Issuer Credit Research
Bank of China Additional Discussion Report: Downside Monitoring
Bank of China Additional Discussion Report: Downside Monitoring
- Report date: 2026-05-30
- Issuer / Theme: Bank of China Limited / Downside monitoring for real estate, policy lending, foreign-currency and overseas operations, and local-government-related risks
- Report type:
additional_discussion - Discussion scope: Organises the SSC discussion held on 2026-05-29 as supplementary points to the existing issuer_summary
- Reference context: Bank of China issuer_summary dated 2026-05-18, issuer_notes / knowledge_snapshot / source_registry, and discussion dated 2026-05-29
1. Purpose and Treatment
This report is a supplementary report that organises the additional Q&A on Bank of China Limited (“BOC”) conducted during the discussion on 2026-05-29, while cross-checking it against the existing issuer_summary. The observations here are not definitive credit judgements based on new primary-source research. They are intended to record issues, warning lines, and unconfirmed items that should be checked when reading or updating the existing report.
The existing issuer_summary frames BOC as a large Chinese state-owned G-SIB whose credit profile is supported by its huge deposit base, relationship with Huijin and the MOF, adequate capital and liquidity, and high likelihood of government support. At the same time, credit improvement is constrained by low NIM, real-estate-sector NPLs, personal consumption loans, personal business loans and credit cards, policy-driven credit supply, and the complexity of its overseas, foreign-currency, and subsidiary structure. The latest Q&A does not materially replace that existing view. Rather, it should be treated as a deeper discussion of “which early-stage indicators should be monitored while headline NPLs remain stable.”
This report separates points already confirmed in existing reports, hypotheses raised in the discussion, and unconfirmed items. Items not verified through primary sources in this exercise—such as market spreads, CDS, individual bond prices, order books for new foreign-currency bonds, and the balance, RWA, and modified amount of local-government-related lending—are not treated as confirmed facts.
2. Analytical Read-Through from the Discussion
The main focus of the Q&A was not the NPL ratio of real-estate lending itself, but how to capture at an early stage the transmission channels from real estate into adjacent sectors, personal credit, policy lending, local-government-related lending, and foreign-currency market access. As confirmed in the existing report, BOC’s senior issuer credit is not structured in a way that would deteriorate rapidly because of a short-term liquidity crisis or losses in a single sector. At the same time, with NIM having declined to around 1.26%, any sustained increase in credit costs or expansion in low-yield lending could make internal capital generation and the pace of CET1 recovery more visible issues for the market.
For real-estate lending, the high NPL ratio of 6.26% at end-2025 has already been confirmed in the existing report. However, the discussion framed the key downside question for BOC as whether the deterioration spreads beyond real estate into construction, commerce and services, personal business loans, mortgages, and low-rate refinancing or loan modifications related to local governments. This gives more concrete form to the monitoring points in the existing issuer_summary: “simultaneous deterioration in real estate and household credit” and “long-term refinancing and loan modifications for LGFVs and local-government-related debt affecting NIM, asset turnover, RWA, and profitability.”
On policy lending, the key issue was not identified as one-off capital erosion from M&A or large subsidiary restructuring. Rather, the discussion focused on capital efficiency if high growth in lending to inclusive finance, micro and small enterprises, private enterprises, manufacturing, and strategic emerging industries coincides with low yields, RWA growth, and higher credit costs. This reflects the dual nature of policy alignment: it supports BOC’s franchise and likelihood of government support, but in a low-NIM environment it may also pressure internal capital generation.
For foreign-currency and overseas operations, the discussion did not suggest that BOC currently has weak foreign-currency liquidity. Based on the scale of foreign-currency customer deposits, overseas commercial banking operations, and BOCHK, the Q&A instead emphasised that even where foreign-currency liquidity indicators are strong, spread widening in foreign-currency senior bonds, non-capital TLAC, Tier 2, and AT1 instruments could occur before deterioration is visible in financial metrics. The hypothesis is that if spread widening becomes prolonged, it could feed through into the real economy via foreign-currency funding costs, overseas loan growth, the stickiness of overseas deposits, and the profitability of overseas businesses.
