Futures
Access hundreds of perpetual contracts
TradFi
Gold
One platform for global traditional assets
Options
Hot
Trade European-style vanilla options
Unified Account
Maximize your capital efficiency
Demo Trading
Introduction to Futures Trading
Learn the basics of futures trading
Futures Events
Join events to earn rewards
Demo Trading
Use virtual funds to practice risk-free trading
Launch
CandyDrop
Collect candies to earn airdrops
Launchpool
Quick staking, earn potential new tokens
HODLer Airdrop
Hold GT and get massive airdrops for free
Pre-IPOs
Unlock full access to global stock IPOs
Alpha Points
Trade on-chain assets and earn airdrops
Futures Points
Earn futures points and claim airdrop rewards
Traversing the "Low Profit" Era: The Deep Structural Changes Behind the 2025 Annual Reports of the Six Major Banks
Ask AI · Why has net interest margin become a structural challenge for bank profitability?
With 3.5 trillion yuan in revenue, 1.42 trillion yuan in net profit, total assets exceeding 220 trillion yuan, dividends over 420 billion yuan, and both revenue and net profit maintaining positive growth—China’s six largest state-owned banks’ 2025 annual reports deliver a “steady progress” performance. Industrial and Commercial Bank of China (ICBC) with total assets of 53.48 trillion yuan has also become the world’s first commercial bank to cross the 50 trillion yuan mark, reaching a new historical high in scale.
However, behind the impressive figures, a profound shift is quietly underway. Under the surface appearance of double growth in revenue and profits lies a reshaping of the earnings model, a shift in risk focus, and a change in growth logic.
The old model is under pressure: Why has net interest margin become an “irreversible” fate?
In 2025, the net interest margins of the six major banks all declined. Postal Savings Bank remained the top performer with a net interest margin of 1.66%, followed closely by China Construction Bank at 1.34%. ICBC and Agricultural Bank both recorded 1.28%, Bank of China 1.26%, and Bank of Communications fell to 1.20%, the lowest among the six.
Behind this series of numbers lies a structural challenge that cannot be avoided.
The downward movement of loan interest rates caused by multiple adjustments to the LPR is the primary driver for compressing net interest margins. In 2025, the weighted average interest rate on incremental loans fell by about 30 basis points from the beginning of the year, and the repricing effect of outstanding loans further eroded interest income. Meanwhile, deposit costs showed “rigid” characteristics. Although the listed deposit rates were cut multiple times, residents’ savings appetite remained strong; the proportion of time deposits continued to rise, and the decline in cost on the liability side was much slower than the pace at which asset-side returns fell.
This trend is not a cyclical fluctuation, but the inevitable result of structural transformation. As interest rate marketization reform deepens, and the financing costs of the real economy continue to be reduced, the era when banks could rely on “interest spread” to get by is already gone and will not return.
At the earnings release conference, multiple senior bank executives responded to the outlook for 2026 interest margins and management strategies.
Postal Savings Bank President Lu Wei said, “As for this year, with the central bank’s current symmetric rate cuts, the strengthening of the self-discipline mechanism, and the measures from the State Administration for regulating unfair competition, we can clearly feel that efforts from outside are being made at the same time, and their role in stabilizing net interest margins is obvious. For Postal Savings Bank, we will continue to intensify management, consolidate our leading advantage, and at the same time focus on helping net interest margins settle and stabilize as soon as possible. At the branch level too, we need to strengthen assessments, press down responsibilities, and drive branch deposit-and-loan interest margin to stabilize quickly.”
ICBC Deputy President Yao Mingde expects that in 2026 the interest margin will most likely follow an “L-shaped” pattern. “In 2025, our net interest margin’s downward trend gradually narrowed, with the year-on-year decline shrinking by 5 BP. Although it is still declining, the rate of decline is slowing, and we believe this trend is sustainable. If we do not consider the LPR and the listed deposit rates being adjusted further significantly, we expect our net interest income this year to turn positive year-on-year, bringing an inflection point, and the magnitude of the net interest margin decline will further converge compared with 2025.”
Bank of Communications Deputy President Zhou Wanfu said that the bank will keep net interest margins stable and improve them from the following three aspects: first, strictly conduct deposit-and-loan quantity and price balance management; second, implement deposit-and-loan pricing management in a more refined way, and strictly comply with the requirements of the pricing self-discipline mechanism; third, scientifically optimize the asset-liability structure, and dynamically adjust based on factors including liquidity and interest rate risk.
Bank of China Deputy President Liu Chenggang expects that Bank of China’s net interest margin year-on-year decline will narrow substantially. “There are multiple uncertainties in the external environment. Changes in international geopolitics have squeezed the room for major currencies to cut rates, and domestic banks are still facing a low-interest-rate environment. Bank of China is confident in seizing the market opportunities brought about by the implementation of a package of incremental policies, fully leveraging its global strengths and comprehensive capabilities, and solidly achieving an overall balance of ‘volume, price, risk, and efficiency,’ further enhancing operational resilience and sustainable development capacity.”
