China Minsheng Banking Additional Discussion Report: Credit Cost, CET1, and Capital Policy Review Based on SSC Discussion

• Report date: 2026-06-04
• Issuer / Theme: China Minsheng Banking Corp., Ltd. / Credit impairment, CET1, low ROE, capital instruments, and ongoing review of management policy
• Report type: additional_discussion
• Discussion scope: Summary of the five main questions and each follow-up question covered in the SSC discussion on 2026-06-04
• Reference context: Issuer summary dated 2026-05-18, existing notes on the issuer, and SSC discussion on 2026-06-04

1. Purpose and Treatment

This report is a supplemental report that organises the SSC discussion on China Minsheng Banking Corp., Ltd. (hereafter, China Minsheng Bank, or the bank) in the context of the existing issuer summary. It is not new primary research or an update to the main body of the existing report. Its purpose is to separately preserve the hypotheses raised in the discussion, points already confirmed in the existing report, and unresolved items for future follow-up.

Accordingly, this report is not a final investment view. In particular, references to a move in the CET1 ratio toward 9%, market reaction in AT1 and Tier 2 instruments, and deterioration in the rating outlook are early-warning frameworks used in the discussion. They do not represent a formal outlook change by rating agencies or an investment recommendation on any individual security.

2. Context Confirmed in the Existing Report

The existing issuer summary characterises the bank as a credit that is “systemically important, but with limited financial headroom.” Its D-SIB designation, deposit base, LCR, and scale as a nationwide bank support its senior credit profile. At the same time, low ROE, heavy credit impairment losses, a CET1 ratio in the low-9% range, and asset-quality issues in credit cards, MSE, real estate, and certain corporate sectors remain constraints.

The main confirmed figures are as follows. Credit impairment losses in 2025 were RMB53.950bn, substantially exceeding net profit attributable to shareholders of the parent of RMB30.563bn. ROE was 4.93% in 2025, while the CET1 ratio was 9.38% at end-2025 and declined slightly to 9.35% at end-March 2026. The NPL ratio was 1.49% at end-2025 and 1.46% at end-March 2026, but special-mentioned loans increased from RMB121.195bn at end-2025 to RMB124.538bn at end-March 2026.

By loan category, the NPL ratio at end-2025 was 3.87% for credit cards, 1.63% for personal loans to MSEs, 3.61% for corporate loans to real estate, 2.27% for wholesale and retail, and 0.82% for leasing and commercial services. The NPL ratio for corporate loans to real estate remains high, although it improved from 5.01% at end-2024. In contrast, deterioration was notable in credit cards, wholesale and retail, and leasing and commercial services. The discussion therefore focused on this “non-real-estate deterioration path.”

3. Summary of Q&A

3.1 Asset Categories Likely to Push Up Credit Impairment First

Intent of the question. The first question was intended to test whether the bank’s credit deterioration scenario should be assessed not only through a “renewed real estate deterioration” lens, but also through the possibility that persistently high credit costs in credit cards, personal loans to MSEs, wholesale and retail, and leasing and commercial services are becoming the main route that prevents CET1 accumulation. The purpose was to identify which asset categories could push up credit impairment most quickly over the next one to two years and are most likely to feed through to CET1 ratio decline or deterioration in the rating outlook.

Main points of the answer. In the discussion, corporate loans to real estate were still regarded as important. However, as of 2025, the sources of deterioration more likely to flow quickly into earnings were identified as the combination of credit cards, wholesale and retail, leasing and commercial services, and personal loans to MSEs. Credit cards were viewed as the category most likely to show up first in credit impairment, as the NPL amount increased despite a decline in balances and the NPL ratio rose to 3.87%. Wholesale and retail, and leasing and commercial services, were treated as non-real-estate sources of deterioration on the corporate side. For personal loans to MSEs, the focus was less on the increase in the NPL amount and more on the fact that the NPL ratio did not improve despite balance contraction.

Corporate loans to real estate cannot be ignored given the high absolute NPL ratio of 3.61%. However, because the NPL ratio improved and the NPL amount declined in 2025, at least in this discussion the category was treated not as “the main driver of the first wave of new deterioration,” but as an “existing problem that continues to linger and would amplify losses if it deteriorates again.”

