China Galaxy Securities: The State's Bet on a National Champion Broker

I. Introduction & Episode Roadmap

On the afternoon of February 26, 2025, a Reuters story landed on trading desks from Hong Kong to Singapore with the force of a policy announcement. China's sovereign investment arm, it said, citing five people familiar with the matter, was preparing to merge two of the brokerages it controlled — 中国国际金融股份有限公司 China International Capital Corporation, the country's most prestigious investment bank, and 中国银河证券 China Galaxy Securities, its sprawling retail distribution machine. The combined entity would have carried roughly $193 billion of assets and vaulted straight into the top three of Chinese securities firms.¹

The shares had already moved. Both stocks posted double-digit gains in Hong Kong before the companies could respond.¹ Then, that evening, the two firms filed near-identical statements to the Shanghai and Hong Kong exchanges. They had checked with their controlling shareholder, 中央汇金投资有限责任公司 Central Huijin Investment. Huijin had no plan to combine them. Galaxy added a further, striking line: it had received no information, written or verbal, from any government department, regulatory authority, or shareholder about any such transaction.²³

Consider what that sentence actually concedes. A listed company with more than 125,000 registered shareholders was telling the market that the single most consequential question about its future — whether it would continue to exist as an independent entity — was not a question its own board could answer.⁴ It could only relay what the owner had told it. And the owner, Central Huijin, is a wholly state-funded company whose shareholder rights are exercised by the State Council itself, and whose directors are appointed by the State Council.⁵

That is the tension this story turns on. China Galaxy Securities is, by conventional measures, a serious financial institution: roughly RMB 856 billion in total assets at the end of 2025, more than 19 million brokerage clients, 458 branch outlets across all 31 mainland provinces and municipalities, 14,259 employees, and a dual listing in Shanghai under 601881.SS and Hong Kong under 6881.HK.⁴⁵ In 2025 it earned record net profit. It has an AAA domestic credit rating, an ASEAN brokerage network, and a management team recruited from the most sophisticated investment bank in the country.⁵

It also has never, in its entire corporate existence, made an independent decision about its own strategic destiny.

The central question for anyone underwriting this equity is therefore not the usual one. It is not "can management execute?" — the evidence suggests they can execute reasonably well. It is a stranger question: what is a minority stake worth in a company whose merger partner, capital structure, and payout policy are set one level above the board, by an owner whose objective function is national financial policy rather than per-share value?

The road from here runs through seven stops. First, the origin: Galaxy was not founded, it was assembled — a State Council–approved restructuring vehicle built in 2005–2007 to absorb the wreckage of a predecessor. Second, the two IPOs, four years apart, on two exchanges, and what that dual structure does to capital allocation. Third, the one genuine piece of corporate M&A in Galaxy's history: the staged, six-year buyout of CIMB's Southeast Asian stockbroking arm. Fourth, the core engine — retail wealth management and margin lending — where materiality actually lives, and where the moat is thinner than the branch count suggests. Fifth, the trading book that makes earnings swing. Sixth, the leadership, and the remarkable revolving door between Galaxy and CICC that fed the merger speculation in the first place. And seventh, the consolidation supercycle Beijing set in motion in March 2024, which is now reshaping the industry around Galaxy at a pace Galaxy does not set.

The cold open is the whole thesis in miniature. A rumor moved the stock. A denial moved it back. Neither event had anything to do with what the company did that quarter.

↑ Back to Top

II. Origins: Built, Not Grown (2005–2007)

Most great business stories begin with a person. This one begins with a document number.

On a December day in 2005, the China Securities Regulatory Commission issued approval 证监机构字〔2005〕163号 — a formal reply consenting to the preparation plan for a new entity to be called China Galaxy Securities Co., Ltd.⁵ There was no founder. There was no garage, no first customer, no product insight. There was a regulatory file, five founding shareholders, and an initial registered capital of RMB 6.0 billion, of which a single holding company — 中国银河金融控股有限责任公司 China Galaxy Financial Holdings — subscribed 99.89%.⁵

To understand why the Chinese state was manufacturing a brokerage from scratch, you have to understand what the previous decade had done to the industry. China's securities firms in the late 1990s and early 2000s were, to put it charitably, structurally unsound. Many were regional operations that had grown up alongside local governments, funded themselves by misappropriating client margin deposits, and speculated on their own account with the proceeds. When the A-share market entered a prolonged bear phase from 2001, the model detonated. Dozens of brokerages failed. The predecessor entity here — the original 中国银河证券有限责任公司 China Galaxy Securities Co., Ltd., itself established in August 2000 out of the securities arms of several state financial institutions — was among those needing a rebuild.⁶

The state's solution was not a bailout in the Western sense. It was an architectural one. In August 2005, with State Council approval, the Ministry of Finance and Central Huijin jointly registered China Galaxy Financial Holdings — described by its own rating documentation as the first state-owned financial holding company in China.⁵ That holding company then became the vehicle through which the viable brokerage, investment banking, and related assets of the predecessor were lifted into a clean new joint-stock company, capitalized fresh, and pointed at the future. Two decades later the ownership chain is still visible in the filings: as of the end of 2025, Huijin held 69.07% of Galaxy Financial Holdings, the Ministry of Finance 29.32%, and the National Council for Social Security Fund 1.61%.⁵

This was policy engineering, and Galaxy was one instance of a template. Huijin's portfolio today reads like a map of the Chinese financial system: China Development Bank, ICBC, Agricultural Bank of China, Bank of China, China Construction Bank, China Everbright Group, China Export & Credit Insurance, China Re — and, critically for what comes later, three other securities firms: 申万宏源 Shenwan Hongyuan, 中信建投证券 China Securities Co., and CICC.⁵ Huijin does not own a broker. It owns a brokerage sector.

Why does a twenty-year-old restructuring matter to someone looking at this equity in 2026? Because the DNA never changed. A company that was created by administrative decision, capitalized by administrative decision, and staffed by administrative decision does not spontaneously become an independent capital allocator when it lists shares. Every strategic inflection since — the Hong Kong listing, the Shanghai listing, the ASEAN acquisition, the 2025 merger episode, the dividend policy — has run through the ownership structure rather than around it. The company's own 2026 governance action plan says so in language that leaves nothing to interpretation: it commits to "exploring and practicing a modern enterprise system with Chinese characteristics under Party committee leadership," with the Party committee performing the functions of "setting direction, managing the big picture, and ensuring implementation."⁷

That is not a criticism; it is a specification. It tells you which risks to price. A Western investor evaluating a mid-tier broker asks about underwriting share, cost of funds, and compensation discipline. Those questions matter here too. But they sit underneath a prior question — what does the owner want this entity to be? — and that question is answered in Beijing, on a schedule the market learns about through rumor.

Which is precisely what makes the next chapter interesting. Because between 2013 and 2017, this state-built entity did something genuinely unusual: it went and sold a third of itself to the public. Twice.

↑ Back to Top

III. Two IPOs, One Balance Sheet: Going Public in Hong Kong, Then Shanghai (2013, 2017)

In May 2013, the Chinese brokerage industry was not a place global investors were queuing to put money. The A-share market was grinding through the tail end of a multi-year slump. Commission rates were falling. And yet on May 22, 2013, China Galaxy Securities listed H shares on the Hong Kong Stock Exchange under the code 06881.HK, raising roughly HK$8.15 billion once the partial over-allotment was exercised the following month — around US$1.1 billion, one of the larger Hong Kong offerings of that year.⁵⁶

The choice of venue was not sentimental. It was arithmetic. Mainland IPO approvals had effectively frozen: the CSRC had suspended new A-share listings for much of 2012 and 2013 while it worked through a backlog and a review of issuer quality. A securities firm is a balance-sheet business — margin lending, proprietary trading, market making, and underwriting commitments all consume regulatory capital, and the amount of business a Chinese broker may write is governed by net-capital ratios. Growth without equity is not slow; it is impossible. Hong Kong was the only door open, so Galaxy walked through it. Registered capital rose from RMB 6.0 billion to RMB 7.54 billion.⁵

Then it kept going. In March 2015, in the frenzied run-up to that summer's A-share bubble, Galaxy placed another 2.0 billion H shares offshore, taking registered capital to RMB 9.54 billion.⁵ And in January 2017, when the domestic listing window finally reopened, it issued 600 million A shares and listed on the Shanghai Stock Exchange on January 23 under 601881.SH, lifting registered capital to RMB 10.14 billion.⁵⁶

Four years separated the two listings, and the gap tells you something durable about how Chinese financial issuers think. The H-share market is where you go when you need money and the domestic gate is shut. The A-share market is where you go when you want a valuation. Chinese brokers have historically traded at meaningfully higher multiples onshore than offshore, because mainland retail investors buy brokerage stocks as a leveraged bet on the index while international investors price them as cyclical, state-directed financials with modest returns on equity. Same company, same balance sheet, two prices, and the gap persists because the two share classes are not fungible.

That divergence has a real consequence for capital allocation, and it is not the one management usually emphasizes. When your two listings trade at different prices, buybacks become politically and technically awkward — which share class do you repurchase, and at whose expense? The path of least resistance is cash dividends, which treat every shareholder identically regardless of listing venue. Galaxy has taken exactly that path, and what it has and has not done with capital is examined in Section X.

One further capital-markets move deserves a mention because it cuts against the "disciplined issuer" story. In March 2022, Galaxy issued RMB 7.8 billion of convertible bonds, listed that May on the Shanghai exchange as 中银转债. By the end of 2023 the convertible had been fully redeemed, and cumulative conversion had created 797 million new shares.⁵ That is the mechanism by which registered capital reached RMB 10.93 billion, where it stood as of the end of March 2026.⁵ Put plainly: over the past decade Galaxy has been a net issuer of equity, not a repurchaser of it. Every one of those raises was defensible on its own terms — a capital-hungry business needs capital — but the cumulative effect is that per-share value has been diluted in service of balance-sheet growth, and the return on that growing equity base has been unremarkable.

