China Minsheng Bank: The Private Bank That Isn't

I. Introduction & Episode Roadmap

In January 2026, 中国民生银行 China Minsheng Bank turned thirty. The anniversary was noted in the chairman's letter to shareholders with the language Chinese bank chairmen use for such things — gratitude to the Party and the State, to the economy, to the staff, to the customers — and a phrase that recurs like a mantra throughout the document: the journey toward becoming a "time-honoured bank."¹

It is a strange phrase for this particular institution, because Minsheng was never supposed to be time-honoured. It was supposed to be new. When it opened its doors in Beijing on January 12, 1996, it was the first national commercial bank in the People's Republic founded primarily by private companies rather than by the state.⁴ Its purpose was explicit and, for the time, mildly subversive: to lend money to the private entrepreneurs whom China's state banking system did not serve. Its founding capital of RMB1.38 billion came from businessmen, not ministries.¹⁴

Three decades on, Minsheng carries roughly RMB7.83 trillion of assets, a network of 2,392 offices and 58,982 employees, and a strategic self-description — "a bank for non-state-owned enterprises" — that has survived essentially unchanged.¹ What has not survived is the ownership. The largest shareholder of China's original private bank is now 大家人寿 Dajia Life Insurance, holding an aggregate 17.84% of the shares. Dajia Life is the vehicle created in 2019 to absorb the wreckage of 安邦保险集团 Anbang Insurance Group after regulators seized it; its ultimate controller is China Insurance Security Fund Co., Ltd., an arm of the state.¹²⁶ The bank designed so that no single owner could control it now has, at the top of its register, an entity that exists because a different owner lost control of something else entirely.

That irony is not a footnote to the Minsheng story. It is the story.

Here is the tension an investor confronts in August 2026. On the surface, FY2025 looked like an inflection. Operating income rose 4.92% to RMB139,677 million — the first annual increase after four consecutive years of decline. Net interest margin, the spread that is the beating heart of any bank's profitability, ticked up by a single basis point after years of relentless compression. The real-estate non-performing loan ratio fell sharply. Three credit-rating agencies moved in the bank's favour during the year.¹ Management called it a year of "revenue growth, structure optimisation and capability enhancement."¹

And yet net profit attributable to shareholders fell 5.37%, a second consecutive annual decline. Impairment charges against credit — the provisions a bank books when it expects loans to sour — jumped 18.64% to RMB53,950 million. Operating cash flow was negative for the second straight year. The dividend per share declined for the third year running.¹ Then, in the first quarter of 2026, the same pattern repeated with more force: revenue up 2.74%, net profit down 9.64%, credit impairment charges up roughly 28% year over year.³

A bank whose revenue is rising while its profit falls and its provisions accelerate is a bank still finding problems. Whether it is finding the last of them or the first of a new batch is, at this moment, genuinely unknown — and that uncertainty is why the shares changed hands at RMB3.49 on August 11, 2026, valuing the whole institution at roughly RMB148 billion, or about a quarter of the RMB12.90 book value per ordinary share the bank itself reported.²⁸¹

The road from there to here runs through five acts. First, the founding: how a bank without an owner came to exist, and why the choice to fragment ownership among 59 companies was the single most consequential decision in its history. Second, the boom: the 2000s expansion, the record-setting Hong Kong listing, and the shareholder wars that became a permanent feature of the boardroom. Third, the engine room: what this bank actually earns money on, which is less obvious than the marketing suggests. Fourth, the reckoning: the property crash that made Minsheng, for a period in early 2022, the worst-performing bank stock in the world, and the moment when a founding shareholder became a defendant in the bank's own lawsuits. And fifth, the present: a state-bank professional in the chairman's seat, a construction-bank risk man as president, a founder quietly rebuilding his stake, and an open question about whether any of it adds up to a durable recovery.

Start with the founding, because everything else is downstream of it.

---

↑ Back to Top

II. Founding a Bank Without an Owner (1996–2000)

Picture the problem as a Chinese private manufacturer would have experienced it in 1993. The economy is growing at double digits. Deng Xiaoping's southern tour has legitimised private enterprise. Orders are flooding in. And the local branch of the state bank will not lend, because its mandate, its incentive structure, and its political self-preservation all point in one direction: state-owned enterprises. The private firm has collateral, cash flow, and customers. What it does not have is a lender.

That gap was the business case for Minsheng, and the man who saw it most clearly was 经叔平 Jing Shuping — a figure who, by the time he founded a bank, was already old enough to have watched the entire arc of modern Chinese commerce. Jing had graduated from St. John's University in Shanghai in 1939, studying journalism, and had been in business before the People's Republic existed.⁴ He had survived the nationalisations, the Cultural Revolution, and the long cold decades when the words "private capital" were an accusation rather than a description. By the 1990s he chaired the All-China Federation of Industry and Commerce (全国工商联), the Party-sanctioned umbrella body for private business — a position that made him simultaneously an insider and an advocate for outsiders.

Jing's argument was economic rather than ideological, which was precisely why it worked. He did not propose that China should have private banks because private ownership is superior. He proposed that China needed genuinely commercial banks — institutions that priced risk and chased returns rather than fulfilling quotas — and that letting non-state money into the sector was the practical way to get them. "China needs a group of truly commercial banks," he said, "and it is necessary and feasible for non-state economic sectors to invest in China's financial and banking sector."⁴

Permission came, and with it the design problem that would define the next thirty years. A bank capitalised by private entrepreneurs raises an obvious anxiety for a regulator: what stops the owner from turning the bank into his own private treasury? The answer Jing and the authorities arrived at was fragmentation. Rather than a handful of large blockholders, Minsheng was launched with 59 founding shareholder companies drawn from the Federation's membership, contributing RMB1.38 billion of registered capital between them.¹⁴ No one could control it because, by construction, no one owned enough of it.

The roster read like a directory of 1990s Chinese private capitalism. 刘永好 Liu Yonghao, the Sichuan agribusiness magnate behind 新希望 New Hope, whose family had built a feed-and-livestock empire from a quail-breeding operation. 卢志强 Lu Zhiqiang, who would build 泛海 Oceanwide into a sprawling real-estate-and-finance conglomerate. 张宏伟 Zhang Hongwei of 东方集团 Orient Group. Later, 史玉柱 Shi Yuzhu of 巨人 Giant — a man whose earlier software and health-products venture had famously collapsed under the weight of an unfinished skyscraper before he rebuilt himself. These were not passive financial investors. They were operators with balance sheets of their own, and several of them had exactly the kind of capital-hungry businesses that Minsheng existed to finance.

Hold that thought. It is the fault line.

The elegance of the no-controlling-shareholder design is that it prevents capture by one owner. Its blind spot is that it does nothing to prevent capture by several. A board composed of nine or more shareholder-directors, each running a group that borrows from the bank, each dependent on the others' votes to keep a seat, has an obvious coordination problem — and an obvious equilibrium. Everyone approves everyone's credit. In 1996, with the Chinese economy compounding and property prices rising, that equilibrium looked like relationship banking. Twenty-five years later, it looked like something else.

The commercial start was strong enough to validate the thesis. By 2003, total assets had passed RMB200 billion, from a standing start of RMB1.38 billion in capital — growth that reflected less any particular brilliance than the sheer size of the underserved market Minsheng had walked into.⁴ Lending to private firms that the state banks ignored was not a clever niche; it was the largest financing gap in the world's fastest-growing large economy.

The capital markets came next. On December 19, 2000, Minsheng listed its A shares on the Shanghai Stock Exchange — an important test, because it asked public investors to fund a bank whose credit quality depended on a customer base that Chinese banking had no track record of underwriting. The listing succeeded, and in doing so established a template later followed by every joint-stock bank in the country.

But the A-share listing also introduced a new dynamic that the founders may not have fully anticipated. Once a bank's shares trade freely, ownership is no longer a matter of who was invited to the founding dinner. It becomes a matter of who is willing to buy in the market. A dispersed register, in a liquid market, is not a stable equilibrium. It is an invitation.

That invitation would be accepted, repeatedly and expensively, over the following two decades. First, though, came the growth years.

---

↑ Back to Top

III. Scaling the Private-Enterprise Playbook (2000s)

The 2000s were the years when Minsheng's founding thesis paid off spectacularly, and the mechanism was almost embarrassingly simple: go where the state banks would not.