Local-government-related lending is the area with the largest number of unconfirmed items in this Q&A. It is important to distinguish between the fact that BOC manages local-government-related risks as a monitored risk area and the fact that the specific balance, RWA, Stage 2 amount, and modified amount of such lending were not confirmed in this exercise. The discussion hypothesis is that local-government-related risks may materialise not as a sharp increase in NPLs, but as repayment deferrals, loan modifications, low-rate refinancing, and maturity extension, resulting in lower NIM and weaker capital efficiency.
3. Summary of the Q&A
| Q&A Theme | Purpose of the Question | Key Points from the Response / Discussion | Points Explored in Follow-Up | Credit Analytical Implications |
|---|---|---|---|---|
| Second-round spillover from real-estate lending | The question sought to confirm whether deterioration in real-estate-sector NPLs alone would become a rating or spread issue, or whether it would become material only if it spread to construction, commercial services, personal business loans, and local-government-related exposures. | Based on the already confirmed real-estate-sector NPL ratio of 6.26%, low NIM, and strong capital and liquidity, the discussion concluded that real estate alone is unlikely to immediately undermine senior credit, but that second-round spillover would affect the earnings buffer and internal capital generation. | A rough stress sensitivity was discussed in which the real-estate-sector NPL ratio worsens to 10-15%, construction NPLs rise to 3-5%, commercial and service-sector NPLs rise to 2-3%, and mortgage NPLs rise to 1.0-1.5%. This was an estimate within the discussion and not an official stress test. | The primary indicators to monitor are not limited to the headline NPL ratio. They also include Stage 2 loans, special-mention loans, loan modifications, repayment deferrals, low-rate refinancing, and the chain of sector-level NPL deterioration. |
| Policy lending, capital allocation, and dividends | The question sought to confirm whether growth investments, subsidiaries, capital instrument issuance, and common-share dividends could constrain the speed of CET1 recovery in a situation where low NIM, real-estate risk, and policy-driven credit supply overlap. | Large-scale M&A or subsidiary restructuring was not confirmed as the main source of credit pressure. The more important combination was framed as high policy-loan growth, RWA growth, low yields, rising credit costs, and maintenance of a 30% dividend payout. The 2025 common-share dividend payout ratio of 30% has been confirmed, but the dividend adjustment policy under stress remains unconfirmed. | The discussion covered capital efficiency in inclusive finance and micro and small enterprise lending, manufacturing, strategic emerging industries, green finance, BOC Asset Investment, BOCHK, and BOC Aviation. Subsidiaries and investment businesses should be viewed less as the main cause of direct CET1 erosion and more as sources of capital lock-up, earnings volatility, and constraints on capital mobility. | A 30% dividend payout may be acceptable in normal times. However, if higher credit costs and RWA growth occur simultaneously, slower accumulation of retained earnings and a decline in the CET1 ratio toward the low-12% range would become a market-relevant warning line. |
| Foreign-currency and overseas operations, and market access | The question sought to confirm how far BOC’s foreign-currency funding, overseas businesses, and cross-border operations could be affected by renminbi depreciation, persistently high US dollar interest rates, geopolitics, and reassessment of the Chinese sovereign. | The discussion did not frame foreign-currency liquidity as currently weak. Rather, foreign-currency customer deposits, overseas commercial banking operations, and the BOCHK franchise are strengths. However, spreads on foreign-currency senior bonds, non-capital TLAC, Tier 2, and AT1 instruments may react before deterioration appears in CET1 or liquidity indicators. | The discussion explored whether widening in foreign-currency bond spreads would remain a temporary market risk or become prolonged and lead to constraints on overseas lending, higher foreign-currency deposit costs, reduced risk appetite at BOCHK and overseas branches, and slower overseas earnings. | Even if domestic senior credit remains stable, foreign-currency instruments require separate monitoring of market access, spreads, currency-specific liquidity, and jurisdictional risk at overseas subsidiaries and branches. |