The ownership picture at the end of March 2026 was concentrated and stable. Galaxy Financial Holdings held 47.43%, unpledged. Hong Kong Securities Clearing (Nominees), the custodian standing in for H-share beneficial owners, held 33.74%. Hong Kong Securities Clearing itself, representing A shares bought through the Shanghai–Hong Kong Stock Connect, held 0.94%. Two domestic brokerage-sector ETFs rounded out the top five. Together the top five accounted for 83.09% of the register.⁵

Read that structure carefully and the shape of the investment case emerges. Roughly a third of the company is owned by international investors through Hong Kong; slightly less than half is owned by the state; and the marginal domestic buyer is frequently a passive sector ETF that buys Galaxy because it is a brokerage, not because it is Galaxy. Free-float price discovery is thin, sentiment-driven, and highly correlated to the index. Which makes the question of who really decides — the subject of Section IX — the only question that can produce a genuinely company-specific outcome.

Before that, though, there is one chapter in Galaxy's history where the company genuinely acted like a corporate strategist rather than a policy instrument. It happened offshore.

↑ Back to Top

IV. The Buy-and-Build Playbook: CGS-CIMB and the Making of a Pan-Asian Platform (2017–2023)

In 2016, CIMB Group faced an unglamorous problem. The Malaysian banking group had spent the previous decade assembling a regional stockbroking franchise — buying pieces of what had been Dutch bank ABN AMRO's Asian equities business and stitching them into an operation that spanned most of ASEAN and reached into Hong Kong, Korea, India, the UK, and the US. It was a real network. It was also sub-scale in a business where research, technology, and compliance costs are largely fixed, and where the marginal dollar of profit comes from order flow CIMB did not have enough of.

The obvious partner was Chinese. And in June 2017, CIMB agreed to sell 50% of CIMB Securities International to China Galaxy Securities for approximately S$167 million, or about US$121 million at the time.⁸⁹ The transaction completed in January 2018, the businesses were rebranded CGS-CIMB, and the two firms became equal partners in a stockbroking platform operating across Singapore, Malaysia, Indonesia, Thailand, Hong Kong, South Korea, India, the United Kingdom, and the United States.¹⁰

What made the deal genuinely interesting was not the price. It was the structure. Embedded in the partnership agreement were put and call options allowing Galaxy's Hong Kong subsidiary to acquire CIMB's remaining stakes over time.¹¹ In other words, Galaxy did not buy a business. It bought an option on a business, with a real operating partner obliged to stay in the seat while the option matured.

It exercised in two further steps. In December 2021, Galaxy's Hong Kong holding company exercised the first call option, taking a further 24.99% of the Singapore entity and 25% of the Malaysian holding company.¹¹ Then, on December 29, 2023, CIMB completed the divestment of its residual 25.01% and 25% stakes for gross proceeds of approximately RM 780 million.¹¹¹²¹³ CIMB was fully out. The platform became wholly owned, and was renamed — first CGS International, and subsequently, in Galaxy's own regulatory disclosure, 中国银河国际证券 China Galaxy International Securities (Singapore) and China Galaxy International Securities Malaysia, the entities Galaxy's filings now group together as 银河海外 Galaxy Overseas.⁵

What did the whole thing cost? CIMB has disclosed that its total proceeds from selling 100% of its original stockbroking business since 2018 came to roughly RM 2.5 billion.¹¹ At prevailing exchange rates over the period, that is somewhere in the region of half a billion US dollars, spread over six years, for an integrated brokerage and investment-banking network across ten markets.

Now benchmark it. In 2012–2013, 中信证券 CITIC Securities bought CLSA from Crédit Agricole for US$1.25 billion, structured as an initial 19.9% stake for US$310.3 million followed by the remaining 80.1% for US$941.7 million, completed on July 31, 2013.¹⁴¹⁵ CITIC paid roughly two and a half times what Galaxy paid, for one franchise — a storied research house with deep institutional relationships, but a single brand rather than a multi-market retail and corporate network. On raw price per market entered, Galaxy did well.

But the more useful comparison is on structure, and here the story is more nuanced than the tidy "staged M&A is smarter" lesson suggests. CITIC also bought CLSA in two steps. The difference is what happened in between. CITIC's first tranche was a nineteen-month bridge to full control on pre-agreed terms. Galaxy's staging ran six years, with the seller operating the business alongside it, sharing the integration risk, and — crucially — retaining the option value of walking away. That is genuinely different risk transfer. Galaxy bought optionality and paid for information; CITIC bought an asset and paid for speed.

Did Galaxy pay up at the end? Almost certainly, at least a little. By late 2023, CGS-CIMB was a proven, integrated business, CIMB knew Galaxy's alternative was an incomplete network with a partner's name on the door, and the final quarter cost more in absolute terms than the first half had. That is not a scandal; it is the tuition fee for deferring commitment. The relevant question is whether the total was worth it, and for that you need the segment numbers.

They are sobering, and they are also improving. In 2025, Galaxy's international segment generated RMB 2.569 billion of revenue, up 18.23% year on year — roughly 9.1% of the company's principal-business revenue, having risen about 3.0 percentage points as a share over the prior year.¹⁶ Across 2023 to 2025 the international business compounded revenue at 13.17% a year.⁵ Discrete profit contribution from the ASEAN platform specifically was not disclosed separately.

So: after roughly half a billion dollars and eight years, the entire international franchise — Hong Kong plus ASEAN — is under a tenth of the top line. That is the honest sizing. It is optionality, not an engine.

The operational evidence underneath is genuinely better than the revenue share implies, though. In 2025, Galaxy Overseas ranked first by trading value in Singapore, second in Malaysia, fifth in Thailand, and sixth in Indonesia; it completed 71 equity and debt financing transactions totalling S$4.7 billion; and it was ranked first among ASEAN local brokers in Extel's 2025 Asia local-brokerage survey.⁴⁵ Separately, 银河国际控股 China Galaxy International Holdings — the Hong Kong arm — broke into the top ten by Hong Kong IPO sponsorship volume for the first time, ranked fourth among Chinese brokers by number of deals, and fifth by offshore bond issuance count, in a year when Hong Kong reclaimed the global IPO fundraising crown with HK$644.4 billion raised.⁴

The positioning has become explicit. In 2024, Galaxy and CGS International jointly convened a forum branding the platform as a financial gateway between China and Southeast Asia, and through 2025 the group co-hosted the inaugural Malaysia Economic Forum, the ASEAN Investment Forum, and the ASEAN Business Forum as the only Chinese financial institution involved.¹⁷⁴ That is state-adjacent capital deliberately positioning itself in the corridor along which Chinese manufacturing, energy, and capital are moving south.

Here is the balanced read. Galaxy structured a cross-border deal better than most Chinese financial acquirers have, held its nerve through a six-year integration, and now owns the leading local brokerage franchise in several ASEAN markets — a genuinely scarce asset that would be hard to replicate at any price today. It has also, so far, bought itself a segment that is one-eleventh of revenue and growing from a small base. The bet is real. It is not yet material. Anyone claiming it transforms the investment case is ahead of the evidence.

For materiality, you have to go back onshore, to the branch network.

↑ Back to Top

V. The Core Engine: Wealth Management and Retail Brokerage

Walk into one of Galaxy's 458 securities outlets — in a provincial capital, a prefecture-level city, an industrial district — and the business model is immediately legible.⁵ There is a screen wall of quotes. There are account-opening desks. There are investment advisers, 4,320 of them registered across the firm at the end of 2025, whose job is to convert a customer who came in to trade stocks into a customer who also holds funds, bonds, structured products, and a pension account.⁴ This physical footprint, spread across all 31 mainland provinces, is the single largest thing Galaxy owns that its competitors cannot cheaply copy.⁵

It is also, in 2026, the most contested asset in Chinese finance.

Wealth management is Galaxy's core by a wide margin. In 2025 the segment produced RMB 13.472 billion of revenue, up 28.36%, or about 48% of the company's total.¹⁶ It comprises four related activities: brokerage commissions on stock and fund trades; distribution fees on financial products; advisory services; and the credit business — margin financing, securities lending, and stock-pledge repurchase lending. The first three are fee businesses that consume little capital. The fourth is a lending book, and it is where the interesting risk sits.

The scale is genuine. Galaxy ended 2025 with more than 19.3 million clients, up 11.56%; RMB 251.948 billion of financial products held on the platform, up 19.3%; and more than 230,000 cumulative personal pension accounts, roughly double the prior year.⁴¹⁶ Those are real distribution numbers, and the growth in product holdings matters more than the growth in accounts: it means clients are leaving assets with Galaxy rather than just executing through it.

Now the harder part. Look at what those assets earn.

Galaxy's margin-financing balance stood at RMB 136.2 billion at the end of 2025, with an average maintenance ratio of 268% — meaning collateral worth roughly two and a half times the loan.⁴ Its rating agency, 联合资信 China Lianhe Credit Rating, put the year-on-year growth of the margin book at 42.72%.⁵ And margin interest income rose 5.00%.⁵

Read that again, because it is the most analytically important pair of numbers in this section. The loan book grew by more than forty percent. The interest it generated grew by five. Meanwhile stock-pledge interest income fell 12.55%, which the rating agency attributed directly to lower rates charged.⁵ Galaxy is lending far more money for materially less spread.

That is what competitive commoditization looks like in the securities industry. Margin financing is an undifferentiated product: the client posts collateral, the broker lends against it, and the only variables are the rate and the haircut. There is no brand, no switching cost, no proprietary technology that makes one firm's margin loan better than another's. When a bull market arrives and every broker wants to grow its credit book, the rate is the only lever, and it goes one way. Galaxy's 2025 credit business is therefore a textbook demonstration that scale in this segment does not equal pricing power.