Points explored in the follow-up. The follow-up question asked whether, if these asset categories deteriorate at the same time, the bank could still absorb a normal increase in credit costs through provisions, or whether it would enter a stage where a move toward a CET1 ratio below 9% and deterioration in the rating outlook should be considered. The answer was that the assessment needs to be staged using a combination of credit impairment losses, special-mentioned loans, NPL formation, CET1, and ROE, rather than a single indicator.

The normal absorption range is a state in which the NPL ratio is around 1.5%, special-mentioned loans are in the 2.7%–2.8% range, credit impairment losses pressure earnings but CET1 remains around 9.3%. The watch range is a state in which special-mentioned loans clearly increase, the NPL ratio rises into the 1.6% range, credit impairment losses remain elevated above RMB55bn–60bn per year, and ROE returns to below 5%. The warning range was defined as a state in which the NPL ratio exceeds 1.7%, special-mentioned loans move to around 3% or above, credit impairment losses increase toward RMB60bn, and CET1 declines toward around 9%.

Credit analytical implications. The implication of this Q&A is that the bank’s credit deterioration should not be assessed only by whether real estate NPLs have jumped. If credit cards, MSE, wholesale and retail, and leasing and commercial services deteriorate simultaneously, special-mentioned loans and credit impairment losses increase, and CET1 declines toward 9%, the issue should be viewed not as short-term liquidity risk but as a problem of loss-absorption capacity and internal capital generation under low ROE.

Unresolved items. NPL formation, delinquencies, write-offs, recoveries, and the breakdown of credit impairment by product and sector from 2026 onward cannot be sufficiently confirmed from quarterly materials alone. Which combination of indicators rating agencies would use as a direct trigger for an outlook change also needs to be further checked against the full text of the latest primary releases.

3.2 Whether Low ROE Is a Temporary Credit Cost Issue or a Structural Decline in Internal Capital Generation

Intent of the question. The second main question was intended to assess whether the bank’s low ROE can be explained only by temporary credit cost factors, or whether internal capital generation has structurally weakened because of lower NIM, deposit costs, risk-asset contraction, and weak fee income. The focus was whether the bank has become one in which ROE does not recover sufficiently even if credit costs normalise to some extent, making it difficult to accumulate CET1 through retained earnings.

Main points of the answer. In the discussion, it was considered risky to attribute low ROE solely to credit costs. Operating income increased in 2025, but final profit was consumed by credit impairment. NIM was low at 1.40%, and although it appeared to improve slightly to 1.43% in the first quarter of 2026, the improvement in 2025 was mainly supported by lower deposit costs. If the support came from lower liability costs rather than a recovery in average loan yield, then the bank’s capacity for NIM recovery after the room for further cost reduction narrows remains unconfirmed.

In addition, net fee and commission income was not growing strongly, and part of non-interest income was viewed as sensitive to market conditions and investment income. Therefore, even if credit costs decline somewhat, if NIM returns to a path below 1.4%, fee income remains weak, and ROE stays below 5%, the bank’s ability to organically accumulate CET1 will be limited.

Points explored in the follow-up. The follow-up question organised the measures the bank could take to protect ROE and CET1 if the room for further deposit cost reduction narrows. Realistic defensive measures identified included controlling RWA growth, carefully rebalancing the loan mix, reducing credit costs, restoring stable fee income, and, if necessary, restraining dividends. In contrast, a strategy of raising ROE by returning to higher-yield credit card, MSE, and private enterprise loans was considered helpful for short-term earnings, but with the side effect of increasing future credit impairment again.

A shift toward low-risk assets and high-quality corporates is positive for asset quality and CET1, but tends to slow ROE recovery. Conversely, renewed expansion into higher-yield assets may support reported ROE, but could amplify the credit cost pressure already visible in credit cards and MSE. The bank was therefore viewed as facing a dilemma between “accepting low ROE in order to protect capital” and “taking on credit risk again to improve ROE.”