A game of new drivers: Does a high proportion of non-interest income equal a successful transformation?
As the space for traditional interest income continues to be compressed, non-interest income is becoming an important support for bank profitability.
In 2025, ICBC’s non-interest income was 203.144 billion yuan, up 10.2% year on year. Postal Savings Bank’s net fee and commission income rose 16.15% year on year, and other non-interest net income increased 19.73%. Bank of China’s non-interest income ratio reached 33.06%, a record high, with standout contributions from wealth management, settlement and clearing, and financial market transaction business. Bank of Communications’ non-interest net income grew 2.22% year on year. Agricultural Bank achieved net fee and commission income of 88.085 billion yuan, up 16.6% year on year. Among this, agency business grew 87.8%. The bank explained that this was mainly due to the deepening of its wealth management business transformation, which increased income from wealth management products and fund distribution.
But when broken down: Is non-interest income growth truly a “real transformation,” or just a “cover-up”?
From a structural perspective, non-interest income growth mainly comes from three levels. The first is wealth management business, including wealth management products, fund distribution, insurance agency, etc. This business has high value and strong sustainability, but it is sensitive to market conditions. The second is card business. Due to weak consumer activity, its growth space is limited. The third is basic businesses such as settlement and custody, which are relatively stable but lack growth elasticity.
At a deeper level, the question is: How much of the current growth in non-interest income has truly escaped the “derivative” nature of credit business? How many banks still heavily rely on cross-selling wealth management products to credit customers? If credit business were separated out, just how strong would independent, market-oriented wealth management capability be?
From international experience, global banks such as JPMorgan Chase and HSBC generally have non-interest income ratios above 40%, while domestic large banks still hover around 20%. This means that the transformation from a “credit bank” to a “comprehensive financial services platform” still has a long way to go.
Risk replaces the anchor: Why has retail credit shifted from “ballast stone” to “storm eye”?
This is the most worrying signal in the 2025 annual reports: bank risks are shifting from traditional corporate areas toward personal retail business.
Postal Savings Bank’s non-performing loan ratio rose to 0.95%, and the proportion of loans under watch clearly increased, reflecting that the accumulation of retail credit risk is not an isolated case.
“Personal loans are indeed a pressure point for our asset quality. The non-performing loan ratio is 1.42%, and at the same time, our personal loan proportion is high, at 50.2%.” Postal Savings Bank Deputy President Yao Hong said, “In 2025, our ‘three rates’—non-performing, special mention, and overdue—each rose year on year, by 0.05, 0.62, and 0.11 percentage points respectively. Among them, the increase in the watch category rate is relatively larger, mainly because we carried out targeted relief and renewal loans for customers with willingness to repay but temporary repayment difficulties. We prudently classified this portion of loans into the watch category.”
By the end of 2025, ICBC’s personal non-performing loans amounted to 142.337 billion yuan, an increase of 39.510 billion yuan. The personal non-performing loan ratio was 1.58%, up 0.43 percentage points. Among them, the non-performing loan ratio for personal housing loans rose from 0.73% to 1.06%, and the non-performing loan ratio for personal consumer loans rose from 2.39% to 2.58%.
China Construction Bank’s personal loan and loan advances non-performing loan ratio rose from 0.98% last year to 1.19%, and Bank of Communications’ personal loan non-performing loan ratio rose from 1.08% last year to 1.58%.
Why has risk migrated to the retail side? On the macro level, unstable household income expectations and a stressed employment market directly affect repayment ability. On the micro level, in recent years, practices such as using business loans to replace mortgage loans are exposing legacy issues during the real estate market adjustment cycle. In addition, problems such as credit card joint-liability risks and multi-borrower credit approvals for consumer loans have also accelerated in exposure during the economic downturn cycle.
What should be done in the future? At the earnings meeting, China Construction Bank Deputy President Li Jianjiang responded: “Facing the situation of rising risks in the retail sector in recent years, the bank has made great efforts to optimize the retail credit risk management mechanism, strengthen risk checks and balances at key points in the credit process, promote centralized and intensive risk control for retail credit, and the multiple risk management measures in 2025 have shown results—resulting in a narrower year-on-year increase in the non-performing loan ratio for personal loans.” Li Jianjiang said that, judging from the current operating trend, risk prevention and control in the retail sector will still be one of the key focuses of its work.