| Policy lending and domestic sector credit sensitivity | The question sought to confirm how sensitive policy-aligned lending to manufacturing, strategic emerging industries, private enterprises, housing, and consumer loans is to the economic, interest-rate, and real-estate cycles. | The discussion did not conclude that manufacturing lending is already deteriorating. Rather, it noted that deterioration in personal consumption loans, credit cards, and personal business loans has been leading. Manufacturing and mortgages are currently stable, but balance growth, RWA growth, and migration into Stage 2 are the next items to confirm. | The discussion covered how quickly deterioration in personal and micro/small-enterprise credit could spread to manufacturing, strategic emerging industries, and mortgages, and how much higher credit costs under low NIM could reduce after-tax profit and retained earnings. | Early-warning indicators are not manufacturing NPLs themselves, but Stage 2 loans, delinquency, average interest rates, and loan modifications in personal consumption loans, cards, personal business loans, inclusive finance, and micro and small enterprise lending. |
| Local-government-related lending and restructuring support | The question sought to confirm how much pressure would be placed on CET1 and internal capital generation if local government finances, a weaker real-estate market, lower collateral values, repayment deferrals, loan modifications, and low-rate refinancing overlap. | The specific balance, RWA, Stage 2 amount, and modified amount of local-government-related lending were not confirmed in this exercise. Therefore, the quantitative scale of pressure by region or sector remains unconfirmed. At the same time, the discussion noted that the first impact may appear through lower NIM, low-rate refinancing, longer maturities, and weaker capital efficiency rather than a sharp increase in NPLs. | The discussion explored how local-government-related risk could spill over into domestic manufacturing, mortgages, and real-estate lending, and how it could be reflected in senior credit and market spreads. However, specific BOC-level data are insufficient. | This issue should be treated not as a “confirmed material loss” but as a “potential earnings pressure that is hard to see because of limited disclosure.” NPLs and loan modifications in public works, water conservancy and environmental projects, and transport infrastructure could serve as proxy indicators. |
4. Points Confirmed in the Existing Report
The foundations confirmed in the existing issuer_summary and knowledge_snapshot are as follows. BOC is a large Chinese state-owned G-SIB, and its senior issuer credit is supported by its huge deposit base, closeness to the government, regulatory importance, and adequate capital and liquidity. At end-1Q2026, the CET1 ratio was 12.18%, the total capital adequacy ratio was 18.23%, LCR was 144.67%, NSFR was 127.59%, the NPL ratio was 1.22%, and the allowance coverage ratio was 203.17%.
On earnings, the important point is that NIM declined from 1.59% in 2023 to 1.40% in 2024 and 1.26% in 2025, and remained at 1.26% in 1Q2026. Even if headline profit is stable, the capacity to absorb credit costs has become thinner than before. Therefore, the combination repeatedly raised in this Q&A—higher credit costs, low-yield lending, RWA growth, and maintenance of dividends—can be viewed as an extension of the existing report’s low-NIM issue into a downside scenario.
On asset quality, the real-estate-sector NPL ratio of 6.26% at end-2025, personal consumption loans at 2.18%, personal business loans at 1.95%, and credit cards at 2.18% have already been confirmed in the existing report as key constraints. By contrast, the mortgage NPL ratio was 0.60% and the manufacturing NPL ratio was 0.88%. At least as of the confirmed end-2025 data, mortgages and manufacturing are not framed as the main sources of deterioration.
The existing report also confirms that securities classes need to be analysed separately: senior, non-capital TLAC, Tier 2, AT1 / perpetual bonds, BOCHK, BOC Aviation, and overseas branch bonds. The latest Q&A on foreign-currency and overseas operations reinforced this distinction in securities-class risk through the lens of foreign-currency spreads and market access.
5. Claims and Hypotheses from the Discussion
The most important hypothesis from the discussion is that BOC’s downside trigger is not a single real-estate NPL ratio, but the combination of second-round spillover and reduced loss-absorption capacity under low NIM. The high NPL ratio for real-estate lending has already been confirmed, but based only on its balance share, it is difficult to argue that it would materially erode CET1 on its own. Rather, the discussion concluded that BOC should be monitored for spread into construction, commerce and services, personal business loans, mortgages, and low-rate refinancing related to local governments, using Stage 2 loans, special-mention loans, loan modifications, and repayment deferrals.
The second hypothesis is that personal consumption loans, credit cards, and personal business loans could serve as leading indicators of domestic credit deterioration at BOC. Manufacturing and mortgages are currently stable, but if deterioration in personal credit and micro/small-enterprise-related credit continues, it may spread later to commerce and services, manufacturing, and mortgages. This translates the existing report’s reading that BOC is “sensitive to household credit and SME business conditions” into a monitoring sequence.
