity recovered to 9.39%, up 3.88 percentage points. Total assets grew 16% to RMB 782.8 billion, shareholders' equity to RMB 122.1 billion, and the board declared a dividend of RMB 3.20 per 10 shares, roughly RMB 1.11 billion in total.⁵¹³ Momentum carried into 2026: first-quarter revenue of RMB 8.825 billion was up 54.3%, and first-quarter net profit of RMB 3.577 billion was up 75.2%.³³

And compensation began to normalise. Average pay per employee rose 24.4% in 2025 to RMB 799,300 — the first increase after four consecutive annual declines — while headcount fell by 432 to 14,218, of whom 13,156 are onshore and 1,062 international.²³ Even after that rebound, average pay remains about 35.5% below the 2020 level.

The bull case

The strongest argument for CICC is positional rather than operational. In an industry where the state has declared consolidation to be policy, CICC has been designated an acquirer. That is a materially different risk profile from being a target, and it is not something a competitor can replicate through better execution — you cannot out-work your way into being your regulator's chosen platform.

The second argument is that the cross-border franchise is genuinely differentiated and the evidence is external rather than self-reported. League-table leadership in Hong Kong IPO sponsorship, Chinese-issuer global IPOs, and China M&A is compiled by third parties. CATL's Hong Kong listing was the largest global IPO in two years and CICC was on it. That business requires bilingual technical capability, offshore regulatory fluency, and onshore relationships in combination — a genuinely narrow skill set that foreign banks lack on one side and domestic brokers lack on the other.

The third argument is that the wealth pivot did what it was designed to do. Investment banking revenue collapsed and recovered violently between 2021 and 2025; the company did not. Wealth management provided a revenue floor through the trough. The 2025 recovery, when it came, was broad — banking, equities, and wealth all up sharply — meaning CICC does not require an IPO boom alone to earn well.

The fourth is that the painful part of compensation reform is behind rather than ahead. Costs were reset over three years, headcount reduced, and the firm returned to growth with a lower fixed cost base. Operating leverage on the way up is real, and the gap between 2025 revenue growth of 33.5% and profit growth of 72% is the arithmetic evidence.

The bear case

The flagship business is hostage to a policy lever. Investment banking's share of revenue is small, and its variance is enormous, and the variance is not driven by market demand — it is driven by a regulator that has demonstrated it will slow IPO approvals to protect secondary market levels. That happened from late 2023, and there is no structural reason it cannot happen again in any year the CSI 300 has a bad quarter. This is not a cyclical risk that mean-reverts on a known schedule. It is a discretionary policy risk with no schedule at all.

The merger is dilutive before it is accretive, and integration risk is concentrated in exactly the places CICC has least experience. CICC has never integrated two companies at once. Its chairman arrived in late 2023 and has not run an entity of the combined size. The acquired businesses are retail branch networks with different cultures, different geographies, and different pay structures — and CICC's own compensation framework is politically constrained, which limits the usual levers for retaining acquired talent.

Compensation is now a structural liability rather than a one-off cost event. The firm cannot pay freely without political risk, at the precise moment Goldman Sachs, Morgan Stanley, UBS and potentially Citigroup are building onshore teams with no such constraint. Creating new title tiers to substitute for pay is a clever near-term workaround and a poor long-term answer. Watch senior banker departures.

Governance has visible cracks, and the S2C penalty is the specific evidence. A sponsor-diligence failure in the flagship product is not a rounding error in a business whose pricing depends on the credibility of its name on a cover page.

And the entire growth strategy is contingent on continued state sponsorship. This is a strength right up until the moment Beijing's priorities change — and Beijing's priorities have changed abruptly before, most recently for the entire Chinese internet sector.

Porter's Five Forces, briefly and specifically

Rivalry is intense and structurally intensifying. Consolidation reduces the number of competitors while making each survivor larger and better capitalised — competition gets harder, not easier, when your rivals are being merged into aircraft carriers too.