The market context makes the point sharper rather than softer. 2025 was an exceptional year for Chinese equities. Total A-share turnover reached RMB 420.69 trillion, with average daily turnover of RMB 1.73 trillion — up 61.7% year on year — while the market-wide margin balance grew 36.3% to RMB 2.54 trillion and A-share total market capitalization crossed RMB 100 trillion for the first time.⁴ The Shanghai Composite hit a ten-year high.⁴ In the single best trading environment in a decade, Galaxy's margin book grew slightly faster than the market's and its interest income barely moved. If that is what the segment earns at the top of the cycle, the question of what it earns in the middle answers itself.

There is a second-order risk embedded here that deserves flagging. Galaxy's credit-business leverage ratio — credit-business balances measured against shareholders' equity — rose to 112.33% at the end of 2025, from 87.68% a year earlier and 85.05% in 2023.⁵ Lianhe described that level as high relative to the industry.⁵ To be clear about what this does and does not mean: the loans are heavily overcollateralized, credit-impaired margin assets were a trivial RMB 132 million with 67.42% already provisioned, and impaired stock-pledge exposures of RMB 12 million were fully reserved.⁵ The book is not deteriorating. But it is getting bigger relative to the equity underneath it, at the top of a bull market, in a business where the collateral is the same asset class whose decline would trigger the losses. That correlation — collateral value and default probability moving together — is the structural weakness of every margin book ever written, and Galaxy has more of one than it did two years ago.

Against whom is this contested? The Chinese brokerage industry is unusually crowded at the top. CITIC Securities has long been the revenue and investment-banking leader. 国泰海通证券 Guotai Haitong Securities, created in 2025 from the merger of 国泰君安 Guotai Junan and 海通证券 Haitong Securities, is now the largest domestic broker by total assets and equity.¹⁸ 华泰证券 Huatai Securities built the leading online and discount franchise, effectively pioneering commission compression in the mainland. 广发证券 GF Securities and 招商证券 China Merchants Securities occupy the same broad tier as Galaxy. The scale gap is not subtle: Guotai Haitong booked RMB 15.74 billion of net profit in the first half of 2025 alone — more in six months than Galaxy earned in the whole of that year.¹⁹ Galaxy is large. It is not dominant.

So where does Galaxy actually win? Two mechanisms hold up to scrutiny, and one does not.

The one that does not is "brand." Galaxy's annual disclosure is dense with industry awards — best brokerage, best wealth-management team, best fintech institution, and a long list of others.⁴ These are marketing artifacts, not evidence of pricing power, and the margin-spread compression above is the proof: a brand with genuine pull would not have to fund forty percent balance growth with rate concessions.

The two that do hold up are distribution reach and sponsor credibility. Distribution is straightforward — 458 outlets across 31 provinces is expensive to build and expensive to run, and it reaches a demographic of mainland retail savers who still want a physical branch and a named adviser. That customer is stickier than an app user, though the cohort is aging and the economics per branch are structurally worse than they were a decade ago. Sponsor credibility is subtler but arguably more valuable: Lianhe's 2026 tracking report explicitly cites Huijin's and Galaxy Financial Holdings' capacity to support the company "in funding and business opportunities" as a rating strength, and reaffirmed the AAA long-term issuer rating with a stable outlook on May 29, 2026.⁵ In a business where the cost and reliability of wholesale funding directly determines how much you can earn on a lending book, being an entity the market believes the state stands behind is a real, quantifiable cost advantage — not a moat over peers who share the same backing, but a genuine advantage over the long tail of privately controlled brokers.

Management's own framing of the wealth-management strategy is worth testing. Chairman 王晟 Wang Sheng and his team have pushed a transformation toward "light-capital" businesses — wealth management, investment banking, and asset management — and in 2025 they hit a milestone they have been advertising for three years: those businesses grew 26.89% and reached 51.47% of total revenue, crossing half for the first time.¹⁶ That is a specific, falsifiable target, publicly set and publicly met, which is more than many management teams manage. It is also worth noting what the metric conveniently obscures: the largest component of "light capital" is retail brokerage commission, which is itself pure market beta. Crossing 50% in a year when daily turnover rose 61.7% is less a structural shift than a cyclical one wearing a strategic label.

The honest conclusion on the core engine is this. Galaxy has scale, reach, and a funding advantage. It does not have pricing power, and 2025's numbers demonstrate that at the moment when pricing power should have been easiest to exercise. The segment will earn well when turnover is high and poorly when it is not, and the direction of turnover is not something any broker controls.

Which brings us to the part of the P&L that makes that cyclicality impossible to hide.

↑ Back to Top

VI. The Trading Book: Proprietary Trading and Institutional Business

Here is a sequence of four numbers, all from the same year, all from the same company. Galaxy's net profit attributable to shareholders in the first quarter of 2025 was RMB 3.02 billion. In the second quarter, RMB 3.47 billion. In the third quarter, RMB 4.48 billion. In the fourth quarter, RMB 1.55 billion.⁴

Nothing broke. No business was sold, no impairment was taken, no regulator intervened. The fourth quarter simply happened. Revenue fell from RMB 9.00 billion in the third quarter to RMB 5.55 billion in the fourth, and net profit fell by two thirds.⁴ That is the signature of an earnings stream driven by mark-to-market positions in securities markets, and it is the single most important thing to understand about how this company makes money.

The Investment & Trading segment produced RMB 7.026 billion of revenue in 2025, up 10.76% — the second-largest segment after wealth management, and roughly a quarter of the total.¹⁶⁵ But revenue share understates its influence on the bottom line, because this business carries very little variable cost. When the book performs, nearly all of it drops through. When it does not, the same is true in reverse.

What is actually in the book? At the end of 2025, Galaxy's proprietary securities holdings totalled RMB 390.86 billion. Bonds accounted for RMB 274.14 billion, or 70.14% — mostly interest-rate government paper and high-grade credit. Equities were RMB 41.77 billion, or 10.69%. Funds were RMB 38.60 billion, or 9.88%. The remainder sat in bank wealth-management products, in-house asset-management schemes, trusts, and restricted shares.⁵

The composition matters more than the headline. This is not a hedge fund taking directional equity risk with shareholder capital; it is overwhelmingly a fixed-income portfolio, and Galaxy's regulatory metrics reflect that. Proprietary equity securities and derivatives stood at 30.76% of net capital against a regulatory ceiling of 100%, while non-equity holdings were 289.21% against a ceiling of 500%.⁵ There is headroom on both. But the direction of travel is worth noting: equity exposure rose during 2025 while bond holdings shrank slightly, as the company shifted allocation toward stocks amid the rally — the classic pattern of adding risk into strength.⁵

Galaxy's own explanation of its equity strategy, delivered by Wang Sheng at the annual results briefing on April 17, 2026, was that high-dividend, low-volatility holdings act as the "ballast stone," while active equity investment targets artificial intelligence, semiconductors, new energy, and biopharmaceuticals for upside.²⁰ President 薛军 Xue Jun added the framing that as a centrally administered securities firm, Galaxy would substantially invest in emerging industries in service of national strategy.²⁰

That second sentence deserves a moment of attention rather than a nod. It is a candid statement that the proprietary book has an objective function beyond risk-adjusted return. Galaxy also disclosed that it participated in the People's Bank of China's swap facility for securities firms, funds, and insurers — a liquidity mechanism designed to channel medium- and long-term money into equities and stabilise the market — and described its own role as helping maintain orderly market operation.⁴ A proprietary book that is partly an instrument of market stabilisation is a proprietary book whose managers cannot always sell when they would like to. Minority shareholders own a claim on the outcome of that book without controlling its mandate.

Alongside the trading desk sits the institutional business, which had a spectacular 2025 in percentage terms and a modest one in absolute terms: revenue of RMB 1.74 billion, up 508.20% from a very small base.¹⁶ The underlying franchise is genuinely growing. Prime brokerage — the plumbing that hedge funds and private funds rent to execute, clear, custody, and report — ended the year with 8,040 client accounts, up 14.3%; RMB 481.4 billion of business scale, up 38.30%; and RMB 5.45 trillion of stock and fund trading volume routed through it, up 52.10%.⁴⁵ Custody and fund-services scale reached RMB 295.08 billion, up 40.7%.⁴ Galaxy also ranked fourth by number of market-making mandates on both the STAR Market and the Beijing Stock Exchange, and ninth in fund market making.⁵

Institutional services are structurally more attractive than retail margin lending, because they carry genuine switching costs — a fund that has integrated a broker's execution, custody, and reporting stack does not move casually. The problem is scale. At RMB 1.74 billion, this is 6% of revenue. It is the most promising thing in the P&L for anyone hoping Galaxy develops a durable competitive position, and it is nowhere near large enough to change the company's earnings character yet.

Futures brokerage sits adjacent as a smaller related book, run through wholly owned 银河期货 Galaxy Futures. Its asset-management scale reached RMB 21.47 billion at the end of 2025, up 92.8%, and the unit built out a niche in physical commodity delivery — its coking coal, coke, and iron ore delivery volumes ranked first in the industry — plus an "insurance plus futures" agricultural hedging programme that covered nine crops and paid out RMB 24.37 million to 69,900 farming households.⁴ The last of those is a policy service more than a profit centre, and Galaxy presents it as such.