China's four large state banks in that era operated with a mandate that was only partly commercial. Their loan books were populated by state-owned enterprises, infrastructure projects, and policy priorities. Their credit officers were not rewarded for figuring out whether a private textile manufacturer in Zhejiang or a components maker in Guangdong was a good risk — and were quite explicitly punished if such a loan went wrong. Into that vacuum walked a bank whose entire identity was built around the customer segment everyone else avoided.

The competitive advantage here deserves precision, because it was real but narrow. Minsheng did not have better technology, cheaper funding, or a superior brand. What it had was willingness — and the relationship networks of its founder-shareholders, who could introduce the bank to hundreds of private firms in their own supply chains and industry associations. That is a genuine distribution advantage, and for roughly a decade it functioned as one. It was also, structurally, a concentration machine. A loan book built through the personal networks of a handful of conglomerate owners is a loan book correlated to those conglomerates.

The bank leaned in. It built out small-business lending, community sub-branches, and a franchise in what Chinese banking calls 民营企业 private enterprises with more conviction than any peer. Even in its 2025 annual report, the strategic positioning is unchanged: to become "the preferred bank of NSOEs and the host bank of SMEs."¹³² Consistency of positioning over thirty years is, in fairness, unusual and worth crediting.

The Hong Kong Moment

The high-water mark of the era arrived in late 2009. Chinese banks were the trade of the moment — global investors wanted proxies for Chinese growth, and Minsheng offered the private-sector version. The H-share offering priced at HK$9.08, at the upper end of a HK$8.50 to HK$9.50 range, raising HK$30.15 billion, or about US$3.89 billion. It was the largest initial public offering in Hong Kong and in Asia that year, eclipsing Metallurgical Corporation of China's US$2.35 billion deal. The retail tranche was roughly 156 times oversubscribed; institutional demand ran to HK$33–34 billion excluding cornerstone investors. UBS was sole global coordinator, with BOC International, China International Capital Corporation, Macquarie and Hai Tong on the syndicate.⁶

The IPO mattered for three reasons beyond the headline. It gave Minsheng the capital base to grow aggressively through the 2010s, which it duly did. It subjected the bank to Hong Kong's disclosure regime, which is why an investor today can read a genuinely detailed annual report in English. And it put a floating, foreign-accessible share class alongside the A shares — which is why, three decades after founding, the single largest line on the shareholder register is HKSCC Nominees Limited at 18.93%, the custodian through which Hong Kong holders sit.² That figure is a plumbing artefact, not an owner. But it further diffused an already diffuse register.

The Battlefield for Capital

Which brings us to the recurring drama that Chinese financial press came to call 资本逐鹿的战场 — the battlefield where capital chases the deer. Board re-election cycles at Minsheng were not administrative events. They were campaigns.

The equilibrium held for roughly eighteen years, and then broke decisively in late 2014, when Anbang Insurance began buying. Over about two months, Anbang increased its position twelve separate times, driving its A-share holding to 22.51% and making it the single largest shareholder — shattering what Chinese commentators described as a two-decade "understanding" among the founding families.⁵

The response was exactly what the structure invited. In June 2016, Orient Group and Huaxia Life — then the ninth and tenth largest shareholders — signed a concert-party agreement that pushed their combined stake past Lu Zhiqiang's Oceanwide bloc by about 1.1 percentage points. Lu answered with cash: over four trading days in July 2016 he spent more than RMB7.5 billion buying Minsheng shares in the market.⁵ Read that again. A non-executive director spent the equivalent of roughly a billion dollars of his own group's money, in four days, to defend board influence at a bank he did not control and could not control. Chinese media nicknamed the whole affair "a war among the strongest men."⁵

For an investor, the analytical content of that episode is not the drama. It is the capital allocation. Every yuan that founder-shareholders spent buying each other's shares was a yuan not invested in their operating businesses — and several of those businesses were levered property developers. The governance structure did not merely fail to prevent risk concentration. It actively encouraged the bank's largest borrowers to take on more leverage in order to hold their board seats.

Two Departures

The decade also produced two leadership discontinuities that in hindsight look like warnings.

董文标 Dong Wenbiao chaired Minsheng from 2006 to 2014 and was, by reputation, the architect of its small-business franchise. He left in August 2014 — the bank's filing cited "other business arrangement" — to become chairman of China Minsheng Investment Corporation (中民投), a newly incorporated Shanghai vehicle with RMB50 billion of registered capital that Chinese media hailed as an "aircraft carrier" of private capital.⁸ The detail that should make an investor sit up: Dong assembled CMIG by persuading 59 private companies to become founding shareholders — the same number, the same template, the same logic as the bank he had just left.⁹ CMIG spent the next five years amassing roughly US$34 billion of debt, and by 2019 its offshore unit was announcing it could not repay a US$500 million bond, having already triggered cross-default clauses on US$800 million of dollar notes.⁹ The template, deployed a second time without a banking licence or a regulator, blew up in half a decade.

Then, in January 2015, the tail risk showed up inside the bank itself. 毛晓峰 Mao Xiaofeng had joined Minsheng in 2002 from the Communist Youth League, held a Harvard master's degree, and had risen through retail, micro-finance and corporate banking to become president in August 2014 — the youngest chief executive in Chinese banking. On January 25, 2015, the Central Commission for Discipline Inspection took him away for questioning. Contact was lost by January 27. On January 31 the bank announced his resignation for "personal reasons," with chairman 洪崎 Hong Qi stepping in as acting president. The case was linked to the corruption investigation into 令计划 Ling Jihua, a former vice chairman of China's top political advisory body.⁷

A bank can survive a bad quarter. What the Mao affair demonstrated was that Minsheng's political and shareholder entanglements were not merely governance untidiness — they were a live channel through which the institution could lose its chief executive overnight. That is a different category of risk, and it priced into the stock for years.

None of it, however, dented the growth. By the end of the 2010s Minsheng was one of China's largest banks by any measure. The question that would decide its fate was not how big it could get, but what it had actually been lending against.

---

↑ Back to Top

IV. What the Bank Actually Runs On: Segments & Economics

Every bank tells a story about itself, and every bank's segment disclosure quietly contradicts it. Minsheng's marketing centres on retail: 143 million retail customers at end-2025, up 6.46%; retail assets under management of RMB3.28 trillion, up 11.46%; private-banking clients up 20.24% to 74,671, holding RMB1.03 trillion between them.¹ Impressive numbers, prominently displayed.

Now read the segment table. In FY2025, corporate banking generated RMB68,470 million of operating income and RMB28,606 million of pre-tax profit. Retail banking generated RMB56,597 million of income — respectable, about 40.5% of the group total — but only RMB11,775 million of pre-tax profit. Corporate banking, in other words, produced roughly 89% of the group's entire pre-tax profit while contributing 49% of revenue.¹

That gap is the single most useful fact about how this bank works. Retail banking at Minsheng is revenue-rich and profit-poor: it carries the cost of 2,389 sub-branches, a large credit-card operation, and 58,982 staff, and it prices into the most competitive segment of Chinese finance. Corporate banking carries far lower distribution cost per yuan of revenue. So when management describes a "large corporations plus large retail" (大公司+大零售) dual engine, the honest translation is that one engine produces most of the thrust and the other is being rebuilt mid-flight.

The third segment, "others" — treasury, investment, subsidiaries including 民生金融租赁 Minsheng Financial Leasing and the fund-management and international arms — lost RMB8,122 million before tax in 2025, after losing RMB10,520 million in 2024. The drag narrowed, and its revenue rose from RMB9,974 million to RMB14,610 million, but it remains a persistent negative.¹ A conglomerate structure that consistently subtracts from group profit is exactly the kind of thing an activist investor would put on the table.

The Margin Squeeze, Explained Simply

The industry backdrop needs explaining in plain language, because it drives everything else.

A commercial bank is, at bottom, a spread business. It borrows short (deposits) and lends long (loans), and the difference between what it earns on assets and pays on liabilities — expressed as a percentage of interest-earning assets — is the net interest margin. Think of it as the gross margin on a manufacturer's product. When it compresses, everything downstream compresses with it.

Minsheng's net interest margin was 1.91% in 2021. By 2024 it had fallen to 1.39%, and in 2025 it recovered to 1.40%.¹ Losing half a percentage point of margin on a RMB7 trillion balance sheet is the difference between a comfortably profitable bank and a marginal one. The cause was systemic: China's central bank cut policy rates repeatedly to support a slowing economy, mortgage rates were repriced downward, and loan demand from creditworthy borrowers weakened. Every Chinese joint-stock bank suffered. Minsheng suffered from a lower starting point.