Buyer power is high and understated. A large Chinese SOE issuer can pick from CITIC, Guotai Haitong, Huatai, Galaxy, CICC and several foreign banks, all of which can execute a large domestic deal. Underwriting fee compression in China has been persistent for a decade, and the reason is that buyers have alternatives.

Supplier power — in a people business, talent is the supplier — swung decisively toward the firm during the pay-cut years, when bankers had nowhere better to go. It is swinging back as deal volume recovers and foreign banks rebuild onshore teams. CICC's inability to respond with pay is the sharpest asymmetry in its competitive position.

Substitution comes from two directions: foreign banks with full onshore licences taking cross-border mandates, and passive and platform-based distribution channels compressing the economics of wealth product distribution.

Barriers to entry remain high — licences, capital requirements, relationships, regulatory approval — but they are lower than they were in 2015, and every step of the opening has been one-directional.

A 7 Powers reading

Hamilton Helmer's framework asks which specific, durable mechanism prevents a competitor from replicating your returns. CICC's claim is unusual and worth stating precisely.

It is not scale economies — CICC is fourth by assets even after the merger, behind rivals it cannot outgrow organically. It is not network economies — a broker's clients do not become more valuable to each other as more join. It is not switching costs in the corporate franchise, where issuers routinely rotate banks deal by deal, though the wealth business is building modest switching costs through the buy-side advisory book. It is not cornered resource in the classic sense, though the cross-border execution team comes closest. Process power is a partial claim: three decades of running Chinese issuers through Western capital markets has produced institutional knowledge that is hard to hire.

The strongest claim is something adjacent to counter-positioning: CICC's advantage is being the state's chosen consolidator, a position that no competitor can replicate through investment or effort because it is granted rather than earned.

But test that claim against a fact established earlier: Central Huijin backs at least six brokerages. State favour that is granted to six firms is not exclusive, and a power that can be reassigned by a single shareholder decision is not durable in the sense Helmer means. What CICC actually has is a currently favourable position within a state portfolio — valuable, real, and revocable. Treating it as a moat requires assuming the allocator's preferences are permanent. Nothing in CICC's own history supports that assumption; the firm has been reassigned a new strategic purpose roughly every fifteen years.

The activist question nobody will ask

It is worth running the thought experiment even though it cannot happen, because it clarifies where the analytical pressure points are. If a concentrated Western fund built a position in CICC and wrote the letter, what would the letter say?

It would start with related-party capital allocation: two of the three defining transactions in the company's modern history — the 2016 purchase of a retail brokerage and the 2025 absorption of two more — were negotiated with the controlling shareholder on both sides, without a competitive process, and paid for in stock issued by the company the controlling shareholder controls. The letter would demand independent valuation and a genuine minority vote, not an abstention.

It would then attack disclosure quality: two reporting bases producing incompatible segment percentages, no disclosed management shareholdings, and no quantified synergy target or integration timetable published alongside a RMB 114 billion transaction.

It would question management accountability: a chairman whose disclosed pay fell by RMB 10,000 in a year when profit rose 72% is not being held accountable in either direction, which means the compensation system is transmitting no information about performance at all.

And it would note portfolio complexity: a securities firm that also runs a RMB 524 billion private equity book, a RMB 597 billion asset management business, a retail brokerage subsidiary, and an offshore investment bank, with no disclosed capital allocation framework governing where incremental capital goes among them.

None of these letters will be written, because the 40% holder makes them futile. But every point in the hypothetical letter is a real analytical weakness, and the fact that no one can force the company to address them is itself part of the investment case — for better and worse.

The KPIs worth tracking

Three, and only three, are worth the reader's ongoing attention.

First: client assets and the buy-side advisory balance within CICC Wealth. Total client account assets and the sub-balance in fee-based advisory mandates are the cleanest measure of whether the wealth pivot is producing recurring economics or just gathering low-margin assets. Rising total assets with a stagnant advisory balance would mean the business is still transactional.

Second: investment banking segment revenue and margin together. Revenue alone tells you about the IPO cycle; margin tells you whether CICC is holding pricing while competing against consolidated rivals for the same mandates. Fee compression will show up in the margin line before it shows up anywhere else.