There is one accounting matter here that materially affects how the trading book looks in the financial statements, and it is worth flagging because it changes the reported revenue line by a third. In July 2025, China's Ministry of Finance issued implementation guidance on the accounting treatment of standard warehouse receipt trading — the commodity transactions Galaxy Futures conducts. Under the new guidance, qualifying trades are recognised net, as investment income, rather than gross as revenue. Galaxy applied it and retrospectively restated 2024.⁴⁵ The effect: reported 2024 operating revenue fell from RMB 35.47 billion as originally published to RMB 22.76 billion restated.⁴ Ernst & Young Hua Ming audited 2023 through 2025 and issued unqualified opinions throughout.⁵

This is not an accounting scandal; it is a presentation change mandated by the standard-setter, applied consistently, and disclosed. But it is a trap for anyone comparing Galaxy's 2025 revenue against a five-year-old data series without noticing, and it is why the honest framing of the 2025 top line is "RMB 28.30 billion on a restated basis," not a headline growth number compared against a gross-basis past.

The deeper point stands regardless of presentation. Strip away the segment labels and Galaxy's earnings are a levered function of Chinese market turnover and asset prices. That is not a criticism of management — it is the nature of the industry, and every listed Chinese broker shares it. It does mean, however, that a year of record profit is weak evidence of improving competitive position, and that the parts of the business capable of producing non-cyclical earnings deserve disproportionate attention.

Those parts are small, and they deserve an honest accounting.

↑ Back to Top

VII. Investment Banking and Asset Management: Smaller, Strategically Important

In 2025, China's equity capital markets came back to life. A-share equity financing including convertibles reached RMB 941.59 billion, up 274.2%. IPO volume alone hit RMB 131.77 billion, up 95.6%. Refinancing rose 339.5%. Domestic M&A transaction value climbed 758.4% to RMB 685.0 billion.⁴ Regulators reopened the STAR Market and ChiNext to unprofitable technology companies and launched a "technology board" in the bond market.⁴

In that environment, China Galaxy Securities completed exactly one IPO.⁴

One. Plus seven refinancing deals, for total equity underwriting of RMB 8.697 billion, which ranked it twelfth in the industry.⁴ For a firm of Galaxy's balance-sheet size, in the best equity issuance year in half a decade, that is a thin result, and it defines the competitive reality of this segment more crisply than any commentary could. Investment banking generated RMB 585 million of revenue in 2025, up 22.18% — roughly 2% of the company's total.¹⁶

Why so weak? Because equity underwriting in China is a relationship-and-reputation business at the top, and the top is occupied. CITIC, CICC, and now Guotai Haitong have decades of accumulated issuer relationships, sector coverage teams, and — most importantly — the institutional distribution to guarantee a book. A broker known primarily for retail branches does not naturally win the mandate to take a semiconductor company public.

But look one column over and the picture inverts. In debt underwriting, Galaxy is genuinely competitive: RMB 682.58 billion underwritten in 2025, up 37.3%, ranking sixth in the industry.⁴ Within that, local government bonds — RMB 433.33 billion, ranked fifth — did most of the work, with financial bonds tenth, medium-term notes tenth, directional debt-financing instruments seventh, and short-term commercial paper sixth.⁴ Galaxy also ranked seventh by transaction value in listed-company share-issuance acquisitions and major asset restructurings, with three deals worth RMB 19.47 billion.⁴

That contrast is the single most revealing thing about where Galaxy's franchise actually sits. It is a top-six underwriter of government and financial-institution paper and a twelfth-ranked underwriter of corporate equity. Those are not the same business. Local government bond mandates are allocated substantially on the basis of institutional relationships, provincial presence, and standing with the issuer's supervising authorities — precisely the assets a state-owned broker with 458 outlets in 31 provinces accumulates by existing. Equity underwriting is won by convincing a founder that you can price and sell their company. Galaxy is very good at the first and unproven at the second.

This has a direct bearing on the "light-capital transformation" narrative. Debt underwriting, particularly of government paper, is lower-margin, more commoditized, and more directly tied to fiscal policy volumes than to any competitive advantage Galaxy has built. It is good business. It is not the high-margin advisory franchise the phrase "investment banking" tends to evoke.

Asset management tells a similar, smaller story. 银河金汇证券资产管理 Galaxy Jinhui, the wholly owned securities asset-management subsidiary, ended 2025 with RMB 90.702 billion of assets under management across 239 products, up 6.5%.⁴ The mix shifted in a healthy direction — collective schemes, where the manager exercises real discretion and earns higher fees, grew to RMB 68.27 billion across 136 products, while single-client mandates, which are essentially channel business, shrank to RMB 22.21 billion.⁴ Revenue was about RMB 510 million, near a five-year high but still under 2% of the group.¹⁶

Set that against the market. Chinese securities firms and their asset-management subsidiaries collectively ran RMB 5.80 trillion of private asset-management products at the end of 2025.⁴ Galaxy's RMB 90.7 billion is a rounding error inside its own industry — roughly 1.6% share in a business where scale drives everything from research budgets to distribution leverage. Meanwhile industry-wide public fund net assets managed by brokers fell 14.5% over the year, so the pool Galaxy is fishing in is not uniformly growing.⁴

The company also runs two private-equity and alternative-investment arms. 银河创新资本 Galaxy Innovation Capital ended 2025 managing 32 funds with RMB 36.479 billion of registered scale, adding nine newly filed funds worth RMB 6.255 billion during the year, with technology-sector investment up 619.3%.⁴ 银河源汇 Galaxy Yuanhui, the alternative-investment subsidiary, newly approved RMB 945 million of investments.⁴ These are policy-aligned "patient capital" vehicles supporting national technology priorities — the annual report says so directly — and they are too small to move consolidated earnings.⁴

So does asset management diversify Galaxy away from the trading cycle? Not yet, and arguably not at all. A management fee on RMB 90.7 billion of AUM is a stable revenue stream in principle, but AUM itself expands and contracts with markets, redemptions rise when performance falls, and the absolute contribution is small enough that it cannot offset a bad quarter in the proprietary book. The diversification argument requires several more years and considerably more scale before it becomes an argument at all.

That is the honest map of the franchise: one large cyclical retail engine, one large cyclical trading book, one competitive government-debt underwriting business, and a set of genuinely interesting but sub-scale institutional, international, and asset-management operations. The people running it inherited that map and have been trying to redraw it. Who are they?

↑ Back to Top

VIII. Current Management: The CICC Bloodline

On a day in October 2023, two Chinese brokerages announced leadership changes that, read side by side, looked less like two appointments than one trade.²¹

陈亮 Chen Liang, chairman of China Galaxy Securities, was appointed chairman of CICC. And 王晟 Wang Sheng, Galaxy's president — who had spent the previous two decades at CICC — moved up to act as Galaxy's chairman.²¹ The two firms had, in effect, swapped the top of their org charts.

Wang Sheng's biography explains why the market read so much into it. Born in 1977, he took a bachelor's degree in management information systems from Tsinghua University's School of Economics and Management in 2000 and a master's in management science and engineering in 2002.²² He joined CICC that same year and stayed for two decades, working on restructurings across telecoms, financials, energy, and pharmaceuticals. From 2010 he held senior roles in CICC's investment banking department, becoming its executive head in June 2016, an assistant president of the firm in March 2018, and in March 2020 a member of CICC's management committee and head of investment banking.²² He left for Galaxy in July 2022 as president and deputy director of the executive committee, became an executive director and vice chairman that August, took the Party secretary role in October 2023, and formally became chairman on October 26, 2023.²²

This matters beyond the resume. CICC is the most Western-flavoured institution in Chinese finance — founded as a joint venture with Morgan Stanley, staffed by bankers who think in league tables and fee pools, culturally the opposite of a branch-network retail broker. Installing a career CICC banker at the top of Galaxy was a deliberate signal about the direction of travel: away from being a distribution utility, toward being a corporate-finance institution. The "light-capital transformation" and the stated ambition to build "an investment bank trusted by entrepreneurs" are Wang Sheng's language.⁴

The results, tested against that ambition, are mixed. Under his chairmanship the light-capital revenue share crossed 50% and the international franchise reached top-ten status in Hong Kong IPO sponsorship. The A-share equity underwriting business, which is the most direct measure of whether a CICC banker can transplant a CICC franchise, remained twelfth.

Alongside him, 薛军 Xue Jun became president in November 2023.²³ Born in 1970, he studied economics at the Central Institute of Finance and Banking and took an MBA at Peking University's Guanghua School of Management. He began his career in 1997 at the CSRC's issuance supervision department — the regulator's approval desk — before senior roles at Qilu Securities, Shenyin Wanguo, and Shenwan Hongyuan, then joined Galaxy as a Party committee member, executive committee member, vice president, and chief financial officer.²³ A regulator-turned-operator serving as president and finance chief alongside a dealmaker chairman is a coherent pairing for a firm whose principal external constraint is regulatory.

The bench has been rebuilt aggressively since. 郭晨 Guo Chen joined from 中国投资有限责任公司 China Investment Corporation in December 2025 as vice president and chief risk officer. 孙青 Sun Qing joined from CICC in March 2026 as vice president and executive committee member, bringing asset-management leadership experience. 曲燕萍 Qu Yanping was elected an executive director in February 2026. 张瑞兵 Zhang Ruibing, a business director and executive committee member, took on the compliance director role concurrently on April 13, 2026. And 梁士鹏 Liang Shipeng departed Galaxy for CICC in early 2026.²⁴

Count the traffic. In under three years, Galaxy's chairman went to CICC, a CICC banker became Galaxy's chairman, a CICC executive became Galaxy's vice president, and another Galaxy executive went to CICC. When Reuters reported in February 2025 that Huijin was preparing to merge the two firms, the market did not need a document to find the story plausible. It could read the personnel file. And personnel moves inside Huijin's portfolio are approved at the shareholder level — which is exactly why the market treated them as signal rather than coincidence.

Now the credibility test, and it is a genuinely interesting one.