What management did about it in 2025 is the most concrete evidence of execution in the whole annual report, and it is worth crediting on its own terms. Rather than chasing loan yield — which in a weak economy means accepting worse credit — the bank attacked the cost side. Deposit costs fell 40 basis points to 1.74%, with personal deposits growing 7.08% while more expensive corporate deposits shrank 2.08%.¹ The president's letter framed this as "establish before dismantling," and noted that margin performance "outperformed the overall market conditions."¹ That claim survives scrutiny: a one-basis-point margin increase in 2025 was genuinely better than the sector trend. By contrast, 招商银行 China Merchants Bank — the sector's profitability benchmark — saw its margin fall 11 basis points to 1.87% in the same year.²⁷

The catch is scale. Total assets grew 0.23% in 2025. Total loans actually shrank 0.45%, with personal loans down 5.18% offsetting a 2.68% rise in corporate lending.¹ A bank that stabilises margin while its loan book contracts has bought time, not growth. Net interest income rose just 1.46%; the real revenue driver was non-interest income, up 14.86%, of which a meaningful portion came from investment gains on a RMB2.09 trillion bond portfolio — a source that is inherently market-dependent and not repeatable at will.¹

The Wealth Management Pivot

The more strategically interesting build-out is in fee income and wealth management, and here the evidence is better than the scepticism might suggest.

Retail agency sales of wealth-management products exceeded RMB1 trillion in 2025, up 9.10%. Sales of private-equity products more than tripled. Assets under custody passed RMB13 trillion for the first time, reaching RMB13.44 trillion. The family-trust platform aimed at entrepreneur clients was upgraded and grew rapidly, though the bank did not disclose its size.¹³¹

Why does this matter beyond the growth rates? Because it is the one part of the strategy that structurally reduces the risk that nearly broke the bank. Selling a wealth product earns a fee without putting the bank's own capital at risk. Custody earns a fee for safekeeping assets the bank does not own. Both are balance-sheet-light. A bank that shifts revenue mix from lending toward fees is a bank whose earnings become less dependent on credit cycles — precisely the correction Minsheng needs.

The honest caveat: net fee and commission income was 13.12% of operating income in 2025, down from 13.71% in 2024 and 16.65% in 2021.¹ The wealth franchise is growing, but from a base too small to change the group's character yet, and fee income overall has been shrinking as a share of revenue because of industry-wide fee cuts on insurance and fund distribution. Call it a real pivot, correctly directed, several years from mattering.

There is one more disclosure worth flagging for anyone assessing whether the bank has structurally de-risked. Loans to the top ten borrowers fell to 8.34% of net capital in 2025 from 10.17% a year earlier, and the single largest borrower to 1.42% from 2.42%.¹ Concentration is falling measurably. That is the kind of quiet operating fact that carries more information than a chairman's letter.

Which raises the question of what all that concentration was invested in — and what happened when it stopped working.

---

↑ Back to Top

V. The Reckoning: Property Exposure and "The World's Worst-Performing Bank" (2021–2024)

On January 11, 2022, Bloomberg published a story with a headline that a bank's investor-relations department can never fully live down: "World's Worst-Performing Bank Lent Billions to China Evergrande." Minsheng's shares had fallen 31% over the preceding twelve months — the worst performance among the 155 members of Bloomberg's World Banks Index. The bank that had once been marketed to global investors as the future of Chinese banking, a privately run lender in a state-run system, had become the index's designated casualty of what the story called an ill-fated push into property lending.¹⁰

To understand how a private-enterprise bank ended up as 中国恒大 China Evergrande's largest bank creditor, you have to understand what "lending to private enterprise" came to mean in China between 2012 and 2020. The original thesis — finance the manufacturers and traders the state banks ignore — ran into a hard commercial reality. Manufacturing is capital-intensive, low-margin, and hard to underwrite. Property development, by contrast, offered collateral you could see, borrowers who wanted enormous tickets, and a decade-long track record of prices only going up. Minsheng's "private enterprise" franchise drifted, as did the entire Chinese banking system's, toward developers.

By mid-2021, as Evergrande's liquidity crisis became undeniable, Minsheng was reported to have pared its exposure to the developer to roughly RMB30 billion from about RMB40 billion over the previous year, telling investors through the Shanghai Stock Exchange platform that the risk was "maintained within a controllable range."¹¹¹² Cutting a position by a quarter as a borrower slides toward default is not risk management; it is triage. And it left the bank still holding tens of billions against a company that would become the largest property default in history.

The systemic picture was worse than the Evergrande headline. Citigroup analysts estimated that Minsheng carried about RMB130 billion of exposure to high-risk developers — equivalent to roughly 27% of its Tier 1 capital, the highest such ratio among China's major lenders.¹⁰ That ratio is the number that matters. Absolute exposure tells you the size of a problem; exposure against capital tells you whether the problem can kill you. At 27% of Tier 1, a severe enough outcome would have consumed a quarter of the bank's loss-absorbing equity.

When a Founder Becomes a Defendant

Then the founding-shareholder fault line opened.

Oceanwide — Lu Zhiqiang's group, and one of the founding investors from 1996 — went into terminal decline as property credit dried up. In January 2023, Minsheng's Beijing branch began filing lawsuits over financial lending contract disputes against a chain of Oceanwide entities: Wuhan Centre Building Development Investment, Wuhan CBD, Oceanwide Holdings, and China Oceanwide Holdings Group. The claims totalled around RMB7 billion (roughly US$1 billion), comprising about RMB3.97 billion owed by one Wuhan vehicle and RMB3.05 billion by the other, with Lu named personally as de facto controller.¹³

Read that sentence slowly, because it describes something genuinely rare. A bank sued one of its own founding shareholders — a man who had served as a non-executive director and who, seven years earlier, had spent RMB7.5 billion defending his influence over its board — for RMB7 billion of unpaid debt. On September 22, 2023, a Bermuda court accepted a winding-up petition against China Oceanwide Holdings Group, whose liabilities stood at just over US$2 billion as of June 2023, with the order following days later. The petition had been brought by lenders over a New York property project and demanded repayment of more than US$175 million.¹⁴

Minsheng won. As disclosed in the FY2025 annual report, the bank obtained enforceable judgments in all seven lawsuits against the two Oceanwide holding entities and moved to enforcement, though two cases have since been ruled terminated for enforcement — the standard outcome when there are no recoverable assets left to seize. The Wuhan cases are in enforcement with one suspended.¹ Winning a judgment against an insolvent borrower is a legal victory with limited economic content.

Management's public posture through the episode was consistent and calmly framed. At the June 2024 shareholders' meeting, chairman 高迎欣 Gao Yingxin characterised the bank's remaining Oceanwide exposure — RMB18.7 billion of outstanding loans at end-2023 — as a small fraction of the loan portfolio, with proactive provisions already taken and mitigation under way.¹⁵ Whether that framing was reassurance or minimisation depends on your read of the provisioning, and the provisioning is where the story gets interesting.

The Credit Cost Arc — and Why It Has Not Ended

Impairment losses on credit are the cleanest window into what management actually believes about its loan book, because provisions are a forecast expressed in money.

In 2021, Minsheng booked RMB77,398 million of credit impairment charges — roughly 47% of that year's entire operating income. That is not a provision; it is a controlled demolition. For the three years that followed, charges settled into a RMB45–49 billion range: RMB48,762 million in 2022, RMB45,707 million in 2023, RMB45,474 million in 2024. The natural reading was that the worst had been absorbed and the run-rate was normalising.¹

Then 2025 broke the pattern. Charges rose 18.64% to RMB53,950 million, with the loan-specific component up 22.80% to RMB47,901 million.¹ And this happened in the same year that management reported improving asset quality — the real-estate NPL ratio nearly halved from 5.01% to 3.61%, with real-estate bad loans falling by RMB4,962 million.¹

How can bad loans in the problem sector fall sharply while provisions rise sharply? The industry breakdown answers it, and the answer is not comfortable. Non-performing loans in leasing and commercial services — a catch-all category that in China frequently houses holding companies, financing vehicles and asset-management structures — rose by RMB3,026 million, taking that sector's NPL ratio from 0.34% to 0.82%. Wholesale and retail NPLs rose RMB2,356 million, with the ratio climbing from 1.44% to 2.27%. And personal loan NPLs edged up, with the retail NPL ratio rising from 1.80% to 1.92%.¹ The group NPL ratio therefore rose, from 1.47% to 1.49%, despite the property improvement.¹

The analytical conclusion is straightforward and important: Minsheng is not finishing a credit cycle, it is rotating through one. Property risk is being resolved. Corporate-holding-company risk and consumer credit risk are appearing. That is a materially different investment proposition from "the worst is behind us."