The third hypothesis is that policy lending expansion, if accompanied by RWA growth and higher credit costs under low NIM, could become a constraint on internal capital generation. Policy lending increases BOC’s institutional importance, but in areas with low average yields and credit costs that tend to rise in an economic downturn, it slows the build-up of the CET1 ratio. Maintaining a 30% dividend payout may be a limited issue in normal times, but in a phase of higher credit costs and RWA growth it could slow the pace of capital recovery.
The fourth hypothesis is that for foreign-currency instruments, spreads could widen before deterioration appears in financial metrics. There is no confirmation that foreign-currency liquidity itself is weak. However, if US dollar interest rates, reassessment of the Chinese sovereign, the renminbi exchange rate, geopolitics, and Hong Kong market sentiment overlap, the market valuation of foreign-currency senior bonds, non-capital TLAC, Tier 2, and AT1 could move first. Only if spread widening becomes prolonged would this become a real-economy issue through higher foreign-currency funding costs, restraint on overseas lending, reduced stickiness of overseas deposits, and slower overseas earnings.
The fifth hypothesis is that local-government-related lending may materialise through lower NIM and weaker capital efficiency due to repayment deferrals, loan modifications, low-rate refinancing, and maturity extension, rather than through NPLs. This is important, but the BOC-specific quantitative information confirmed in this exercise is insufficient. Therefore, local-government-related risk should be treated not as an item where material losses have already been confirmed, but as an item with thin disclosure that can easily be overlooked as a source of potential earnings pressure.
6. Ongoing Follow-Up Items
| Follow-Up Item | Current Status | Practical Warning Lines / Confirmation Triggers | Materials / Information to Check Next |
|---|---|---|---|
| Second-round spillover from real-estate NPLs into adjacent sectors | Hypothesis from the discussion. The existing report confirms the high real-estate-sector NPL ratio, but the scale of second-round spillover is unconfirmed. | Increase in the construction-sector NPL ratio, increase in the commerce and service-sector NPL ratio, increase in personal business loan NPLs, mortgage NPLs rising toward 0.8-1.0%, and increases in Stage 2 / special-mention loans. | BOC annual and interim reports, sector-level NPLs, Stage 2 loans, special-mention loans, modified loans, and disclosures on repayment deferrals. |
| Leading deterioration in personal consumption loans, cards, and personal business loans | Combination of confirmed facts and discussion hypotheses. NPL deterioration has been confirmed, but spillover to other sectors remains unconfirmed. | Continued increase in personal consumption loan and card NPL ratios, personal business loan NPL ratio approaching 2.5-3.0%, increase in Stage 2 loans, higher delinquency rates, and simultaneous decline in average lending rates. | Breakdown of personal lending, Stage 2 loans, delinquencies, special-mention loans, credit cost ratio, and regional / product-level disclosure. |
| Capital efficiency of policy lending, inclusive finance, and micro and small enterprise lending | Hypothesis from the discussion. Manufacturing NPLs were framed as currently stable, but loan growth and weaker capital efficiency remain unconfirmed risks. | Growth in policy lending persistently exceeding profit growth and retained-earnings growth, NIM decline, decline in average interest rates in inclusive finance, increase in Stage 2 loans, and higher credit cost ratio. | Disclosures related to the “five major areas” of finance, inclusive finance and micro/small-enterprise loan balances, average interest rates, NPLs, Stage 2, RWA, and credit costs. |
| 30% dividend payout and pace of CET1 recovery | Combination of confirmed facts and discussion hypotheses. The 30% dividend payout has been confirmed, but flexibility to adjust under stress is unconfirmed. | CET1 ratio declines toward the low-12% range, RWA growth exceeds profit growth, dividend payout ratio of 30% is maintained even under rising credit costs, and dependence on external capital funding increases. | Dividend policy in annual and interim reports, CET1 trend, RWA growth, retained earnings, and capital replenishment plans. |
| Foreign-currency funding and overseas business market access | Hypothesis from the discussion. Strong foreign-currency liquidity has been confirmed in existing disclosure, but real-economy spillover from spread widening remains unconfirmed. | Decline in order books for new foreign-currency bonds, widening new-issue premium, widening TLAC / capital instrument spreads relative to senior bonds, slower overseas loan growth, slower overseas deposit growth, and deterioration in BOCHK’s LCR and deposit trends. | Foreign-currency bond offering materials, investor allocation, BOCHK results, foreign-currency liquidity ratios, and, if available, foreign-currency LCR / USD LCR, Chinese sovereign CDS, and state-owned bank spreads. |
| Hidden earnings pressure from local-government-related lending and restructuring support | Combination of unconfirmed items and discussion hypotheses. Specific balance, RWA, Stage 2 amount, and modified amount are unconfirmed. | Increase in loan modifications for local-government-related and infrastructure-related lending, expansion of low-rate refinancing, widening local government bond spreads, deterioration in local fiscal balances, decline in real-estate collateral values, and increase in NPLs related to public works, water conservancy, and transport. | BOC annual and interim reports, descriptions of local-government-debt-related exposures, LGFV-related reporting, local government bond spreads, Ministry of Finance and local fiscal data, and rating-agency reports on Chinese local government debt. |
7. Candidates for Transfer to issuer_notes.md
This additional_discussion does not update issuer_notes.md, knowledge_snapshot.md, or source_registry.md. However, in future updates to issuer_notes, the following short points merit consideration as candidates for addition to “follow-up on management strategy, investment plan, and financial policy” or to standing credit questions.