ICBC Deputy President Wang Jingwu also said that ICBC’s personal loan asset quality has long been relatively strong, but in the past two years, due to multiple factors—including economic transformation and growth, real estate market adjustment, and temporary imbalances between supply and demand—its non-performing loan ratio has entered a short-term upward channel, basically consistent with the trend across the industry. However, as a package of policies accelerates implementation and policy dividends continue to be released, the foundation of the personal credit market will gradually improve, and the asset quality of personal loans will return to a reasonable level.
“To respond to market changes, ICBC has already made corresponding adjustments in its internal structure and functions in advance. It established a Personal Credit Business Department to realize centralized and specialized personal loan business and further improve operating standards. At the same time, we strengthened digital intelligence, enriched product innovation in personal consumer finance, coordinated and balanced development with safety, and focused on resolving various potential risks and hidden issues, while solidly handling the disposal of non-performing assets.” Wang Jingwu said.
Technology empowerment: The key variable of the future
Worth noting is that in their 2025 annual reports, all of the six major state-owned banks focused on disclosing their application progress of AI technologies, embedding AI capabilities deeply into the entire business process.
Bank of China’s annual report shows that in 2025, based on its three major platforms—computing power, technology, and data—it built two sets of agile, efficient, secure, and reliable AI-enabled governance mechanisms. It developed six typical application paradigms such as intelligent Q&A and report generation, deployed series of large models including DeepSeek and Qwen3, and built more than 400 intelligent assistants. It achieved deep empowerment in key areas such as credit, marketing, operations, office, customer service, and technology.
At the earnings release conference, Bank of China President Zhang Hui said it will further build an “AI+” financial ecosystem. China Construction Bank also systematically advanced the construction of AI applications; related technologies have been scaled to empower 398 scenarios within the Group, deeply penetrating key areas such as wealth management, inclusive finance, risk management, and technology R&D.
ICBC disclosed in its annual report that it innovatively launched the “Leading AI+” initiative, rolling out more than 500 AI applications across over 30 business areas. Its AI digital employees shoulder 55,000 work tasks per year. Meanwhile, ICBC keeps pace with technological development; based on Industrial and Commercial Bank of China’s (ICBC’s) “Zhiyong” platform, it explores and establishes an intelligent agent collaboration system featuring “one master, multiple specialists.” ICBC said it will align with the trend of technological change, seize the opportunities of “AI+,” continuously strengthen the driving force of digital intelligence, and deepen the digital and intelligent transformation of business management and risk governance.
At the earnings release conference, Agricultural Bank President Wang Zhiheng said that Agricultural Bank is firmly grasping the wave of AI technology development. It has specifically set up an office for building a smart bank, increased the overall coordination and advancement of smart bank construction, and also clarified that it will use intelligent agent applications as a starting point and project needs as the driving force, continuously improve the “AI+” capability system, and focus on promoting intelligent and inclusive applications of AI.
Agricultural Bank Deputy President Lin Li, when discussing risk control measures, also said that the bank is currently strengthening technology empowerment, expanding new risk control capabilities, and launching the Agricultural Bank version of “Lobster.” “This is not about chasing trends. We use this tool to automatically process and analyze data, intelligently generate due diligence reports, and make the lending process more convenient, more efficient, and more secure.”
Three judgments: Long-termism amid slow variables
The era of the next five years of meticulous cultivation has already begun. Based on an in-depth analysis of the annual report data of the state-owned large banks, three clear judgments can be made:
Net interest margin has entered the bottom area of the “1% era,” but rebounds lack strength. As the LPR becomes stable and the repricing of deposit costs gradually completes, there is limited room for net interest margins to keep falling significantly. However, given that the financing costs of the real economy still need to be reduced, the margin also cannot show a trend-like rebound. Banks must accept that “thin profits” will become the norm, and, within this norm, find a way to survive.
The core of risk management will shift from “credit risk” to “customer group management capability.” Competition in retail is essentially a competition of scenarios, data, and risk control models. Whoever can accurately identify high-quality customer groups, effectively manage credit risk, and establish differentiated pricing capabilities will take the initiative in the retail market. Retail credit competition is no longer a contest of scale expansion, but of customer group management.
Bank valuation logic will be thoroughly re-evaluated. The market will no longer pay for a bank simply because it is “big,” but will price for “the ability to get through cycles.” Whoever can first prove that it has moved beyond dependence on net interest margins, controls retail risks, and builds a moat of non-interest income will obtain valuation premiums. In the future, differentiation among banks will show not only in operating data, but also in capital market valuations.
When scale is no longer synonymous with safety, when net interest margin is no longer the profit shield, when retail changes from a sweet fruit to a thorny rose, and when AI becomes a key variable, bankers need to answer a fundamental question: what kind of financial institution do we actually want to become?
Should we continue to fight it out in the red ocean of traditional credit, or truly transform into a financial services provider capable of mastering cycles, managing risks, and creating value?
The six major banks have given their own phased answers for their respective stages. But the true endgame will unfold gradually in the slow variables of the future.