The Capital Raise That Never Happened

One final episode from this period deserves attention, because it speaks to capital discipline in a way no ratio can.

For roughly six years, Minsheng pursued a RMB50 billion (about US$6.9 billion) convertible bond — a security that raises debt now and converts to equity later, replenishing the core capital that absorbs losses. The Shanghai Stock Exchange accepted the application in March 2023. On August 11, 2023, the bank abandoned the plan. The explanation offered was that the decision followed "careful analysis and consideration of the capital market environment," with a bank representative noting that "the secondary market has limited capacity at the moment" and that the plan "was initiated a long time ago and is not suited to the current market conditions."¹⁶

Six years to fail to raise capital, at precisely the moment capital was most needed, is a data point about institutional decisiveness. The consequence shows in the core Tier 1 capital adequacy ratio, which grinds upward at glacial pace: 9.04% in 2021, 9.17%, 9.28%, 9.36%, and 9.38% at end-2025.¹ That is organic accretion — retained earnings, nothing more. Meanwhile the bank funded itself with instruments that do not count as core equity: RMB40 billion of tier-2 capital bonds at 2.35% in April 2025 and RMB30 billion of undated (perpetual) capital bonds at 2.30% in June 2025, lifting perpetual bonds outstanding to RMB105 billion from RMB75 billion.¹

The distinction matters for a non-specialist. Core Tier 1 capital is common equity — the money that absorbs losses first and without limit. Perpetual and tier-2 instruments sit above it in the loss waterfall and carry coupons that must be serviced. Filling the capital stack with the cheaper, weaker forms preserves the share count but does less to build genuine resilience. For comparison, CMB reported a core Tier 1 ratio of 11.92% under the weighted approach at end-2025 — more than 250 basis points of additional cushion.²⁷

The dividend tells a related story. Total distributions fell from RMB9,457 million for 2023 (RMB2.16 per ten shares) to RMB8,406 million for 2024 (RMB1.92) to RMB8,274 million for 2025 (RMB1.89).¹ Three consecutive declines. Notably, the payout ratio barely moved — 30.14% of ordinary-shareholder profit in 2025, essentially the same as the prior two years.¹ So the dividend is not being cut by choice. It is shrinking mechanically because earnings are shrinking. For an income-oriented holder, that is the more troubling of the two explanations.

Property was the visible crisis. The less visible one had been running in parallel, inside the bank's own controls.

---

↑ Back to Top

VI. Governance Under Strain: Fines, Fraud, and a Contestable Board

In April 2017, a client of Minsheng's Hangtianqiao (航天桥) sub-branch in Beijing made a routine enquiry about a wealth-management product that had passed its maturity date. The answer came back that the product did not exist. Not that it had underperformed, or been restructured — that it had never been created at all.¹⁷

What unravelled from there was one of the largest single-branch frauds in modern Chinese banking. The branch president, 张颖 Zhang Ying, had since 2013 lured clients with promises of high yields, induced them to sign purchase or transfer agreements for fabricated wealth products, and routed the money into personal accounts she controlled — spending it on property, cars, luxury goods, and cash inducements to attract further deposits. The courts ultimately found she had defrauded 147 victims of more than RMB2.746 billion. She was sentenced to life imprisonment with permanent deprivation of political rights and confiscation of all personal property; Beijing's High People's Court upheld the sentence on second-instance judgment dated November 29, 2020.¹⁸

Many of the victims were long-standing private-banking clients, some cultivated over a decade through a bank-sponsored golf club. One captured the damage in a single line quoted at the time: "If we can't even trust a big national bank, what other financial institutions can we trust?"¹⁷

This matters analytically for a reason distinct from its drama. Property losses were a market-cycle failure — bad judgment about an asset class, shared with the entire Chinese banking system. The Hangtianqiao fraud was an internal-control failure. It ran for four years inside a branch of a systemically important bank, undetected. Those are different diseases with different prognoses, and only one of them gets better when the property market recovers.

A Pattern, Not an Incident

The regulatory record since suggests the control problem was not confined to one branch.

In July 2021, Minsheng was fined RMB115 million for a list of violations spanning incomplete internal systems and non-compliant operations. In February 2023 the China Banking and Insurance Regulatory Commission announced penalties totalling RMB388 million across five institutions, of which Minsheng's share was RMB89.724 million, for conduct including diversion of loan funds, falsification of statistical data, and non-standard review procedures for major transactions.¹⁹ That August brought a further RMB44.3 million penalty. Cumulatively, Chinese financial press tallied more than RMB350 million of large-ticket fines against Minsheng since 2020 — enough for the label 罚单之王, "king of penalty notices."²⁰

The most important entry, though, is the most recent. On October 31, 2025, the National Financial Regulatory Administration fined the bank RMB58.65 million, with six named individuals warned and fined a combined RMB360,000. The cited grounds: imprudent management of loan, bill, interbank and related businesses, and non-compliant regulatory data reporting.²¹ Minsheng was one of five lenders penalised that day, alongside 中国银行 Bank of China (RMB97.9 million), 农业银行 Agricultural Bank of China, 平安银行 Ping An Bank and 浦发银行 Shanghai Pudong Development Bank — so this was a sector-wide sweep rather than a Minsheng-specific indictment.²² By incomplete tallies, the bank and its branches were penalised more than RMB86 million across 2025.²¹

Two things are worth saying about that. First, the sector-wide nature of the October action genuinely mitigates the reading — regulators were disciplining an industry, not singling out one bank. Second, and less comfortably, "non-compliant regulatory data reporting" in 2025 is a repeat of the "falsification of statistical data" cited in 2023. When the same category of finding recurs after two years and an announced compliance overhaul, the natural inference is that remediation has been slower than the rhetoric.

And the rhetoric is emphatic. The FY2025 report describes a "Year of High-Quality and Compliant Development" (高质量合规发展年), a group-wide internal control and compliance system, and the philosophy that "compliant operation is the core competitiveness."¹ There is also a notable governance change: following NFRA approval in September 2025, the bank no longer has a Board of Supervisors, with its functions folded into the board's structure under a revised Articles of Association.¹ That is a system-wide Chinese reform rather than a Minsheng initiative, but it does concentrate oversight in a board that has historically been the arena for shareholder conflict rather than a check on it.

The Related-Party Question

Which brings us to the issue a sceptical investor should press hardest.

Chinese financial commentary has argued for years that the founder-shareholder groups — 泛海系 the Oceanwide bloc, 东方系 the Orient bloc, 新希望系 the New Hope bloc — enjoyed preferential access to Minsheng credit. The bank's own related-party disclosures do not settle the argument, but they do document the scale. At end-2025, related-party loan balances included RMB3,712 million to Shanghai Giant Investment Management, RMB2,817 million to Orient Group Incorporation, RMB2,542 million to Oriental Group, RMB1,266 million to Tongfang Guoxin Investment, and a string of real-estate development entities identified as related parties of New Hope Liuhe Investment and other shareholders.¹ The bank states these transactions are conducted on normal commercial terms consistent with pricing for unrelated parties.¹

The single most striking related-party disclosure has nothing to do with the founders. During 2025, Minsheng granted Dajia Life Insurance — its largest shareholder — a comprehensive credit limit of RMB26 billion with a two-year term, against which RMB4.625 billion was drawn at year-end.¹ A bank extending a RMB26 billion facility to the institution that holds nearly 18% of its equity is, on its face, exactly the circular arrangement the 1996 fragmented-ownership design was meant to make impossible. It is disclosed, it is presumably approved, and it is within regulatory limits — the bank reported that all large-exposure requirements were met.¹ It is also precisely the sort of item an activist would put on the first slide.

Set against that, the concentration data cited earlier — top-ten borrower exposure falling to 8.34% of net capital — is genuine evidence that the structural problem is being reduced rather than merely rebranded.¹ Both things are true. The credibility test for the current management team is whether that trend continues through a period when the property market stops improving, or whether it reverses the moment the pressure comes off.

Which requires knowing who the current management team actually is.

---

↑ Back to Top

VII. Who's Running It Now: Gao Yingxin, Wang Xiaoyong, and the New Shareholder Register

In mid-2020, with the property cycle turning and the bank's shareholder politics still unresolved, the board of China's original private bank went looking for a chairman. It did not promote from within, and it did not select a private entrepreneur. It hired a career state banker.