Third: average compensation per employee alongside headcount. This is the single best proxy for the political constraint, the retention risk, and the operating leverage all at once. Pay rising in line with profit means the constraint has relaxed. Pay flat while profits rise means it has not — and the departures will follow.

↑ Back to Top

XI. Risks to Watch

Some risks in this story are generic to any financial institution. These five are specific to this one.

Policy and regulatory risk sits above everything else. In most markets, the pace of new issuance is set by supply and demand. In China, it is set by an approval queue that the regulator controls, and that regulator has demonstrated it will slow the queue to defend secondary market levels — as it did from late 2023, when a fully implemented registration-based system was expected to expand issuance and instead coincided with a sharp contraction in CICC's underwriting volumes. This is the highest-margin part of CICC's business and it can be throttled without warning, without market cause, and without any recourse for shareholders. Any model that treats banking revenue as mean-reverting on a market cycle is mis-specified.

Execution and integration risk is now the dominant near-term variable. Merging two retail brokerages into a firm whose culture was built around institutional advisory is hard under any conditions. Doing it simultaneously, under regulatory review, with a chairman under three years in post and a president under a year, raises the difficulty. The specific things to watch are branch rationalisation — which is politically sensitive because it means jobs — systems consolidation, and whether the combined firm's ROE recovers within the first two full years or drifts.

Talent risk is the constraint CICC cannot buy its way out of. The compensation framework is set by political tolerance, not by market clearing. Meanwhile UBS, Goldman Sachs and Morgan Stanley operate onshore with majority or full ownership and no such ceiling, and Citigroup has been reported as moving toward a wholly-owned licence. The title-tier workaround reported in 2025 is an acknowledgement of the problem, not a solution to it. Senior banker attrition in the cross-border franchise would attack the one advantage that is genuinely differentiated.

Cross-border and geopolitical risk reaches CICC indirectly, which makes it easy to underestimate. CICC itself carries no US ADR — it is an A+H structure — so it faces no direct US delisting exposure. But its clients do. A large portion of the deal flow behind CICC's cross-border franchise involves Chinese issuers with US listings, VIE structures, or US-exposed supply chains, and tightening US listing scrutiny and export controls suppress that flow at the source. CICC does not need to be sanctioned to be hurt; its clients need only become unfinanceable.

Technology disruption is the risk least discussed and most plausibly underpriced. Two of CICC's businesses are, at bottom, information-processing services sold at high margins because the processing was expensive. Sell-side equity research is one: analysts read filings, build models, and write notes. Mass-affluent wealth advice is the other: a relationship manager explains products to a client with a few hundred thousand renminbi. Both are exactly the shape of work that large language models have been compressing in cost across every market they have touched. The likely effect is not that CICC's research or advisory disappears; it is that the price of adequate research and adequate advice falls toward the cost of compute, and the premium narrows to the genuinely relationship-dependent top end. A firm whose wealth strategy depends on converting millions of retail accounts into fee-paying advisory mandates is selling into a market where the alternative is getting cheaper and better every year. CICC has disclosed no quantified financial impact from AI adoption, positive or negative, which is itself worth noting: the absence of disclosure is not evidence of absence of effect.

Governance and quality-control risk remains live. The S2C sponsorship penalty was small in money and large in meaning, and the relevant question is whether it was an isolated lapse or a symptom of diligence standards eroding under fee pressure and headcount cuts. Sponsor liability in China has been tightening steadily, with penalties extending to individual sponsor representatives. A second incident would move this from a data point to a pattern, and it would arrive at a time when the firm is absorbing two other underwriting franchises whose diligence cultures are unknown quantities.

To that list, one accounting and disclosure caveat that is not a risk so much as a discipline. CICC reports under both mainland and Hong Kong standards, and — as established when we worked through the segment mix — the two produce materially different revenue bases and therefore materially different segment percentages. Anyone comparing CICC's business mix across sources, or against peers, needs to confirm which basis is being used before drawing conclusions. This is not an accounting irregularity. It is a reconciliation trap that has produced genuinely contradictory published figures about the same company in the same year.