Credit analytical implications. The implication of this Q&A is that the CET1 ratio should be viewed not as a static figure in the capital adequacy table, but in combination with ROE, NIM, fee income, credit impairment, and RWA growth. If low ROE becomes structural, weak internal capital generation may become a constraint on ratings and spread valuation even without a sharp rise in the NPL ratio.

Unresolved items. It remains unconfirmed whether future NIM will be supported only by lower deposit costs, how far loan yields and fee income will recover, and whether ROE can stably recover to the 5% range after credit costs normalise. From the 2026 interim report onward, it will be necessary to check the breakdown of NIM, average loan yield, deposit cost, net fee income, and the full-year direction of credit impairment losses.

3.3 Whether D-SIB Status, Deposit Base, and LCR Support Senior Debt and Capital Instruments in the Same Way

Intent of the question. The third main question was intended to assess how far D-SIB designation, the deposit base, and high LCR support liquidity and refinancing risk for senior bonds, and, separately, how far they mitigate loss absorption, coupon cancellation, and market-access deterioration risks for capital instruments such as AT1 and Tier 2.

Main points of the answer. In the discussion, senior bonds and capital instruments were clearly separated. D-SIB designation, the deposit base, LCR, and regulatory supervision are important supports for senior credit, making it difficult to treat an abrupt liquidity crisis or inability to refinance as the base case. However, D-SIB designation is not a government guarantee and does not eliminate the risk of loss absorption, non-call, coupon cancellation, or market-access deterioration for capital instruments.

AT1 and Tier 2 can react earlier to CET1 decline, low ROE, persistently high credit impairment, dividend restraint, and non-call concerns even if the issuer remains a going concern. AT1, in particular, is more sensitive than senior debt to CET1 and loss-absorption headroom, and should not be treated as safe based solely on D-SIB support expectations.

Points explored in the follow-up. The follow-up question asked about the order in which the market would begin to react if CET1 declined further from the low-9% range, and what external signals should be monitored. The answer was that the first items to watch would not be an inability to refinance senior debt, but rather a combination of worsening issuance terms for AT1 and Tier 2, concerns over non-call of outstanding AT1 instruments, relative spread widening in subordinated bonds, indications of common dividend restraint, and deterioration in the rating outlook.

A rating outlook change would indicate deterioration in the issuer’s overall credit profile, but stress within the capital structure could appear earlier in AT1 and Tier 2 prices, call expectations, and new-issue terms. Higher deposit costs and lower LCR would indicate broader liquidity and funding stress. In this discussion, however, the priority was to first monitor market stress in capital instruments.

Credit analytical implications. The implication of this Q&A is that the risk sensitivity of senior bonds, Tier 2, and AT1 differs even under the same issuer name. For senior credit, D-SIB status, deposits, and LCR can be assessed positively. For AT1 and Tier 2, however, CET1, ROE, credit impairment, common dividends, non-call possibility, and individual contractual terms need to be separately reviewed.

Unresolved items. The triggers, interest and principal suspension, write-down, and call provisions of individual AT1, perpetual capital bonds, and Tier 2 capital bonds have not been confirmed. Live spreads, prices of outstanding bonds, OAS, CDS, and same-tenor comparisons were also not checked in this discussion. The rating agencies’ subordinated-debt notching and the details of the latest outlook also require further confirmation.

3.4 Whether Management Policy Is Capital Defence or Renewed Risk-Taking to Restore Earnings

Intent of the question. The fourth main question was intended to determine whether the bank’s business and financial policy over the next one to two years is moving toward “reducing risk and protecting capital” or toward “taking a certain amount of risk again to restore earnings.” The focus was particularly on the combination of credit cards, MSE, private enterprise loans, real estate loans, low-risk assets, dividend policy, and RWA management.

Main points of the answer. In the discussion, based only on official messaging, the bank appeared to be mindful of risk control, structural adjustment, and capital stability. However, if low ROE is left unaddressed, internal capital generation will remain weak, giving management an incentive to tilt back toward higher-yield assets. Therefore, for credit assessment purposes, the discussion concluded that actual RWA growth, credit card and MSE balances, real estate loan balances, the ratio of low-risk assets, and the actual dividend payout ratio should be monitored, rather than management messaging alone.