高迎欣 Gao Yingxin, born in 1962, arrived in July 2020 after two and a half years as vice chairman of the board and chief executive of BOC Hong Kong (Holdings) and Bank of China (Hong Kong).¹ Caixin had reported his likely appointment that May.²³ His résumé before Hong Kong reads as a complete circuit of 中国银行 Bank of China's institutional machinery: executive vice president of the parent from February 2015, an executive director from December 2016, a deputy chief executive in Hong Kong from 2005, president and chief operating officer of BOC International in 2004, and — most relevantly — general manager of Bank of China's head-office corporate banking department from 1999 to 2004 and deputy general manager of its credit department from 1996. He holds a master's degree in engineering from East China University of Science and Technology and the professional title of senior economist.¹

He succeeded Hong Qi, who had held the chair for six years, and Chinese commentary noted he was the first external appointment to the chairmanship in more than a decade.⁵ The signal in that hire is not subtle. A board that had spent twenty years being fought over by entrepreneurs chose someone whose entire formation was in the credit and corporate-banking departments of the most internationally exposed state bank in China. It was, in effect, an admission that the private-governance experiment needed a state-trained risk professional to stabilise it.

The president's chair went the same direction. 王晓永 Wang Xiaoyong, born in 1970, joined as president in April 2024 and became vice chairman and an executive director that August. He came from 中国建设银行 China Construction Bank, where between September 2016 and February 2024 he had served as president of the Chongqing branch, general manager of the channel and operation management department, and president of the Sichuan branch. Crucially, his earlier career was almost entirely in risk: the audit department, the office of the risk and internal control management committee, and the risk management department through 2006, then assistant general manager of risk monitoring and control and deputy general manager of credit management. He holds a doctorate in management science and engineering from Tianjin University.¹

Wang inherited a bank mid-repair rather than mid-crisis. His predecessor 郑万春 Zheng Wanchun had overseen the disposal of the bulk of the legacy problem assets during 2020–2022; the heavy provisioning year of 2021 predates Wang entirely. The current management team's task is not triage. It is proving that a stabilised bank can also be a profitable one.

The Track Record, Tested

So test it. The most unforgiving metric for a bank is return on equity — what shareholders actually earn on the capital they have tied up.

Minsheng's return on weighted average equity fell every single year under Gao's chairmanship: 6.59% in 2021, 6.31%, 6.10%, 5.18%, and 4.93% in 2025.¹ Five consecutive declines. Return on average assets followed the same path, from 0.50% to 0.39%.¹ To put that in perspective, CMB earned a return on average equity of 13.44% in 2025 — nearly three times Minsheng's — on an asset base roughly two-thirds larger.²⁷

Some of that decline was unavoidable. Absorbing property write-downs while margins compressed sector-wide would have hurt any bank. But five straight years is long enough that "cyclical" stops being a complete explanation.

The 2025 result was designed to be the turn, and in one respect it was: revenue grew for the first time since 2021, and the composition of that growth — cheaper deposits, higher fee-and-investment income, a stabilised margin — reflects deliberate management action rather than luck.¹ Management's own presentation of the year, delivered in April 2026, led with the framing that the strategy was "effective, with visible results."³¹

But the profit line did not follow, and the first quarter of 2026 made the divergence starker rather than narrower. Operating income rose 2.74% to RMB37,822 million; net profit attributable to shareholders fell 9.64% to RMB11,514 million; annualised return on weighted average equity dropped to 8.08% from 9.22%; and credit impairment charges rose to RMB13,892 million from RMB10,858 million — an increase of roughly 28%.³ The favourable items were there too: margin up two basis points to 1.43%, NPL ratio down to 1.46%, cost-to-income improving to 25.92%, and corporate loans up 5.26% in a single quarter.³ But when a bank's provisions grow ten times faster than its revenue, the provisions are the story.

Is this a genuine inflection or a stabilised decline? On the evidence available in August 2026, the honest answer is that revenue has inflected and earnings have not, and that the gap between them is entirely explained by credit costs that management has not yet been able to bring down.

The Credibility Assessment

A word on how this management team communicates, because it bears on how much weight to put on their claims.

The consistency is genuinely high. The three strategic positionings — a bank for NSOEs, an agile and open bank, a bank with considerate services — appear in identical language across years of filings.¹³² There is no evidence of the strategy-of-the-year churn that plagues turnaround stories, and the 2026 priorities in the report's Prospects chapter are recognisably continuous with 2025's.¹

The specificity is much lower. The FY2025 report's forward-looking section and the results presentation's "Future Prospects" section contain no public quantitative targets — no revenue guidance, no ROE goal, no credit-cost expectation, no capital-ratio objective. What they contain is directional language: "stabilise growth, optimise structure, enhance foundation and improve quality and effectiveness."¹³¹ The 2025 report also carries a formal statement that the bank "has not found any material risks that will adversely affect" its strategy and objectives.¹

This is normal for Chinese listed banks and should not be read as evasion. But it does mean investors cannot hold this team to targets, because there are none. Accountability has to be constructed from the disclosed metrics instead — which is why the KPI discipline in the final section matters more here than it would for a company that guides.

One more datum on alignment. Chairman Gao's pre-tax remuneration from the bank for 2025 was RMB3.292 million; President Wang's was RMB3.211 million. Gao personally holds 500,000 A shares, and executive director 张俊潼 Zhang Juntong holds 350,000 — positions worth roughly RMB1.7 million and RMB1.2 million respectively at current prices. Liu Yonghao, as a non-executive director, received RMB695,000 and holds no shares directly, though his group holds a substantial stake. Independent directors received between RMB260,800 and RMB900,000.¹ Senior executive share ownership is modest in absolute terms — an alignment gap that is standard in Chinese state-influenced banking and worth noting rather than dramatising.

The Register, Rearranged

The most consequential development at Minsheng over the past decade was not a strategy or a hire. It was a change in who owns it.

Dajia Life Insurance holds an aggregate 17.84% of the bank's shares, split across a universal-insurance product account (10.30%) and a traditional product account (6.49%) — making it unambiguously the largest shareholder, with the caveat that the bank continues to state it has no controlling shareholder or de facto controller.¹² Dajia Life's controlling shareholder is Dajia Insurance Group, whose de facto controller and ultimate beneficiary is China Insurance Security Fund Co., Ltd.¹ Dajia Insurance Group was established on June 25, 2019, with registered capital of RMB20.36 billion, owned 98.23% by the Insurance Security Fund alongside token stakes held by SAIC Motor (1.22%) and Sinopec (0.55%), specifically to absorb Anbang Life, Anbang Pension and Anbang Asset Management.²⁶

Follow the chain. The stake that Anbang assembled in a two-month buying spree in late 2014 to seize control of China's private bank passed, after Anbang's seizure and Wu Xiaohui's imprisonment, to a resolution vehicle controlled by the state's insurance guarantee fund. Nobody set out to nationalise the shareholding of the private bank. It happened as a byproduct of cleaning up someone else's failure. The "private bank" branding is now largely historical — a description of origin, not of ownership.

The second development is more human. Liu Yonghao is still on the board, thirty years after helping found the institution, as vice chairman and a non-executive director.¹ And beginning in late 2024, his group started buying again. On November 18, 2024, New Hope's board approved a plan for New Hope Chemical Investment to lift the group's holding above 5% of the A shares through market and block purchases, which would have made it the fourth-largest shareholder, up from sixth at 4.31% as of the third quarter of 2024.²⁴

The history behind that decision explains the emotional charge Chinese commentary attached to it. Liu had held 8.38% in 2014 and then sold into the shareholder wars — 185.4 million shares in December 2014 for about RMB2 billion, 261.9 million shares in July 2015 for RMB2.734 billion, and more in 2016 — reducing to 4.43% by 2018.²⁴²⁵ Further back, in 2006, he had been removed from the board despite being a major shareholder, an episode he later described in terms that made clear it had stung.²⁵ The 2024 stake-building was read by Chinese press as an eighteen-year-delayed reassertion by a founder finally acting from strength rather than defence.²⁵

By end-2025, the New Hope group's aggregate deemed interest stood at 1,930,715,189 A shares — 5.44% of the A-share class and 4.41% of all issued ordinary shares, with Liu's daughter 刘畅 Liu Chang and his spouse 李巍 Li Wei both deemed interested through the group structure.¹ As of March 31, 2026, New Hope Liuhe Investment ranked fifth on the top-ten register at 4.18%.²

How should an investor read insider buying by a founder who sits on the board? Carefully. It is a signal of conviction from someone with unusually good information, and that counts for something. But it is also a signal from someone with reasons beyond return — a thirty-year identification with the institution, a board seat to protect, and a public history to vindicate. And a 4.4% position does not confer control of anything. What it does confer is the ability to reopen the very contest that has periodically consumed this board since 2009.