- Continue to monitor second-round spillover from real-estate-sector NPLs into construction, commercial services, personal business, and mortgages as a key downside trigger for BOC, rather than focusing only on real-estate-sector NPLs in isolation.
- Continue to monitor personal consumption loans, cards, and personal business loans as early-warning indicators of domestic credit deterioration at BOC.
- Continue to check whether policy lending expansion becomes a constraint on internal capital generation if it is accompanied by RWA growth and higher credit costs under low NIM.
- Maintaining a 30% dividend payout is a limited issue in normal times, but in a phase of higher credit costs and RWA growth it may constrain the speed of CET1 recovery.
- For BOC’s foreign-currency instruments, spreads could widen before deterioration appears in financial metrics, and if prolonged, this could constrain overseas business earnings and foreign-currency loan growth.
- Local-government-related lending may materialise through lower NIM and weaker capital efficiency caused by repayment deferrals, loan modifications, and low-rate refinancing rather than through NPLs, and BOC-specific disclosure needs to be checked.
These are candidates for transfer and have not been reflected in issuer_notes.md in this exercise. In particular, the quantitative scale of local-government-related lending, the amount of pressure by region and sector, and market reactions in individual spreads remain unconfirmed. They should therefore continue to be treated as unconfirmed items when transferred.
8. Unconfirmed Items
The following items are important in the discussion but were not verified through primary sources in this exercise, or remain unconfirmed in the existing report.
- Detailed Stage 2 balances, modified loans, repayment deferrals, and low-rate refinancing by real estate, construction, commerce and services, personal business loans, and mortgages.
- Balance, RWA, NPLs, Stage 2, modified amount, yield decline, and regional and sector breakdown of LGFV and local-government-related lending.
- Risk-adjusted returns, RWA density, and credit cost ratio for policy lending and loans to inclusive finance, micro and small enterprises, private enterprises, manufacturing, and strategic emerging industries.
- Whether the 30% common-share dividend payout ratio would be maintained under stress, and how much room there is for adjustment.
- Capital mobility constraints, dividend capacity to the parent, and additional support risk under stress at major subsidiaries such as BOCHK, BOC Aviation, BOC Financial Leasing, and BOC Asset Investment.
- Foreign-currency LCR, USD LCR, foreign-currency NSFR, liquidity buffers by major overseas location, and the foreign-currency bond maturity ladder.
- Live spreads, new-issue premiums, order books, investor allocation, and relative comparison with major state-owned bank peers for foreign-currency senior bonds, non-capital TLAC, Tier 2, and AT1.
- Primary-source confirmation of Moody’s issuer-specific outlook update for BOC.
9. Reference Context
This report was prepared by reference to the existing Bank of China issuer_summary, issuer_notes, knowledge_snapshot, source_registry, and the discussion dated 2026-05-29. The existing report referenced was the Bank of China issuer_summary dated 2026-05-18.
Among the figures and claims used in the discussion, items that can be confirmed in the existing issuer_summary and issuer_notes / knowledge_snapshot are treated as confirmed points within that scope. By contrast, the stress sensitivities, warning lines, market spreads, and views on local-government-related lending presented in the Q&A are treated in this supplementary note as discussion hypotheses or unconfirmed items.