NextFin News - Asia's banks are reporting some of their best numbers in years, and investors are being asked to believe the worst is behind them. China's five largest lenders just posted their biggest quarterly profit increases in recent years, Japan's megabanks are running on record earnings, and Singapore's big three kept profits rising even as net interest margins compressed. But the earnings season now sweeping the region is delivering a less comfortable message: the peak has passed, loan demand is softening, and credit costs are starting to climb.
The divergence is the story. Reported profits are strong because they reflect rate hikes and margin recovery that happened months ago. Forward-looking commentary from bank executives tells a different tale — slowing credit demand, more selective lending, and provisions moving in the wrong direction. When earnings look backward and guidance looks forward, the forward view wins.
The Numbers Look Strong — Until You Ask What's Driving Them
On the surface, the third quarter of 2026 was kind to Asia's lenders. China's mega banks reported their largest quarterly profit increases in recent years as margins recovered from record lows and bad-loan discipline held. Industrial and Commercial Bank of China, the nation's largest lender, saw third-quarter profit rise 10% to 111.96 billion yuan, the biggest gain in five years. Agricultural Bank of China added 11%, Bank of Communications 10.1%, China Construction Bank 9.8%, and Bank of China 6.4%.
In Japan, the three megabanks posted record profits for the fiscal year ended March 2026 — Mitsubishi UFJ Financial Group at 2.4 trillion yen, up 30%; Mizuho Financial Group at 1.25 trillion yen, up 41%; and Sumitomo Mitsui Financial Group at 1.58 trillion yen — and they are forecasting further gains. MUFG expects a 12.5% increase in net profit for the current business year, and Mizuho is targeting a 4% rise, even as tariff uncertainty clouds the outlook.
Singapore's big three — DBS Group Holdings, Oversea-Chinese Banking Corp., and United Overseas Bank — delivered combined first-half net profit up 7.1% year-on-year, even as squeezed net interest margins erased S$520 million in lending income. Fee businesses carried the load: DBS saw wealth assets under management cross S$500 billion for the first time, while at OCBC trading income surged 85% year-on-year to a record S$695 million and wealth management fees rose 44% to S$470 million in the second quarter. Non-interest income now accounts for more than 40% of revenue at both DBS and OCBC.
So what is the problem? The profit momentum is backward-looking. It was generated by a rate-hiking cycle that has already peaked across most of Asia, by a one-off recovery in Chinese net interest margins from distressed lows, and by a temporary shift of corporate funding from bond markets back onto bank balance sheets. Each of those drivers is mean-reverting. The question is not whether earnings were good — they were. The question is whether they are repeatable.
The Transmission Mechanism: Why the Roll-Over Is Built In
The mechanism running under the surface is straightforward, and it operates through three channels.
First, net interest margins have limited room to expand. Most of Asia's rate-cutting cycle is already concluded, according to CreditSights' APAC financials outlook, which means banks will see only limited further gains from their investment books. In Singapore, the compression is already visible in the reported numbers: OCBC's net interest margin fell 22 basis points year-on-year to 1.70% in the second quarter, while DBS and UOB reported half-year margins of 1.96% and 1.91% respectively, down 4 and 13 basis points. UOB's net profit fell 28% year-on-year to S$3.27 billion. Margins are not collapsing, but the direction of travel is down, and deposit competition keeps the pressure on.
Second, loan growth is slowing and changing quality. China recorded its slowest first-half loan growth since at least 2019 in the first half of 2026, and much of the expansion at the big state banks has been driven by discounted bills rather than longer-term loans tied to real economic activity. That is a warning sign: bill financing is short-duration, low-yield, and often used by companies to manage liquidity rather than to invest. New yuan loans came in at 1.08 trillion yuan in September 2026, below the 1.29 trillion yuan extended a year earlier, while outstanding loan growth slowed to 4.8% year-on-year. ICBC expects full-year loan growth of around 6%, a modest figure for the world's largest bank by assets. In Hong Kong, HSBC's S$7 billion quarter-on-quarter loan growth came mainly from retail mortgages — a lower-yielding, rate-sensitive segment — rather than corporate expansion.
Third, credit costs are moving up from historically low levels. In Singapore, analysts project third-quarter credit costs of 17 basis points for the sector, up from 10 basis points in the second quarter, with management guidance for specific provisions in the 17-to-20 basis-point range. SMFG is guiding to 300 billion yen in total credit costs for the current fiscal year, a level that keeps provisioning elevated even as it steps back from the prior year's peak. CreditSights flags deteriorating asset quality in specific pockets: Indian non-bank unsecured retail stress, Hong Kong commercial real estate exposure, and elevated provisioning at Thai banks. Credit costs for 2026 are expected to be similar to 2025 — which is to say, no longer a tailwind.
The sequence matters. Margin expansion peaked first. Loan demand is softening now. Credit costs lag both, because provisions are recognized only after borrowers show stress. That lag is why the earnings still look good while the commentary turns cautious. It is also why the next two quarters matter more than the last two.
What the Executives Are Saying — and Why It Matters More Than the Headlines
The most telling evidence comes from the banks' own forward-looking statements. Mizuho's chief executive, Masahiro Kihara, acknowledged at a press briefing in Tokyo:
Mizuho's pipeline of equity capital markets and mergers and acquisitions deals is currently backed up as shifting business conditions make decision making on large-scale investment challenging.
That is a direct signal that the fee engine many investors are counting on to replace margin income is itself running into headwinds. It also undercuts the cleanest part of the bull case: if deal activity is backing up, the advisory and ECM fees that powered the record year cannot be counted on to accelerate.