The register also still carries the ghosts. Shenzhen Liye Group at 4.49% with 929 million shares pledged; 同方国信 Tongfang Guoxin at 3.33% with 320 million shares pledged; 上海巨人 Shanghai Giant Lifetech, Shi Yuzhu's vehicle, at 3.15%; and 中国长城资产 China Great Wall Asset Management — a state bad-bank — at 3.66%, deemed to hold a further 490 million H shares alongside Central Huijin.²¹ Pledged shares are shares posted as loan collateral. When a shareholder's own borrowings sour, the pledge can be enforced and the block can move. That is not a theoretical risk at this bank; it is how Oceanwide's position unwound.

An institution this contestable, with returns this thin, invites an obvious question: how does it compare with the banks it competes against every day?

---

↑ Back to Top

VIII. Competitive Landscape: Where Minsheng Sits Among the Joint-Stock Banks

Chinese banking is often described as if it were monolithic, which obscures the structure that actually matters. At the top sit the large state commercial banks — Industrial and Commercial Bank of China, China Construction Bank, Agricultural Bank of China, Bank of China — which operate at a scale and with a policy mandate that puts them in a different category entirely. Below them sit the twelve national joint-stock commercial banks, of which Minsheng is one. These are the true competitors: nationally licensed, commercially driven, and fighting each other for the same corporate relationships, the same deposits, and the same affluent retail customers.

Within that peer group, Minsheng sits in the middle-to-lower tier by size and near the bottom by profitability — and the second fact matters far more than the first.

Measure it by market value, which is the market's summary judgment on both. On August 11, 2026, Minsheng's A shares closed at RMB3.49, capitalising the bank at roughly RMB148 billion.²⁸ CMB's Hong Kong shares traded at HK$48.34, for a market capitalisation of approximately HK$1.16 trillion.²⁹ Industrial Bank, CITIC Bank, SPD Bank and Ping An Bank all carry substantially larger market values than Minsheng despite balance sheets that are not proportionally larger. Minsheng's asset base is roughly 60% of CMB's; its market value is closer to one-sixth.

Where does that six-fold gap come from? Almost entirely from return on equity, and the arithmetic is unforgiving. A bank earning 13.44% on equity and trading near book value is priced consistently with a business that creates value. A bank earning 4.93% on equity — below almost any plausible estimate of its cost of capital — mathematically should trade at a discount to book, because each additional yuan of retained capital earns less than shareholders require.¹²⁷ Minsheng's price-to-book of roughly 0.27 is not the market mispricing an asset. It is the market pricing an ROE.

What CMB Has That Minsheng Doesn't

The comparison is worth making concrete, because it isolates the mechanism.

CMB's advantage is a deposit franchise. In 2025 its average cost of interest-bearing liabilities was 1.26%, and the daily average balance of demand deposits was 49.40% of total deposits.²⁷ Demand deposits are the cheapest money in banking — current-account balances that customers leave with a bank because it is convenient, not because they are being paid for it. Roughly half of CMB's funding comes from money that barely needs a yield.

Minsheng's deposit cost ratio, after a hard-won 40-basis-point reduction, was 1.74%.¹ That gap of nearly half a percentage point on the liability side is, functionally, the ROE gap. It exists because CMB spent two decades building a retail wealth-management franchise deep enough that affluent Chinese households keep their primary banking relationship there — 224 million retail customers, with retail finance generating 56.63% of net operating income and 50.66% of pre-tax profit.²⁷ Compare that with Minsheng, where retail generates 40.5% of revenue but under 30% of the pre-tax profit produced by the two profitable segments combined.

CMB also carries a cost-to-income ratio of 32.01% against Minsheng's 36.52%, an NPL ratio of 0.94% against 1.49%, and allowance coverage of 391.79% against 142.04%.²⁷¹ That last comparison is the one an investor should sit with. Allowance coverage measures reserves against recognised bad loans. CMB holds nearly four yuan of reserve for every yuan of NPL. Minsheng holds about one and a half, against a regulatory floor of 130%.¹ The practical implication: CMB has a large buffer it can release into earnings during a downturn, while Minsheng has very little room before it bumps into the regulatory minimum. This is why Minsheng's provisions flow so directly to the profit line, and why its earnings are more volatile than its balance sheet suggests.

Minsheng's asset-quality trajectory is improving from a worse base; its capital position is thinner; its reserve buffer is thinner still. None of those is a matter of interpretation.

The Five Forces, Applied Honestly

Porter's framework produces an unusually clear read on Chinese joint-stock banking, and it explains why scale does not decide outcomes here.

Barriers to entry are close to absolute. Nobody founds a new national commercial bank in China; the licence regime forecloses it. But this protects all twelve incumbents equally, which means it is an industry-level moat, not a Minsheng advantage. In Hamilton Helmer's terms, it is not a Power because it confers no differential.

Buyer power is high and rising. A creditworthy Chinese corporate can source financing from twelve joint-stock banks, four state giants, dozens of city commercial banks, and the bond market. A retail depositor can move money between banks and wealth platforms with a phone. Neither side faces meaningful switching costs. Minsheng's own president's letter, with its emphasis on winning "high-quality liabilities" through payroll agency and settlement scenarios, is implicitly an acknowledgement that deposits must be earned transaction by transaction.¹

Rivalry is intense and price-based, which is the worst kind. When products are near-identical and switching is easy, competition runs through pricing — and in a falling-rate environment that compresses margins for everyone. The FY2025 report's own outlook anticipates continued industry supply-side reform and consolidation among smaller institutions.¹

Supplier power — the cost of funding — is set principally by central bank policy, not negotiated. Banks are price-takers on the macro rate, and differ only in deposit mix. Which is exactly where CMB wins.

Substitution threatens fee income more than lending. Wealth platforms and fintech distributors have taken share in fund and insurance distribution, and industry-wide fee reductions have compressed what remains. Core corporate lending is harder to disintermediate, because underwriting a mid-sized private company still requires balance sheet and local knowledge.

Net assessment: an industry with formidable entry barriers and almost no differentiation among the incumbents. That combination guarantees the sector survives and guarantees that survival is not enough. Returns accrue to whichever bank has the cheapest funding and the best credit discipline — which is a statement about operating capability, not about market position.

Run Helmer's Seven Powers against Minsheng and the result is sparse. Scale economies: present in absolute terms, but Minsheng is the smaller party in every comparison that matters. Switching costs: modest, though payroll-agency and supply-chain integration — the bank reported 43,368 supply-chain corporate financing clients introduced by strategic clients, up 12,870 in a year — create some genuine stickiness, since a company whose payroll runs through your systems does not casually leave.¹ Branding: real, and unusual — thirty years of consistent NSOE positioning gives Minsheng a recognisable identity, and its MSCI ESG rating of AAA, maintained for two consecutive years, is a legitimate differentiator among Chinese banks.¹ Cornered resource: none. Counter-positioning: this was the founding Power in 1996, and it has been fully competed away. Network economies and process power: no evidence.

So the honest competitive conclusion is that Minsheng has an identity rather than a moat. That is not fatal — plenty of money is made in undifferentiated industries by operators who execute — but it means the entire investment case rests on execution, not on structure. Which is precisely the question the bull and bear cases have to resolve.

---

↑ Back to Top

IX. Bull Case, Bear Case, and the "Value Trap" Question

There is a specific kind of investment that fascinates and destroys value investors: the deeply discounted bank. The pitch writes itself. Assets are real, the franchise is decades old, the licence is irreplaceable, and the shares trade at a fraction of stated book value. Sometimes it is one of the great asymmetric opportunities in finance. Sometimes the book value is wrong and the discount is the market being right.

Minsheng in August 2026 is a textbook instance of the problem, and the intellectually honest position is that it has not been resolved.

The Bull Case

Start with the valuation, because it is the foundation of everything else. At RMB3.49 the shares trade at roughly 5.9 times trailing earnings and around 0.27 times the RMB12.90 book value per ordinary share the bank reported at end-2025, having risen further to RMB13.14 by March 2026.²⁸¹³ The trailing dividend yield sits above 5%.²⁸ The market is pricing either a large future write-down of stated book value, or a permanent inability to earn an adequate return on it. If neither proves true, the arithmetic is powerful — a bank that merely re-rates from 0.27 to 0.4 times book delivers a substantial return without any earnings growth at all.