Actions consistent with capital defence would include controlling RWA growth, selective operation of credit cards and MSE, management of real estate loan balances, maintaining the ratio of low-risk assets, and restraining dividends where necessary. Renewed risk-taking to restore earnings would likely appear in the form of renewed acceleration in credit card, MSE, and private enterprise loans while CET1 remains in the low-9% range, with RWA growing faster than profit accumulation.

Points explored in the follow-up. The follow-up question asked which actions should be used to determine whether the bank is pursuing a capital-defensive policy. The answer was that the most important combination is RWA growth and credit card/MSE balances. If the bank is controlling RWA while operating credit cards and MSE selectively, this can be read as capital-defensive. Conversely, if RWA growth reaccelerates while CET1 remains in the low-9% range, credit card, MSE, and private enterprise loans increase, and the ratio of low-risk assets declines, there is a high likelihood that the bank is taking risk again to improve low ROE.

On the dividend payout ratio, if the bank continues to maintain a ratio around 30% under low ROE, this should be viewed as a constraint on CET1 accumulation. However, dividend restraint also affects the message to shareholders and market valuation, so it should not be judged simply as good or bad, but should be checked as an action that indicates the bank’s capital-defence stance.

Credit analytical implications. The implication of this Q&A is that the bank’s creditworthiness should be assessed not only by whether it has weak asset categories, but also by whether management re-expands or restrains those sources of deterioration. Conservative operation would delay ROE recovery, but would be positive for CET1 and asset quality. A rushed recovery in earnings may support margins in the short term, but could increase future credit impairment and CET1 decline risk.

Unresolved items. From the 2026 interim report onward, it will be necessary to check the breakdown of RWA, loan mix, credit card and MSE balances, real estate-related exposures, the ratio of low-risk assets, and dividend policy. Whether management’s stated policy is consistent with actual asset allocation remains unconfirmed.

3.5 Weakest Channel Under a Macroeconomic Deterioration Scenario

Intent of the question. The fifth main question was intended to identify which macro and business-cycle factors the bank is most vulnerable to if China’s economic slowdown, real estate adjustment, deterioration in private-enterprise funding conditions, deterioration in consumer credit, and the low interest-rate environment continue simultaneously. The hypothesis tested was that credit deterioration is more likely to become visible in a situation where simultaneous deterioration in private enterprises, small business owners, and consumer credit overlaps with delayed NIM recovery under low interest rates, rather than through a standalone real estate shock.

Main points of the answer. In the discussion, the bank’s downside was viewed as more likely to emerge in an environment where low growth, low interest rates, worsening private-enterprise earnings, and household credit deterioration proceed simultaneously, rather than through a standalone renewed deterioration in the real estate cycle. Real estate was treated less as a standalone main channel and more as a factor that amplifies spillovers to wholesale and retail, commercial services, MSE, and consumer credit.

Under this view, even if the NPL ratio does not jump sharply, credit headroom can be eroded if an increase in special-mentioned loans, persistently high credit impairment losses, a ceiling on NIM, weak ROE, and insufficient CET1 accumulation appear at the same time. Even if short-term liquidity remains stable, this could become a stage where credit headroom near the lower end of investment grade should be reassessed.

Points explored in the follow-up. The follow-up question asked what combination of indicators would mark the point at which the situation should no longer be treated as mere cyclical weakness, but as a stage where deterioration in the rating outlook or spread widening should be considered. The answer identified the watch lines as a move in the credit card NPL ratio toward above 4%, deterioration in MSE-related delinquencies, increases in NPLs in wholesale and retail / leasing and commercial services, a move in the special-mentioned loan ratio toward 3%, NIM below 1.4%, ROE below 5%, and a move in CET1 toward below 9.2%.

The warning lines were further defined as a state in which CET1 approaches 9%, credit impairment losses move toward RMB60bn, ROE falls to the low-4% range or below, special-mentioned loans move toward above 3%, and worsening AT1 and Tier 2 issuance terms or non-call concerns emerge. At this stage, even if real estate NPLs have not surged, the deterioration in private enterprises, small business owners, and consumer credit should be seen as feeding through to earnings capacity and capital generation.