MUFG chief executive Hironori Kamezawa has described the bank's main scenario for Japan as one in which solid consumer consumption and inbound tourism continue to support a mild economic recovery. But even that "mild recovery" framing is a far cry from the expansion narrative that accompanied the record profits. SMFG's chief executive, Toru Nakashima, said the group's 1.2 trillion yen exposure to private credit drew no great concern because it dealt primarily with high-quality funds — but the very fact that the number was volunteered at an earnings briefing shows how closely investors are scrutinizing the riskier corners of the loan book.
HSBC, reporting third-quarter results that missed revenue and earnings expectations, maintained its full-year 2026 credit guidance at around 45 basis points and described Hong Kong commercial real estate charges as very low for the quarter. Holding guidance while wealth management underperforms and revenue misses is not a red flag by itself, but combined with the regional pattern it reinforces the picture: banks are holding the line on credit for now, but they are not expanding it.
Cyclical or Structural: The Call That Determines the Conclusion
Is this a cyclical dip or a structural shift? The answer is both, and separating them is essential.
The cyclical leg is the rate cycle. Asia's banks are classic rate-cycle beneficiaries, and the cycle has turned. Net interest margin expansion from higher rates is a one-time repricing, not a permanent step-up in profitability. History supports this: every regional rate-hiking cycle since the 1990s has produced the same pattern — margin expansion, record profits, then compression as rates peak and credit costs normalize. The mean-reversion pattern is the rule, not the exception. The short-term driver is clear: rates have peaked, deposit competition is intensifying, and loan growth is slowing. On a 12-to-18-month horizon, profit growth will decelerate from current levels.
The structural leg is more important and less priced in. Asia's banks are being forced to change their business model. Net interest income is no longer a reliable growth engine in a structurally lower-rate, slower-growth regional economy. The banks that win are the ones successfully pivoting to fee-based businesses — wealth management, transaction banking, markets and advisory. DBS is the clearest example: with wealth AUM past S$500 billion and a target of S$1 trillion by 2030, it is building an asset-light income stream that does not depend on the spread between loans and deposits. OCBC's group wealth income rose 27% year-on-year to S$3.29 billion in the first half, now accounting for 41% of total income. This is a durable shift, and it will separate winners from losers over the next five years.
But the structural pivot does not offset the cyclical roll-over in the near term. Fee income is growing from a smaller base and cannot fully replace the margin dollars coming under pressure. The two forces point in opposite directions across different time horizons: cyclical headwinds dominate the next four to six quarters, while structural differentiation determines the five-year outcome.
The Counter-Thesis — and What Would Prove It Right
The strongest case against this reading is simple: China's economy is stabilizing, and if it does, the banks' asset quality will hold and loan demand will recover faster than expected. Beijing is injecting fresh capital into the state lenders — the Ministry of Finance is selling 300 billion yuan in special bonds to replenish core Tier-1 capital at eight institutions, following a 520 billion yuan recapitalisation in 2025 — and fiscal-financial coordination policies are expected to accelerate investment from the fourth quarter into next year, according to ICBC management. If Chinese loan growth re-accelerates toward 8% to 10% and corporate demand returns, the regional banks' earnings revisions would turn positive again, because China remains the marginal driver of Asian credit demand.
That counter-thesis is credible, and it is the scenario that would falsify the "tougher times" view. The specific signal to watch is China's monthly new bank loan data and the composition of that growth. If new loans grow at a double-digit annualized pace for two consecutive months and the share of medium- and long-term corporate loans rises above 60% of the total — rather than bill financing dominating — the cyclical downturn thesis is wrong. Until that data prints, the burden of proof sits with the optimists.
What to Watch Next
The immediate catalyst is the rest of the earnings season. DBS reports third-quarter results on November 5, 2026, and investors will be watching three numbers: net interest margin trajectory, loan growth composition, and credit cost guidance for 2027. The UOB Kay Hian preview expects sector credit costs of 17 basis points in the third quarter, up from 10 basis points, with DBS holding a S$2.4 billion management overlay for general provisions — a cushion that signals management is preparing for a less benign environment. OCBC, meanwhile, has raised its loan growth guidance to high single-digit to low double-digit from mid-single-digit, while expecting credit costs at the lower end of its 20-to-25 basis-point range — evidence that even within the slowdown, the best-franchised banks are still finding volume.
Across the region, the watch list is:
- China: monthly new bank loan data, especially the split between bill financing and medium-to-long-term corporate loans; whether the fiscal stimulus package translates into real credit demand by the fourth quarter.
- Japan: whether the megabanks can keep credit costs near current guidance while US leveraged lending and commercial real estate remain under stress, and whether backed-up fee pipelines clear in 2027.
- Singapore: whether fee income growth can outpace net interest margin compression as the rate-cut cycle deepens.
- Hong Kong: commercial real estate asset quality, given the exposure flagged across regional banks.
Scenarios for the next 12 months: the base case is mid-single-digit profit growth for the region's best-franchised banks, with margin compression partially offset by fee growth and stable credit. The upside case requires a China credit rebound that re-accelerates loan demand and keeps provisions benign. The downside case is a sharper-than-expected slowdown in China plus a US recession that hits leveraged lending and commercial real estate books simultaneously — in which case credit costs could run well above the 40-to-50 basis-point range that banks are currently guiding.
The earnings are not bad. They are late. Asia's banks are reporting the profits of a cycle that has already peaked, while the executives standing behind the numbers are quietly preparing for the one that comes next. The market will find out which view was right not in the next earnings release, but in the loan data that arrives every month between now and then.
Explore more exclusive insights at nextfin.ai.