Second, the property problem is measurably shrinking. Total corporate credit exposure to real estate — loans, off-balance-sheet credit and bond investments combined — fell 3.37% to RMB381,865 million in 2025, with the loan component down 2.40% to RMB325,443 million and non-performing real-estate loans down by nearly RMB5 billion. Off-balance-sheet real-estate-linked business not bearing credit risk shrank 7.85% to RMB49,292 million, a modest RMB49 billion.¹ The developer overhang that defined 2021 is being worked down, not warehoused.

Third, external credit assessment moved in the bank's favour during 2025. Fitch upgraded Minsheng's long-term issuer default rating one notch to BBB− from BB+, citing the bank's rising systemic importance.³⁰ For a bank that funds itself partly in wholesale markets, a move into investment grade is not cosmetic — it lowers the cost of the perpetual and tier-2 issuance the bank has been relying on.

Fourth, the operating cost story is real. Cost-to-income improved to 36.52% from 38.55%, deposit costs fell hard, and the first quarter of 2026 showed cost-to-income at 25.92%.¹³ Combined with the reduction in top-ten borrower concentration and the wealth-management build-out described earlier, this is a management team demonstrably pulling the levers available to it.

Fifth, an insider with thirty years of information is buying rather than selling.

The Bear Case

Now the other side, and it is not a list of macro worries — it is a set of specific contradictions inside the same disclosures.

The central one: two consecutive years of falling net profit, extended to a third by the first quarter of 2026, while revenue rose. In a bank, that pattern has exactly one mechanical cause — credit costs — and Minsheng's credit costs accelerated in 2025 and accelerated harder in early 2026. Management is still finding problems.

The rotation of those problems is the more troubling detail, and it is the argument against the bull case's central plank. If property NPLs are falling while leasing-and-commercial-services and wholesale-and-retail NPLs are rising fast enough to push the group ratio up, then the improvement in the property book is being offset by deterioration elsewhere. A bull who buys the "property cycle is over" thesis has to explain why the sector composition of bad loans is changing rather than shrinking.

Operating cash flow was negative in both 2024 and 2025 — RMB231,638 million and RMB157,476 million respectively.¹ For a bank this figure is heavily influenced by deposit and interbank flows and should not be read as a solvency signal. But two consecutive large negatives, in a year when total deposits grew only 0.66% and total assets grew 0.23%, describe a balance sheet that is not funding growth.¹ The Q1 2026 figure improved dramatically to negative RMB357 million from negative RMB117,455 million, which is worth noting as a genuine improvement.³

The shrinking dividend, on an unchanged payout ratio, means income holders are exposed directly to the earnings decline with no policy buffer.

Regulatory findings continued into late 2025, including a repeat category from 2023.

And the governance structure remains exactly as contestable as it was in 2009 — no controlling shareholder, a state-linked insurer in the top seat, a founder rebuilding a stake, two significant holders with large share pledges outstanding, and a RMB26 billion credit line extended to the largest shareholder.

Where an Activist Would Push

A skeptical investor with a seat at the table would open on three fronts.

Capital. Why did a RMB50 billion convertible bond take six years to not happen? A bank with the thinnest core capital ratio among its major peers, staring into a property downturn, is exactly the institution that should have raised equity early and at a discount. Instead it waited, watched the window close, cited market conditions, and then funded itself with instruments that service coupons rather than absorb losses. That sequence is either a failure of decisiveness or evidence that existing shareholders' aversion to dilution outweighed the bank's need for resilience. Either explanation is a governance finding.

Provisions. Why are impairment charges rising in a year management characterises as a turnaround, and why is allowance coverage sitting at 142% when the best-run peer holds nearly 392%? Management would answer that it is disposing of problem assets aggressively, which is consistent with the surge in substandard and doubtful loan migration ratios — 90.11% and 89.55% respectively in 2025, up sharply from 45.06% and 46.63% in 2023.¹ Those migration figures are actually supportive of management: they indicate recognised problem loans are being pushed through to resolution rather than parked. But it is also the case that a bank with a thin reserve buffer has less freedom to absorb the next surprise.

Portfolio and register. What is the strategic case for a subsidiary and treasury segment that has lost money before tax in both of the last two years? And do the Dajia and New Hope dynamics point toward a coherent strategy, or toward another turn of the battlefield-for-capital cycle? The board approved a merger by absorption of the Minsheng rural banks on December 26, 2025, subject to regulatory approval — a sensible simplification of a small, 77-outlet operation, and the sort of tidying that suggests some awareness of the complexity problem.¹

The Verdict That Isn't One

Which brings us to the question in the section title, and to a genuinely unsatisfying answer.

The bull case requires believing that credit costs are peaking and will fall, releasing the gap between revenue and profit. The bear case requires believing that credit costs are structural — that a bank with this reserve buffer, this capital position, and this history of finding new problem sectors will keep provisioning at a level that suppresses returns.

The available evidence supports the bear case on timing and leaves the bull case unresolved rather than refuted. Revenue has genuinely inflected; the operating improvements are real and management-driven, not accidental; the property overhang is genuinely shrinking. But the single metric that would confirm a turnaround — the earnings trend — has gone the wrong way for three consecutive reporting periods, and the reason is provisions that are still rising.

The distinction that matters is between a trough in bad news and an inflection in returns. Minsheng has plausibly reached the former. There is not yet evidence for the latter. Anyone underwriting this as a recovery is underwriting the credit cost line, which means there are only a handful of numbers worth watching.

---

↑ Back to Top

X. What to Watch: KPIs, Risks, and the Road Ahead

Strip away the strategy language, the anniversary rhetoric and the segment detail, and the Minsheng investment case reduces to three trackable things.

One: the net profit trend, quarter by quarter. This is the master metric, because it is where every other tension resolves. Revenue growth has already inflected. The question is whether earnings follow. A first quarter of profit growth after three consecutive declines would be the strongest available evidence that the credit cycle has genuinely turned. Continued decline through 2026 would suggest the FY2025 revenue recovery was a cost-and-mix story rather than a franchise recovery — real, but not sufficient. The relevant test comes soon: interim results for the first half of 2026 are expected in late August 2026, consistent with the August 29 timing of the prior year's interim announcement.¹

Two: the NPL ratio and the credit impairment charge, read together. Never separately. Either number alone can be made to tell a flattering story — a falling NPL ratio suggests healing, and rising provisions can be spun as prudence. Together they reveal what management actually believes. Falling NPLs alongside falling provisions would be a genuine turn. Falling NPLs alongside rising provisions, which is the current configuration, means recognised problems are being resolved while new ones are being identified. The sector breakdown deserves the same attention: whether leasing-and-commercial-services and wholesale-and-retail deterioration stabilises will determine whether the property improvement translates into anything at group level. Allowance coverage against the 130% regulatory floor is the constraint to watch alongside it.

Three: return on equity, versus CMB and the joint-stock peer average. This is the valuation story compressed into a single number. The six-fold market-value gap with CMB is an ROE gap and nothing else. Minsheng does not need to close it — it needs to stop widening it. A single year in which return on equity rises rather than falls would break a five-year pattern and would be the most economically meaningful event in the bank's recent history.

The Risk Radar That Actually Applies

Not every macro risk is relevant here. Several are.

Residual property exposure remains the largest single credit concentration, at RMB381,865 million of corporate real-estate credit — roughly half the bank's entire net capital base of RMB784,943 million.¹ Even after paydown, a renewed leg down in Chinese property prices would flow through this book directly. The bank's stated approach is "controlling total amount, revitalising existing business and optimising incremental business," with priority given to affordable and rental housing.¹ That is sensible positioning. It is not insulation.

Refinancing and cost-of-capital risk follows from the capital structure described earlier. Having abandoned its convertible bond and relying on perpetual and tier-2 issuance, Minsheng must return to wholesale markets periodically to roll and grow that capital. The Fitch upgrade helps. A reversal would hurt more than it would at a better-capitalised peer.

Regulatory and compliance risk is live rather than historical, given the October 2025 penalty and the repeat data-reporting finding. Fines at this scale are not financially material against a RMB139 billion revenue base. What they signal about control quality — and what a further escalation could mean for business permissions — is the actual exposure.