Galaxy and CICC both denied the merger in February 2025, and both specified that they had confirmed with Central Huijin.²³ Nine months later, on November 19, 2025, CICC announced a merger — with 东兴证券 Dongxing Securities and 信达证券 Cinda Securities, not with Galaxy.²⁵²⁶ The three signed a binding cooperation agreement, and on December 17, 2025 executed the definitive merger agreement setting exchange terms.²⁷

So: was the denial honest? On the evidence available, yes. CICC did not merge with Galaxy; it merged with two other firms. The statement described the controlling shareholder's plan accurately as of the date it was made, and the subsequent facts are consistent with it. Galaxy told the truth.

But notice what "telling the truth" bought minority shareholders. It bought nine months of accurate information about a decision they still had no influence over, followed by an announcement that reordered their industry without involving them. This is not a governance failure in the ordinary sense — Galaxy disclosed promptly and correctly. It is a structural feature: the company can only ever report the owner's current intention, and intentions change. The right lesson is not that Galaxy's management is untrustworthy. It is that its statements about consolidation carry a much shorter shelf life than equivalent statements from an independently controlled company, because the underlying decision does not belong to the speaker.

That structural reality also shapes incentives, and here the contrast with a founder-led firm is stark. Galaxy carries essentially no stock-based compensation and no discretionary R&D-style spend of the kind that reveals a management team's private conviction about the future. Pay is set within state-enterprise frameworks, and the 2026 action plan speaks of "optimising compensation management and performance assessment mechanisms" without disclosing linkage to per-share outcomes.⁷ Notably, the CSRC's May 2024 corrective-action order against Galaxy specifically cited inadequate implementation of deferred compensation rules — a small window into how compensation governance was actually operating.²⁸

What management has done consistently is pay cash. The company conducted interim cash dividends for a second consecutive year in 2025, and adopted a formal market-value management system approved by the board on March 28, 2025.⁷ What it has not done is repurchase shares. The 2026 action plan, which runs to six sections covering business focus, national strategy, governance, investor returns, communication, and executive accountability, contains commitments to "enhance the stability, sustainability and predictability of cash dividends" — and no buyback commitment whatsoever.⁷

Pressed on this at the results briefing in April 2026, with the stock having fallen 32.53% from its August 2025 peak of RMB 19.15 to RMB 12.92 despite record earnings, Wang Sheng attributed the decline to sector-wide pressure and cyclical fluctuation, and pledged better operating quality, stable dividends, improved disclosure, and more investor communication.²⁰ That is a defensible answer and a notably passive one. A management team with genuine capital-allocation latitude, looking at its own shares down a third on record profits, has one obvious tool. Galaxy did not reach for it — not because management is unaware of it, but because the tool is not theirs to pick up.

Which raises the question of what the owner is actually doing with the sector.

↑ Back to Top

IX. The Consolidation Supercycle: "Aircraft-Carrier-Class" Brokerages and Where Galaxy Sits

On March 15, 2024, the CSRC published a document that has done more to determine Chinese brokerage share prices than any earnings report since: the Opinions on Strengthening the Supervision of Securities Companies and Public Funds and Accelerating the Construction of First-Class Investment Banks and Investment Institutions.²⁹

It set out a timetable in two steps. Within roughly five years — so by around 2029 — China would aim to have about ten high-quality leading institutions driving the industry's development. By 2035, it would have two to three investment banks and investment institutions with genuine international competitiveness and market leadership, ranking among global peers on strategy, professional capability, governance, compliance, talent, and culture.²⁹ Twenty-five specific measures across seven areas supported it.²⁹

Strip away the policy language and the message to a mid-tier broker was unambiguous: there is a defined number of seats, it is smaller than the current number of firms, and the state has told you when the music stops.

The market understood immediately, and the deals followed. The defining transaction was Guotai Junan and Haitong. Announced in September 2024, it moved through approvals and completion at a pace essentially without precedent in mainland listed M&A: the merged Guotai Haitong Securities began trading on the Shanghai Stock Exchange on April 11, 2025, roughly six months after announcement, having issued 5.99 billion A shares and 2.11 billion H shares to Haitong's shareholders through a share-exchange absorption.¹⁸¹⁹ The result was the largest mainland broker by total assets and equity, displacing CITIC.¹⁸ Revenue in its first half as a combined entity reached RMB 23.87 billion.¹⁹ Other combinations moved through the pipeline in the same window, and market commentary at the time expected the Guotai Haitong precedent to accelerate consolidation across a sector then valued at roughly US$1.6 trillion in assets.³⁰

Then the wave reached Huijin's own house. The CICC–Dongxing–Cinda transaction announced in November 2025 was structured as a share-swap absorption in which CICC survives and assumes the assets, liabilities, businesses, and personnel of the other two, which delist and deregister.²⁵²⁷ The terms set CICC's swap price at RMB 36.91 per share, Dongxing's at RMB 16.14, and Cinda's at RMB 19.15, implying exchange ratios of 0.4373 and 0.5188 CICC shares respectively, with CICC issuing 3.096 billion new A shares.²⁷ On the disclosed combined basis the enlarged entity carried RMB 1,009.58 billion of total assets, RMB 171.54 billion of net assets, RMB 27.39 billion of revenue, and RMB 9.52 billion of net profit — making it, by asset size, one of the largest brokers in the country.²⁷

Now hold those numbers up against Galaxy's. Galaxy ended 2025 with RMB 855.74 billion of total assets, RMB 147.78 billion of equity attributable to shareholders, RMB 28.30 billion of revenue, and RMB 12.52 billion of net profit.⁴ On revenue and net profit, Galaxy standing alone is larger than CICC will be after absorbing two firms. On assets and equity, it is modestly smaller.

That comparison reframes the entire consolidation question. The lazy version of the Galaxy story is that it is a mid-tier firm awaiting absorption. The numbers say something more ambiguous: it is a top-tier firm by earnings, sitting inside a portfolio where its sibling just executed a three-way combination it was not part of. Either Galaxy is already deemed large enough to be one of the ten, or it is being held for a different pairing, or the sequencing simply has not reached it. The disclosures do not distinguish between these, and speculation about which is exactly that.

What is documented is the pattern of rumor. On November 13, 2023, Galaxy and CICC both issued clarifications denying knowledge of any merger plan. Exactly one year later, on the afternoon of November 13, 2024, both stocks surged on renewed speculation — CICC up as much as 9.02%, Galaxy up as much as 6.44% — after reports circulated that a plan was near, with one version holding that CICC's wealth-management division would be transferred into Galaxy while CICC remained independent. Both firms again said they were unaware.³¹ Then came February 2025 and the Reuters report. Three episodes in sixteen months, three denials, and one eventual merger involving one of the two parties.

This is where the skeptical-investor lens matters, and it has to be applied in a form appropriate to the market. There is no activist campaign to run here. China's state-controlled brokers do not face proxy fights, and a 47.43% holder answering to the State Council is not going to be pressured by a letter. What substitutes for activism is the market's own discount — the persistent gap between what these companies earn and what investors will pay for the earnings, which is the price of unpriceable optionality about corporate structure.

A skeptical investor would push on three specific things. First, disclosure asymmetry: when the controlling shareholder deliberates, the company by construction cannot know, which means every "we are unaware" statement is simultaneously true and uninformative. Second, terms risk: in a share-swap absorption the exchange ratio is negotiated between entities under common control, and the reference prices are struck at a moment the parties choose — the CICC deal's arithmetic shows exactly how much value can be allocated by picking a swap price. Third, strategic latency: a management team that cannot rule out being merged has a structural reason to avoid irreversible long-horizon commitments, and there is a live question whether Galaxy's relatively cautious deployment of capital — heavy in government bonds, light in equity underwriting build-out — partly reflects that.

Against all of that, the bull framing deserves a fair hearing too. Consolidation reduces the number of competitors bidding for the same mandates. If Beijing genuinely intends ten leading institutions and Galaxy is one of them, then the firms disappearing are competitors, and the survivors inherit share in a less crowded market. Galaxy's balance sheet, distribution reach, and AAA funding profile are exactly the attributes that make an acquirer rather than a target. And a firm that emerged from the last state-directed brokerage consolidation as the beneficiary has, at minimum, precedent on its side.

The honest answer is that nobody outside a small room in Beijing knows, and the position sizing should reflect that rather than pretend otherwise. Meanwhile, the operating business keeps producing numbers, and those numbers tell their own story about what this equity actually earns.

↑ Back to Top

X. Financial Trajectory and Capital Returns

Trace Galaxy's profit line across six years and you get a shape that looks less like a growth company and more like a seismograph.

In 2020, the company earned RMB 7.244 billion of net profit attributable to shareholders on revenue of RMB 23.749 billion.³² In 2021, as the market ran hot, profit jumped 43.99% to RMB 10.43 billion.³² In 2022 it fell 26.2% to about RMB 7.8 billion as revenue slipped 6.5%.³³ In 2023 it was RMB 7.879 billion — essentially flat at the bottom, and lower than the level of three years earlier.⁴ Then 2024 brought RMB 10.031 billion, and 2025 brought RMB 12.520 billion, up 24.81% and the highest in the company's history.⁴

Six years, and the trough sat 24% below the peak that preceded it. That is the amplitude an investor is underwriting.

One caveat on the revenue line, because it has already tripped up published comparisons of this company: the 2020 and 2022 revenue figures were struck on the pre-restatement gross basis, while 2025's RMB 28.302 billion is on the restated net basis described earlier. On like-for-like restated numbers, 2023 revenue was RMB 18.939 billion, 2024 was RMB 22.762 billion, and 2025 was RMB 28.302 billion — a genuine 24.34% increase in the most recent year.⁴ Anyone reading a five-year revenue chart on this company without adjusting for the warehouse-receipt reclassification is reading a chart of an accounting policy, not a business.