Credit analytical implications. The implication of this Q&A is that China Minsheng Bank should not be simplified as a “China real estate risk name.” The bank’s weakness becomes more pronounced when exposures to private enterprises, SMEs, consumer credit, low NIM, low ROE, and thin CET1 headroom overlap. D-SIB status, the deposit base, and LCR reduce the risk of abrupt payment concerns on senior debt, but they do not automatically restore credit headroom near the lower end of investment grade.

Unresolved items. It remains unconfirmed which loan categories’ delinquencies and special-mentioned loans would show the first signs of macro deterioration, and the category-level flows from 2026 onward are not yet confirmed. The degree to which renewed real estate deterioration would spill over to related sectors, households, and collateral values is also not quantitatively confirmed.

4. Monitoring Framework

Based on the overall discussion, it is practical to monitor the bank across the following three layers.

The first layer is leading indicators of asset quality. The key is whether a move in the credit card NPL ratio toward above 4%, deterioration in MSE-related delinquencies, simultaneous deterioration in the NPL amount and ratio for wholesale and retail / leasing and commercial services, and a move in the special-mentioned loan ratio toward 3% occur together. This layer checks whether the weaknesses partly confirmed in 2025 continue across categories from 2026 onward.

The second layer is earnings capacity and capital generation. The key is whether a move in NIM toward below 1.4%, continued ROE below 5%, credit impairment losses above RMB55bn, a move in CET1 toward below 9.2%, and reacceleration in RWA growth occur together. If this layer deteriorates, the issue becomes not just asset quality, but internal capital generation.

The third layer is market stress within the capital structure. The key is whether group CET1 moves toward below 9.2%, the bank-level CET1 moves toward below 9%, and signs emerge of worsening AT1 and Tier 2 issuance terms, concerns over non-call of outstanding AT1 instruments, and indications of common dividend restraint. This layer is intended to capture the possibility that capital instruments react before acute deterioration in senior credit.

5. Candidates for Inclusion in issuer_notes.md

The following are candidates to consider transferring to the “Follow-up on management strategy, investment plan, and financial policy” section of issuer_notes.md from the next update onward. This report only identifies the candidates and does not update issuer_notes.md itself.

6. Unresolved Items

NPL formation, delinquencies, write-offs, recoveries, and the breakdown of special-mentioned loans by credit cards, MSE, wholesale and retail, and leasing and commercial services from 2026 onward remain unconfirmed. These need to be checked in the 2026 interim report, 2026 annual report, or more detailed company disclosures.

It remains unconfirmed which categories are generating credit impairment losses, and whether those losses will remain elevated above RMB55bn or move toward RMB60bn. The first-quarter 2026 materials alone are insufficient to assess full-year credit costs or seasonality.

It remains unconfirmed how far NIM improvement depends on lower deposit costs, and whether loan yields and fee income will contribute sufficiently to ROE recovery. It will be necessary to check whether a move in NIM toward below 1.4%, ROE below 5%, and CET1 toward below 9.2% emerge at the same time.

It remains unconfirmed whether management policy is actually capital-defensive. This needs to be assessed through actual RWA growth, credit card and MSE balances, real estate loan balances, the ratio of low-risk assets, and dividend payout ratio.

The individual terms of AT1, perpetual capital bonds, and Tier 2 capital bonds, outstanding bond prices, spreads, new-issue terms, non-call concerns, and rating-agency subordinated-debt notching remain unconfirmed. This report does not make an investment judgement on individual securities.

It remains unconfirmed which indicators rating agencies would use as direct triggers for an outlook change. The full text of the latest primary releases, rating rationales, support assessments, and downgrade triggers need to be further checked.

7. Reference Context

This report refers to the China Minsheng Banking issuer summary dated 2026-05-18, existing notes on the issuer, and the SSC discussion on 2026-06-04. The confirmed primary materials mainly used in the existing issuer summary are China Minsheng Banking Corp., Ltd.’s 2025 annual report, first-quarter 2026 report, 2025 annual results announcement, PBOC/NFRA materials related to D-SIBs, and HKEX disclosure search.

The warning lines in the SSC discussion are analytical hypotheses based on figures confirmed in the existing report. They are not formal thresholds published by the company, rating agencies, or regulators.