↑ Back to Top

XII. Durable Lessons & Epilogue

Thirty-one years on from a joint venture designed to teach Chinese bankers how to write a prospectus, what generalises from CICC's story?

The controlling shareholder's agenda is the strategy. CICC's stated strategy has changed three times, and each time it changed after its controlling shareholder decided what the firm should become. The 1995 mandate — import Western capital-markets capability — came from Beijing. The 2016 pivot into mass wealth management was executed by buying an asset from the controlling shareholder, in shares, on terms the controlling shareholder set. The 2025 consolidation is Central Huijin restructuring its own portfolio using CICC as the vehicle. In every case, reading the owner would have told you more about the company's next decade than reading the company's own strategy deck. For any investor in a state-controlled enterprise anywhere, that is the transferable lesson: the disclosure that matters most is often not in the annual report.

Reputational and political risk is a modelable line item, not a soft factor. A single post on a lifestyle app, showing a payslip that was below the firm's own average, set in motion a regulatory response, three consecutive years of compensation cuts, a demotion cycle, headcount reductions, and a retention problem that CICC was still engineering workarounds for three years later. No competitor could have inflicted that. In an environment where political tolerance for financial-sector compensation is itself a variable, the cost base of a Chinese financial institution is only partly under management's control — and analysts who model it as a normal operating expense will be wrong in both directions.

Diversification bought optionality, not immunity. The wealth management build-out kept revenue from collapsing in the trough, which is exactly what it was supposed to do. It did not prevent net profit from falling more than 45% from peak, because the segments that fell hardest were the highest-margin ones. Hedges of this kind reduce the depth of a drawdown; they do not eliminate the cyclicality of a capital-markets business, and they should not be sold as though they do.

So where does this leave things, as of August 2026?

CICC is a company in the middle of a transformation, not on the other side of one. The recovery is real: revenue up a third and profit up seventy-odd percent in 2025, with first-quarter 2026 growing faster still, and pay rising again for the first time in five years. The recovery is also incomplete: return on equity at 9.39% remains well below the 14.64% of the 2021 peak, and net profit is still short of the RMB 10.8 billion the firm earned that year. Five years on, CICC is not yet back to where it was.

The merger has cleared its shareholders and sits with the regulator, which means the hardest part — actually combining three firms, three cultures, and 436 branches into one operating company — has not started. The bull case rests on a position within a state consolidation agenda that CICC did not choose and cannot control. The bear case rests on the observation that everything valuable about that position was granted by a shareholder who has granted similar positions to five other firms.

What is verifiable is the franchise: number one in Hong Kong IPO sponsorship, number one in Chinese-issuer global IPOs, number one in China M&A, and an offshore arm generating close to half of group earnings. What is unverifiable, and will remain so for at least two years, is whether bolting a trillion renminbi of retail-heavy balance sheet onto that franchise makes it stronger or merely bigger.

Beijing has decided those are the same thing. The next few years will show whether they are.

China International Capital Corporation Limited

Table of Contents

CICC: China's Investment Bank, Reinvented Three Times

I. Cold Open & Roadmap

II. Origins: Building China's First Joint-Venture Investment Bank (1995–2010)

III. Central Huijin Takes the Wheel & the Dual Listing Era (2010s–2020)

IV. The Pivot Nobody Expected: From Elite M&A Shop to Mass Wealth Manager (2016–today)

V. What Investment Banking Still Means for CICC — and the Competitive Landscape

VI. Boom, Bust, and the Viral Payslip: 2021–2024

VII. Current Management: Who's Running CICC Now

VIII. The 2025 Merger: Building an "Aircraft Carrier" Investment Bank

IX. Who Actually Owns CICC Now

X. Strategic Position: Why CICC Wins From Here, and Why It May Not

The bull case

The bear case

Porter's Five Forces, briefly and specifically

A 7 Powers reading

The activist question nobody will ask

The KPIs worth tracking

XI. Risks to Watch

XII. Durable Lessons & Epilogue

References