Return on equity is where the analysis gets uncomfortable, and it is the cleanest single measure of whether the balance sheet is being used well. Galaxy's weighted average ROE was 7.52% in 2023, 8.30% in 2024, and 9.84% in 2025.⁴ Equity attributable to shareholders grew from RMB 130.47 billion to RMB 147.78 billion over the same span, while total assets grew from RMB 663.21 billion to RMB 855.74 billion.⁴

Sit with that for a moment. In the best Chinese equity market in a decade — index at a ten-year high, daily turnover up 61.7%, record profits — Galaxy earned just under 10% on its book value. Global investment banks in a good year earn low-to-mid teens; the strongest earn better. Galaxy's high-water mark for the cycle is roughly what a mediocre year looks like elsewhere. And the reason is visible in the balance sheet: total assets grew 29% over two years while equity grew 13%, meaning the firm is deploying an ever-larger balance sheet — much of it in government bonds and margin loans that earn thin spreads — to generate returns that a fee-based business would produce on a fraction of the capital.

This is the fundamental economic critique of the Chinese brokerage model, and Galaxy is a fair representative of it rather than an outlier. Capital-intensive businesses with regulated leverage limits, commoditized products, and pricing set by competition rather than by value delivered will produce single-digit-to-low-double-digit returns on equity across a cycle, and the market will value them accordingly. The "light-capital transformation" is precisely an attempt to escape that arithmetic. As of 2025 it has crossed a symbolic 50% threshold, but ROE has not yet escaped anything.

The funding side deserves its own note, because it is where a genuine vulnerability sits. Lianhe flagged that Galaxy's debt is large and skewed short: short-term debt was 71.36% of total debt at the end of 2025, down slightly from 73.87% a year earlier but still high, and the agency explicitly listed debt tenor and liquidity management as a rating concern.⁵ Against that, regulatory liquidity and capital metrics were comfortable — a risk coverage ratio of 250.17% against a 100% minimum and a 120% early-warning threshold, a capital leverage ratio of 14.92% against 8% and 9.6%, and a net stable funding ratio of 154.59%.⁵ The mismatch is structural rather than acute: a broker funding long-dated margin loans and bond inventory with short-dated wholesale paper is fine until the wholesale market has a bad week, which is precisely when its collateral values are also falling. The AAA rating and the Huijin lineage are what make this manageable, and that is not a small part of what the state affiliation is worth.

Capital returns are the one area where behaviour has genuinely improved, and it deserves credit without inflation. For 2025 Galaxy declared a final dividend of RMB 2.25 per 10 shares, totalling RMB 2.46 billion, on top of an interim dividend of RMB 1.25 per 10 shares totalling RMB 1.37 billion, for a full-year distribution of RMB 3.83 billion.⁴ Against RMB 12.52 billion of net profit, that is a payout ratio of roughly 31%. Basic earnings per share were RMB 1.03, up 27.16%.⁴

Two things are true about that. First, moving to twice-yearly dividends and sustaining it for a second consecutive year is a real, observable change, aligned with Beijing's broader push for state-owned listed companies to distribute more cash, and it is the kind of policy shift that survives management turnover because it originates above management.⁷ Second, a 31% payout on a business earning under 10% on equity, with no buyback, in a year when the shares fell by roughly a third, is a modest response. The retained 69% is being reinvested into a balance sheet that has not demonstrated it can earn an attractive incremental return.

An activist would put exactly that question — why retain capital at a sub-10% return rather than distribute it? — and would get, in this case, a structurally honest answer: because the controlling shareholder's objective includes the firm's capacity to serve national strategy, absorb market stress, and remain a viable consolidation participant, none of which is served by shrinking the balance sheet. That answer is coherent. It is simply an answer optimised for a different objective than per-share returns, and investors should price it as such rather than assume alignment.

Which sets up the competitive question directly: given all of this, what does Galaxy actually have that its rivals do not?

↑ Back to Top

XI. Competitive Position: Powers and Forces

Run China Galaxy Securities through Michael Porter's five forces and the picture that emerges is of an industry that is comfortable and a company that is ordinary within it.

Start with buyer power, which in this business means the retail and institutional clients on the other side of every trade. It is high, and structurally so. A mainland retail investor can open accounts at three brokers in an afternoon and route orders to whichever quotes the lowest commission. There is no data lock-in, no proprietary workflow, no meaningful integration cost. The evidence in Galaxy's own 2025 numbers is decisive: a margin book that grew more than forty percent while its interest income grew five, and stock-pledge income that fell outright on rate concessions.⁵ When buyers can walk, incumbents fund growth with price.

Threat of new entrants runs the other way and is genuinely low. Securities licences in China are issued by the CSRC, net-capital rules govern how much business a firm can write, and the regulator's declared policy direction is to reduce the number of significant players rather than expand it.²⁹ Nobody is starting a national brokerage in China in 2026. But note the analytical trap: this protects the incumbent class, not Galaxy specifically. Every one of its dozen large rivals sits behind the identical wall.

Substitutes are moderate and shifting. Bank wealth-management products, which held RMB 33.29 trillion at the end of 2025, and public mutual funds, at RMB 37.71 trillion, compete directly for the same household savings that Galaxy wants to capture through its branches.⁴ A saver choosing between a bank's product shelf and a broker's is exercising substitution, and China's banks have vastly more branches.

Supplier power is the least conventional force here, and the most important. In a securities business the critical inputs are capital, licences, and people — and for Galaxy, the supplier of the first two is the controlling shareholder, and the third has been recruited substantially from one specific competitor. This is a firm whose most important supplier is also its owner and its regulator's peer.

Rivalry is intense. A dozen large firms with near-identical product sets compete for the same institutional commissions, the same IPO mandates, and the same margin clients. Guotai Haitong has the largest balance sheet, CITIC the deepest corporate franchise, Huatai the strongest digital distribution, and CICC — post-merger — the most prestigious brand plus new scale. There is no segment in which Galaxy is the clear structural leader.

Now apply Hamilton Helmer's 7 Powers, which asks a stricter question: what allows a firm to sustain differential returns against determined competitors?

Scale economies. Partial, and mostly in the wrong place. The branch network delivers reach, but retail brokerage economics do not improve materially with more outlets once national coverage exists — the incremental branch serves an incremental customer at roughly the same cost. Where scale does help is the trading and market-making book, where fixed technology and research costs spread across a RMB 391 billion portfolio.⁵

Network economies. Weak in retail, real but small in institutional. Prime brokerage exhibits genuine network characteristics — more funds on the platform means better liquidity, better analytics, and better matching — and Galaxy's 8,040 PB accounts and RMB 5.45 trillion of routed volume are evidence it is building one.⁵ The ASEAN platform has a related property: a network of local licences and local execution across nine markets is worth more than the sum of the parts to a client wanting one relationship across the region. Both are early.

Counter-positioning. Absent. Galaxy does not do anything its competitors would find painful to copy. It is not a low-cost disruptor; if anything it carries the heavier physical cost base.

Switching costs. Low in the segment that produces half the revenue, moderate in the segment that produces six percent of it. That ratio is the compressed version of the entire competitive analysis.

Branding. Recognised, not premium. The industry-award list is long and the pricing evidence says it does not translate into willingness to pay.⁴

Cornered resource. This is the one genuine candidate, and it needs naming precisely. Galaxy's cornered resource is not a patent or a mine; it is its position inside the Huijin system. That translates into three concrete things: an AAA issuer rating explicitly supported by shareholder strength, which lowers funding costs on a business where funding cost is a direct input to margin; access to state-linked deal flow, visible in the fifth-place ranking in local government bond underwriting; and, less comfortably, a permanent seat at the table when the state decides which firms survive.⁵⁴ That resource is real, valuable, and — the crucial qualifier — shared with CICC, Shenwan Hongyuan, and China Securities, its Huijin siblings.⁵ It is a moat against the industry's long tail, not against its immediate peers.

Process power. Not demonstrated. Nothing in the disclosed operating record suggests Galaxy executes a repeatable process materially better than peers.

The evidence check the outline calls for cuts both ways, and it should be stated plainly. Where Galaxy claims leadership, some claims survive contact with league tables and some do not. Sixth in total bond underwriting, fifth in local government bonds, fourth in STAR Market and Beijing Stock Exchange market making, first by trading value in Singapore — these are verifiable rankings that support a claim of genuine competitive standing in specific niches.⁴⁵ Twelfth in A-share equity underwriting, roughly 1.6% share of the securities-firm asset-management pool, and one IPO completed in a boom year are equally verifiable and support the opposite conclusion in the businesses that carry the highest margins.⁴

The synthesis: Galaxy is a well-run, well-capitalised, state-backed firm with a strong position in commoditized businesses and a weak position in differentiated ones. Its clearest edge is a cost-of-capital and relationship advantage derived from its owner — which is precisely the advantage that becomes irrelevant the moment the comparison set is narrowed to firms with the same owner.

That framing sets up the two competing stories an investor has to choose between.

↑ Back to Top

XII. Bull vs. Bear Case

The bull case begins with a simple observation: Beijing has announced that it is building roughly ten national champions in this industry, and Galaxy earns more money than several firms that are being explicitly designated as building blocks of those champions.²⁹²⁷ If the destination is a consolidated sector with fewer, larger, better-capitalised players, a firm with RMB 856 billion of assets, RMB 148 billion of equity, national distribution, and an AAA rating is far more plausibly a survivor than a casualty.⁴⁵

The earnings leverage is real and it works in both directions, but the up-leg is powerful. Galaxy's 2025 net profit rose nearly 25% on a market where daily turnover rose 62%, and investment income specifically rose 37.45% to RMB 15.10 billion.²⁰ A broker with this asset base does not need to gain share to grow earnings substantially; it needs the index to cooperate. For an investor who believes Chinese equities are in a multi-year re-rating driven by household allocation shifting out of property and deposits, Galaxy is a direct instrument on that thesis with a real business underneath.