Governance risk is structural. Two decades of boardroom contests, a register with no controlling owner, significant pledged blocks, and a related-party credit line to the largest shareholder together describe an institution where ownership disputes are a recurring feature rather than a tail event.

China macro and policy risk shapes everything and is controllable by nobody at the bank. Management's own 2026 outlook expects nominal growth to rebound, liability costs to fall further, and asset quality to stay stable.¹ Those are reasonable expectations. They are also, notably, the assumptions on which the entire recovery case depends — and the bank has published no scenario for what happens if they do not hold.

One technology note, since it comes up. Minsheng has deployed AI across sales, operations and management, with named internal systems — 慧销 Huixiao for retail sales assistance, 慧芯 Huixin for process automation, and a decision-making agent called 慧眼 Huiyan — described in terms of a "one employee plus N digital avatars" service model.¹ The cost-to-income improvement is consistent with efficiency gains from this kind of automation. But no disclosed metric isolates the contribution, and AI adoption in Chinese banking is universal rather than differentiating. Treat it as table stakes, and watch cost-to-income rather than the announcements.

Near-Term Catalysts

Three things will move the story in the coming months. The interim results in late August 2026 will show whether the Q1 profit decline persisted. Continued disclosure on Oceanwide enforcement recoveries will indicate how much of the written-down exposure is economically recoverable, given that two of the seven judgments have already been terminated for enforcement. And whether New Hope's stake-building resumes toward the 5% threshold and board representation will indicate whether founder re-engagement is a passive value bet or the opening move in something larger.

---

↑ Back to Top

XI. Durable Lessons & Epilogue

Thirty years is long enough for an institution to run a complete experiment, and Minsheng's produced results worth generalising.

The first lesson concerns the founding design itself. Fragmenting ownership among 59 companies solved the problem the regulators feared — no single tycoon could turn the bank into a personal treasury — and created a problem they apparently did not anticipate. Dispersed ownership diffused capture rather than preventing it. When a board seats nine shareholder-directors whose groups all borrow from the institution they govern, the resulting equilibrium is not discipline; it is mutual accommodation. Dispersed ownership is not automatically clean ownership. What prevents insider lending is not the shape of a shareholder register but the independence of a credit function and the willingness of a board to say no to its own members. Minsheng's own related-party disclosures, three decades on, still list shareholder-affiliated borrowers by name.

The second lesson is older than China's banking system. Lending to your own shareholders' business empires is one of the recurring failure modes in banking history, from the 19th-century American unit banks to the Japanese keiretsu, to the Icelandic and Irish institutions of 2008. It fails the same way every time: the loans look best exactly when the collateral is most inflated, and the relationship that made the credit easy to approve makes it impossible to enforce. Minsheng's Oceanwide and Orient Group experience is a fresh, unusually well-documented case study — with the added detail that the bank ultimately had to sue a founding shareholder and win judgments that turned out to be largely uncollectable.

The third lesson is for the deep-value investor, and it is the hardest. Distinguishing a genuine return inflection from a temporary lull in bad news is the central skill of bank investing, and it cannot be done from the valuation alone. A price-to-book of 0.27 says nothing about which of the two is happening. Only the credit cost trend does. Minsheng's FY2025 and Q1 2026 numbers are an almost laboratory-grade illustration: revenue turned, costs improved, the problem sector healed, ratings improved — and earnings still fell, because provisions rose faster than everything else. The market's discount is not obviously wrong. It is a bet on which way the provisions go next.

And there is the epilogue, which no one designed.

A bank was founded in 1996 to prove that private capital could run a commercial institution better than the state. It listed in Shanghai, then set a record in Hong Kong. Its founders fought over its board for two decades, one of them spending a billion dollars in four days to hold his seat. It lent aggressively into Chinese property, became the largest bank creditor to the developer whose collapse defined the cycle, and was for a period the worst-performing bank stock on earth. Its president was taken by anti-corruption investigators. A branch manager stole nearly RMB2.75 billion from its private-banking clients. Its former chairman rebuilt the same 59-shareholder template outside the banking system and watched it default. It sued one of its own founders and won.

And today, thirty years after being launched as an experiment in private capital, the largest shareholder of China's first private bank is a state insurance vehicle that exists only because a different institution's owner was imprisoned for fraud. The bank's strategic positioning still reads "a bank for non-state-owned enterprises." The words are unchanged. Almost nothing behind them is.

---

References

1. Results Announcement for the Year Ended 31 December 2025 (containing the full 2025 Annual Report) — China Minsheng Banking Corp., Ltd., 2026-03-30 ↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩↩

2. Top 10 Ordinary Shareholders (as at 31 March 2026) — China Minsheng Banking Corp., Ltd. Investor Relations ↩↩↩↩

3. 2026 First Quarterly Report — China Minsheng Banking Corp., Ltd., 2026-04-29 ↩↩↩↩↩↩

4. Founder of China's Private Bank Confident About Future — People's Daily Online, 2003-01-12 ↩↩↩↩↩↩

5. 民生银行6年董事长退出"一群最强男人的战争" — 腾讯新闻 QQ News, 2020-07-03 ↩↩↩↩

6. Minsheng Bank raises $3.89 billion in largest Asian IPO this year — FinanceAsia, 2009 ↩

7. Minsheng Bank President Resigns Amid Corruption Investigation — ChinaFile / Caixin Media, 2015 ↩

8. Ex-Minsheng Chairman Takes Helm of Private Capital 'Aircraft Carrier' — Caixin Global, 2014-08-22 ↩

9. Investment Group CMIG to Miss $500 Million Debt Repayment — Caixin Global, 2019-07-20 ↩↩

10. World's Worst-Performing Bank Lent Billions to China Evergrande — Bloomberg, 2022-01-11 ↩↩

11. Evergrande Major Creditor Minsheng Says It Has Cut Exposure — Bloomberg, 2021-06-24 ↩

12. China Evergrande's creditor Minsheng Bank pares loans to developer amid concerns about leverage, default — South China Morning Post, 2021 ↩

13. Developer Oceanwide and Tycoon Lu Zhiqiang Sued in $1 Billion Loan Dispute — Caixin Global, 2023-01-23 ↩

14. Property Firm China Oceanwide Forced to Wind Up Operations by Bermuda Court — NTD, 2023-09 ↩

15. Minsheng Bank Says Exposure to Troubled Shareholder Manageable — Caixin Global, 2024-06-28 ↩

16. China Minsheng Bank Drops USD6.9 Billion Convertible Bond Plan — Yicai Global, 2023-08 ↩

17. China's Minsheng Bank sees investor trust vanish – along with 3b yuan — South China Morning Post, 2017 ↩↩

18. 民生银行一支行行长"自编自导"惊天大案 "假理财"诈骗147名投资人27亿元 — 新浪财经 Sina Finance, 2020-12-14 ↩

19. China's major banks fined for violations of laws and regulations — Global Times, 2023-02-18 ↩

20. 315金融观察｜两个月被罚超2400万元，民生银行登榜"罚单之王" — 新浪财经 Sina Finance, 2025-03-19 ↩

21. 民生银行被罚5865万：因相关贷款、票据等业务管理不审慎等 — 腾讯新闻 QQ News, 2025-10-31 ↩↩

22. 因业务管理不审慎 中行、民生等五家银行领千万元以上罚单 — 财新网 Caixin, 2025-11-02 ↩

23. Bank of China Executive Gao Yingxin May Lead China Minsheng Banking — Caixin Global, 2020-05-26 ↩

24. 新希望拟举牌民生银行 或成该行第四大股东 — 21经济网 21st Century Business Herald, 2024-11-19 ↩↩

25. 出手民生银行！憋了18年，刘永好终于反击 — 新浪财经 Sina Finance, 2024-11-20 ↩↩↩

26. Takeover unveiled for insurance giant — China Daily, 2019-07-12 ↩↩

27. China Merchants Bank 2025 Annual Results Highlights — China Merchants Bank Co., Ltd., 2026-03-27 ↩↩↩↩↩↩↩

28. China Minsheng Bank (600016) share price and valuation data — Investing.com, 2026-08-11 ↩↩↩↩

29. China Merchants Bank share price and valuation data — Investing.com, 2026-08-11 ↩

30. Fitch Upgrades China Minsheng Banking to BBB- on Rising Importance — Futu News, 2025 ↩

31. 2025 Annual Results Announcement (investor presentation) — China Minsheng Banking Corp., Ltd., 2026-04-03 ↩↩↩

32. Development Strategy — China Minsheng Banking Corp., Ltd. Investor Relations ↩↩