The capital-return trajectory supports the case at the margin. Twice-yearly dividends, sustained for a second year, plus a formal market-value management framework, represent a genuine improvement in shareholder orientation from a state issuer, and the policy impetus behind it is durable.⁷

The international platform is the most interesting non-consensus element. Owning the leading local brokerage in Singapore and the number-two in Malaysia, plus meaningful positions in Thailand and Indonesia, plus a top-ten Hong Kong IPO sponsorship business, is an asset that could not be assembled today for anything like what Galaxy paid.⁴⁵ As Chinese corporates, capital, and supply chains move into ASEAN, the firm sitting on both sides of that corridor with local licences has genuine optionality. It is 9.1% of revenue today; the argument is about what it is in a decade.¹⁶

And the tail scenario cuts positively as well as negatively. If Galaxy becomes an acquirer in the consolidation rather than a target — and its earnings scale makes that at least as plausible as the alternative — the market's structural discount for merger uncertainty could compress substantially.

The bear case starts by noting that almost everything in the bull case is a bet on the Chinese equity market rather than on China Galaxy Securities.

Earnings are structurally tied to trading volumes and asset prices the company does not influence. The quarterly progression through 2025 — three quarters of expansion followed by a collapse to RMB 1.55 billion of net profit in the fourth — demonstrates the mechanism in the space of twelve months, without any change in strategy, competitive position, or management.⁴ The 2023 trough, with profit below the 2020 level three years earlier, shows what a sustained downcycle does.⁴³²

Returns on that risk are mediocre. Sub-10% ROE at the peak of the best cycle in a decade, on a balance sheet growing faster than equity, is the arithmetic of a business that has to keep getting bigger to keep earning the same amount.⁴

The core segment faces continued fee compression with no visible defence. Forty-three percent balance growth producing five percent interest-income growth is not a one-year anomaly; it is what happens when a dozen well-capitalised firms compete for the same commoditized loan.⁵

The merger overhang cuts two ways, and both are costs. If Galaxy is absorbed, the exchange ratio is negotiated between entities under common control and set at reference prices chosen by the parties — the CICC transaction's swap arithmetic shows precisely how consequential that mechanic is for who captures what.²⁷ If Galaxy is not absorbed, it operates under indefinite strategic uncertainty, which is a real if unquantifiable tax on long-horizon commitment.

Governance constrains the response. No buybacks, a payout ratio near 31%, capital allocation routed through a shareholder answering to the State Council, and a proprietary book with an explicit market-stabilisation dimension — these are not defects, but they are constraints on the tools available when the shares fall a third on record earnings.⁷²⁰⁴

And a skeptical investor would add three specific stress points to the list. The credit-business leverage ratio climbing past 112% of equity at a cyclical high, described by the rating agency as high for the industry.⁵ The short-dated funding profile at above 70% of total debt.⁵ And the compliance record: the CSRC's corrective-action order of May 2024 cited insufficient diligence on projects, failure to identify non-market-based bond issuances, undisclosed self-subscription in bond offerings, employees evading tax through third-party bonus collection, inadequate deferred-compensation implementation, weak quality control, and poor compliance oversight of subsidiaries — followed by a warning letter from the Beijing regulatory bureau on December 26, 2024 over investor-suitability and internal-control weaknesses in the derivatives business.²⁸³⁴ None of these individually threatens the franchise. Collectively they describe an internal-control function that has been playing catch-up.

The net. The case for Galaxy winning from here rests on two pillars: state-directed scale advantage in a consolidating industry, and cyclical earnings leverage to recovering Chinese markets. Both are legitimate. Neither is company-specific execution. Strip out policy and beta and what remains is a competent operator with a strong niche in government-bond underwriting, an early-stage institutional franchise, a genuinely interesting ASEAN option, and a large commoditized retail business earning less spread each year.

The case against is not that Galaxy is badly run. On the available evidence it is run reasonably well by capable people who set a public target on business mix and met it. The case against is that it is a well-run, replaceable piece on a board where someone else moves the pieces.

↑ Back to Top

XIII. Risk Radar

The risks worth pricing here are not a generic list; they are a small number of specific mechanisms, several of which correlate with each other in exactly the wrong way.

Market-cycle risk is the master variable. Galaxy's proprietary book, its margin lending, its brokerage commissions, and its asset-management fees all move with the same underlying driver — Chinese asset prices and turnover. There is no natural hedge inside the business. The demonstration is in the 2025 quarterly sequence, where a single weak quarter cut net profit by roughly two thirds without any operational change, and in the three-year path from RMB 7.88 billion of profit in 2023 to RMB 12.52 billion in 2025.⁴ An investor holding this equity is holding a levered claim on the A-share index with a dividend attached.

Forced-consolidation risk is the distinctive risk of this particular name. The mechanism is not that a merger destroys value in aggregate; it is that in a state-directed share-swap between commonly controlled entities, the exchange ratio determines how value is divided, and minority shareholders on both sides are price-takers. The Guotai Junan–Haitong precedent established that these transactions can move from announcement to listed combined entity in about six months once decided.¹⁸¹⁹ A holder therefore faces a risk whose timing is unknowable, whose terms are set by the counterparty's controlling shareholder — which is also their own — and whose execution window is short enough that there is limited opportunity to react.

Regulatory and compliance risk is elevated sector-wide and non-trivial at the firm level. The CSRC reported that in 2025 it handled 701 cases, issued 661 penalty decisions covering 1,506 persons and entities — up 13.49% year on year — imposed RMB 15.474 billion of fines and confiscations, barred 142 individuals from the market, up 20.34%, and referred 172 suspected criminal cases involving more than 500 violators to public security authorities.³⁵ The regulator described the shift explicitly as moving from "settling cases with fines" to making violators bear substantive costs.³⁵ Galaxy's own enforcement history sits well below the systemically serious end of that spectrum, but it exists: beyond the corrective-action order and the derivatives warning letter already described, the Jiangsu bureau issued a warning letter on December 12, 2024 to an employee of a Nanjing branch for using inappropriate language that misled clients while marketing private fund products.³⁶ In a licensed business, the tail risk from compliance failure is not the fine — it is business-scope restriction, which directly caps revenue.

Geopolitical risk attaches specifically to the international segment, which is the growth story management most wants to tell. The ASEAN platform operates under nine separate licensing regimes, with clients and counterparties in markets navigating US–China strategic competition. Cross-border licensing, data-transfer rules, correspondent banking relationships, and sanctions compliance are all points at which a Chinese state-controlled broker's expansion could be constrained by decisions taken in Washington, Brussels, or the region's own capitals. The segment's operational momentum through 2025 is documented; its regulatory environment for the next decade is not.⁴

Correlated credit risk is the mechanism most often underestimated in brokerage balance sheets, and it deserves the plainest possible explanation. Galaxy lends money to clients so they can buy shares, holding those shares as collateral. If the market falls, two things happen at once: the collateral is worth less, and the borrower is more likely to default. The maintenance ratio of 268% and coverage ratio of 331% on stock pledges provide substantial cushion for ordinary volatility, and current impairments are negligible.⁴⁵ But this is not a diversifying exposure sitting alongside the trading book — it is the same risk factor, expressed twice. In a severe drawdown, proprietary losses, commission declines, and margin-book impairments arrive together.

Funding risk compounds the above. Galaxy funds a long-duration asset book substantially with short-term debt, at above 70% of total debt.⁵ A market stress that impairs collateral and depresses revenue is also the environment in which short-term wholesale funding becomes expensive or scarce. The AAA rating and state affiliation are the mitigant, and the rating agency's own downgrade triggers name the scenario precisely: material losses eroding capital, a significant rise in risk appetite pushing control metrics below regulatory minimums without a credible recapitalisation plan, or a major risk, compliance, or governance event.⁵

What is notably absent from this list is technology disruption of the kind reshaping Western financial services. Chinese brokerage licensing insulates incumbents from a Robinhood-style entrant, and Galaxy is deploying AI internally rather than defending against it — a large-model marketing assistant, a bond quotation robot selected for a CSRC fintech pilot, AI-driven operational systems in custody and fund services, and a digital-human service framework that handled 1.3 million customer interactions cumulatively.⁴ Whether that produces efficiency gains or merely keeps pace with peers doing the same thing is unresolved, but it is not an existential threat vector in this market.

↑ Back to Top

XIV. Playbook: Business and Investing Lessons

Three lessons travel beyond this company, and each is more interesting for its limits than for its headline.

Staged acquisition converts integration risk into a priced option — but you pay for the privilege. The CGS-CIMB structure is a genuinely instructive piece of deal engineering. Galaxy took 50% with call options rather than 100% outright, kept the seller in the seat as an operating partner for six years, watched the platform perform through a full cycle, and paid the largest per-percentage-point prices only at the end when the asset was proven.⁸¹¹ The value of that structure is informational: Galaxy learned things about the business, the markets, and its own integration capability that no due-diligence process would have surfaced, and it learned them while sharing the downside.

The limit is equally instructive. Optionality is never free. By the final tranche in December 2023, CIMB held all the negotiating leverage — a partner with an incomplete network is not a credible walk-away — and Galaxy's cost per point of ownership in the last quarter exceeded its cost in the first half.⁸¹¹ Note too that CITIC's CLSA purchase used a two-step structure as well; staging is not a Chinese-broker innovation.¹⁴ The transferable lesson is narrower and more useful than "stage your deals": when the asset's value depends on integration you have not yet proven you can execute, buy time and pay for it explicitly. When it does not, speed is worth more.

Policy risk in a state-owned company is not a discount factor — it is the primary variable. Most investment frameworks treat regulation and politics as an overlay applied to a fundamental analysis. In an SOE broker, that ordering is wrong. Who Galaxy merges with, on what exchange ratio, how much capital it retains, what its proprietary book is expected to do during market stress, and how much of its earnings it distributes are all determined at or above the shareholder level.⁷⁵ These are not risks to the thesis; in aggregate they are more consequential to the outcome than any segment's unit economics.

The practical implication is a different research process. The documents that matter most are not only earnings releases but policy documents — the CSRC's March 2024 opinions did more to determine this sector's valuations than any single company's results.²⁹ The signals that matter most include personnel flows between commonly controlled entities, because in a system where executives are appointed at the shareholder level, an org-chart swap is a strategic disclosure that arrives before the announcement does.

Official denials from state-controlled companies are true and short-lived, and both halves matter. The February 2025 episode is a near-perfect case study. Galaxy said the controlling shareholder had no plan to merge it with CICC. That was accurate — CICC merged with two other firms instead.²²⁵ The denial was not misleading and management was not evasive.

And yet the denial provided almost no durable information, because the speaker was relaying a third party's current intention rather than declaring its own. When a founder-controlled company denies a merger, the denial has weight because the person speaking is the person deciding. When a state-controlled company denies one, the statement's half-life is bounded by however long the owner's view holds. The three separate rumor-and-denial cycles across November 2023, November 2024, and February 2025 are what that looks like in practice.³¹²

The generalisable rule: assess a corporate statement by asking whether the speaker controls the outcome. Where they do not, treat the statement as an accurate snapshot with an unknown expiry, and size the position for the possibility that it expires without warning.

↑ Back to Top

XV. Epilogue & What to Watch

Three things are worth watching, and they are deliberately few, because most of what will be written about this company over the next two years will be noise about quarterly beta.

First, the light-capital revenue share. Galaxy management set a public target — shift the business mix toward wealth management, investment banking, and asset management — and reported crossing 51.47% of revenue in 2025 for the first time.¹⁶ The metric is management's own, which makes it the fairest test of whether they deliver what they promise. The question for 2026 and beyond is whether that share holds or improves in a year when market turnover is not rising 62%. If it does, the transformation is structural. If it retreats the moment volumes normalise, it was cyclical mix shift with a strategic label attached. The company's first-quarter 2026 results — revenue of RMB 7.355 billion, up 15.66%, net profit of RMB 3.32 billion, up 10.09%, weighted ROE of 2.53% — showed continued growth at a slower pace than 2025's, which is the beginning of that test rather than its conclusion.³⁷

Second, the spread on the credit book, not its size. Margin balance growth is the number every broker advertises and the number that means least. The relationship worth tracking is between balance growth and interest income growth. In 2025 those diverged violently — the book up 42.72%, the income up 5.00%.⁵ If that gap persists or widens, it confirms that competitive intensity has permanently repriced the most capital-intensive part of the franchise, and the returns on Galaxy's growing balance sheet will keep grinding lower regardless of how strong the market is. If it narrows, pricing discipline is returning to the industry.

Third, the international segment's profit contribution — not its revenue. Revenue reached 9.1% of the principal business in 2025 with a 13.17% three-year growth rate, and the operational rankings across ASEAN and Hong Kong are genuine.¹⁶⁵ But revenue growth in an expanding brokerage network can be bought with headcount. What would constitute evidence that the roughly half-billion-dollar ASEAN bet has paid for itself is a disclosed, growing profit contribution from the international segment, sustained through a weaker year in Southeast Asian markets. That disclosure has not yet been made separately, and the absence is itself informative.

Everything else — the merger question — will announce itself, and there is no research process that front-runs it. What an investor can do is decide in advance what an announcement would mean at various exchange ratios, and hold a position sized for an outcome they cannot forecast.

Which returns us to the February evening when two companies told the market, within hours of each other, that they had checked with their owner and the answer was no.

The remarkable thing about that episode, on reflection, is not the rumor or the denial. It is that both were entirely rational responses to a system working as designed. The Chinese state built China Galaxy Securities out of a restructuring in 2007, capitalised it, listed it twice, let it buy a Southeast Asian brokerage network, staffed it from the country's most prestigious investment bank, and has been rewarded with a firm that earns record profits and pays rising dividends. That is a competent outcome by any standard. The state also reserves the right to fold it into something else, at a price it will help determine, on a timetable it will not pre-announce.

In a sector Beijing is actively engineering into a handful of national champions, is Galaxy more likely to be one of the firms doing the consolidating, or one of the pieces consolidated? Its earnings say the former. Its ownership says the question is not its to answer. The next one to two years of state-directed brokerage M&A will settle it, and the settlement will arrive the way the last three did — as a report on a wire, on an afternoon nobody had circled.

References

1. China Brokers CICC and Galaxy Deny Report They Are Set to Merge — Bloomberg, 2025-02-26 ↩↩

2. Chinese investment firms CICC and China Galaxy firmly deny merger rumours — South China Morning Post, 2025-02-27 ↩↩↩↩

3. 中金公司、中国银河双双回应合并传言：经与控股股东确认不存在筹划传闻所称事项 — 新浪财经, 2025-02-27 ↩↩

4. 中国银河证券股份有限公司 2025 年年度报告摘要 — 上海证券交易所 / 巨潮资讯, 2026-03-31 ↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩

5. 中国银河证券股份有限公司 2026 年跟踪评级报告 — 联合资信评估股份有限公司 / 上海证券交易所, 2026-05-29 ↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩

6. 公司简介 — 中国银河证券股份有限公司 ↩↩↩

7. 中国银河证券股份有限公司关于 2026 年度"提质增效重回报"行动方案的公告（公告编号：2026-040）— 上海证券交易所, 2026-04-30 ↩↩↩↩↩↩↩↩

8. Malaysia's CIMB Group eyes Chinese market with stake sale of stockbroking arm to China Galaxy Securities — CNBC, 2017-06-08 ↩↩↩

9. CIMB and China Galaxy Formalise Strategic Partnership in Stockbroking — CIMB Newsroom, 2017 ↩

10. CIMB and China Galaxy Securities Commence Stockbroking Partnership — CIMB Newsroom, 2018 ↩

11. CIMB Completes the Divestment of Its Residual Stake in CGS-CIMB — CIMB Newsroom, 2023-12-29 ↩↩↩↩↩↩

12. CIMB completes divestment of residual stake in CGS-CIMB — The Edge Malaysia, 2023-12-29 ↩

13. Voluntary Announcement — Completion of the Acquisition of the Remaining Interests in CGS-CIMB — HKEXnews, 2023-12-29 ↩

14. Citic Securities to pay $1.25 billion for CLSA — FinanceAsia ↩↩

15. CITIC Securities and Crédit Agricole CIB jointly announce completion of the sale and purchase of CLSA — Crédit Agricole, 2013 ↩

16. 中国银河证券净利润创历史新高 轻资本业务首占半壁江山 — 21世纪经济报道 / 21经济网, 2026-03-31 ↩↩↩↩↩↩↩↩↩↩↩

17. China Galaxy Securities and CGS International Successfully Conclude Forum to Boost China-Southeast Asia Cooperation — PR Newswire ↩

18. China's Biggest Brokerage Merger Is Sealed as Guotai Haitong Debuts on Shanghai Bourse — Shanghai Stock Exchange / YICAI, 2025-04-14 ↩↩↩↩

19. Guotai Haitong Securities Co., Ltd. 2025 Annual Report — HKEXnews, 2026-04-08 ↩↩↩↩

20. 投资收益增37%！直击中国银河业绩会：股价一年内跌超32%，高管回应股价波动 — 腾讯新闻, 2026-04-18 ↩↩↩↩↩

21. 中金与银河证券换帅官宣！陈亮担任中金董事长、吴波代行总裁，总裁王晟代行银河董事长 — 财联社, 2023 ↩↩

22. 中国银河(601881) 公司资料 F10 — 同花顺金融服务网 ↩↩↩

23. 中国银河新任总裁人选已定，公司党委委员、执委会委员薛军将出任 — 界面新闻, 2023-11 ↩↩

24. 不止换人那么简单！银河证券密集调整高管，意在一流投行？— 新浪财经, 2026-04-17 ↩

25. China's Top Investment Bank CICC Plans Three-Way Merger — Bloomberg, 2025-11-19 ↩↩↩

26. CICC to Acquire Two Smaller Rivals to Create 1-Trillion-Yuan Brokerage — Caixin Global, 2025-11-20 ↩

27. China International Capital Corporation Limited — Merger Agreement Announcement, HKEXnews, 2025-12-17 ↩↩↩↩↩↩

28. 关于对中国银河证券股份有限公司采取责令改正措施的决定 — 中国证券监督管理委员会, 2024-05-31 ↩↩

29. 关于加强证券公司和公募基金监管加快推进建设一流投资银行和投资机构的意见（试行）— 中国证券监督管理委员会, 2024-03-15 ↩↩↩↩↩↩

30. China to spur mergers in US$1.6 trillion stockbroking sector after Guotai-Haitong deal — South China Morning Post ↩

31. 中金、银河股价异动再引合并传闻 双方回应：公司尚不知情 — 财联社, 2024-11-13 ↩↩

32. China Galaxy Securities Co., Ltd. 2021 Annual Report — HKEXnews, 2022-04-18 ↩↩↩

33. China Galaxy Securities Co., Ltd.'s Net Profit Dropped 26.2% in 2022 — Caixin Global, 2023-03-31 ↩

34. 因衍生品业务等违规收警示函，银河证券内控管理存在什么问题？— 新浪财经, 2025-01-06 ↩

35. 坚持依法从严 持续提升执法有效性和震慑力——2025年中国证监会执法情况综述 — 中国证券监督管理委员会 ↩↩

36. 中国银河证券频遭监管处罚，主营业务增长乏力，合规问题成隐患 — 新浪财经, 2024-12-22 ↩

37. 每周股票复盘：中国银河（601881）一季度净利33.2亿增10.09% — 东方财富网, 2026-05-02 ↩