Bank of East Asia Additional Discussion Report: Hong Kong CRE, Low Profitability, and LAC Refinancing Risk

• Report date: 2026-05-30
• Issuer / Theme: Bank of East Asia / Hong Kong CRE, low profitability, and LAC / Tier 2 refinancing risk
• Report type: additional_discussion
• Discussion scope: Supplemental organisation of the SSC discussion dated 2026-05-29, based on the existing issuer_summary
• Reference context: Current issuer summary dated 2026-05-11, issuer notes / knowledge snapshot / source registry, and discussion generated on 2026-05-29

1. Purpose and Treatment

This report organises the additional questions and answers raised in the discussion on 2026-05-29 as a supplementary document to the existing issuer_summary for Bank of East Asia (BEA). It is not a formal issuer_summary update based on new primary sources. Its purpose is to preserve, separately, hypotheses raised in the discussion, issues already confirmed in the existing report, and items requiring further verification.

The discussion was not intended to update a final investment view. Rather, it was intended to identify points that could easily be missed in future monitoring of BEA. Accordingly, the sections below do not reach a “buy or sell BEA” conclusion, but instead organise Hong Kong property investment lending, low ROE, PPOP, non-interest income, LAC / Tier 2 refinancing, market valuation, and rating outlooks as a future verification framework.

The baseline view already confirmed in the existing report is that BEA is a Hong Kong bank with a strong deposit base, liquidity, and CET1 ratio, but also with property-related problem assets and low profitability. The current discussion does not change that baseline view. Rather, it should be positioned as a more concrete articulation of “in what sequence deterioration would appear, and which indicators should lead to a higher alert level.”

2. Analytical Read-Through from the Discussion

The central point of the discussion was that it is premature to anticipate credit improvement in BEA based only on the reduction of mainland China property risk; Hong Kong property investment, low ROE, and LAC / Tier 2 refinancing costs need to be assessed together. The existing issuer_summary acknowledged the reduction in mainland China property exposure, high CET1 ratio, and strong liquidity as positive factors, while also identifying the scale of Hong Kong property investment exposure, Hong Kong impaired lending, low ROE, and insufficient transparency around large-name exposures as constraints. The discussion translated these constraints into a more practical monitoring sequence.

First, Hong Kong property investment was framed not only as a question of “how large the balance is,” but also as a question of “what remains after the reduction.” The discussion hypothesis was that, if stronger exposures are repaid, sold, or refinanced first, and what remains consists of illiquid collateral, property investment companies with weak refinancing capacity, and large-name exposures requiring individual negotiations, a reduction in the outstanding balance may indicate higher risk density rather than credit improvement. This hypothesis is unverified, but the proposed monitoring axis is useful: place greater emphasis on arrears, Stage 2, collective provisions, specific provisions, collateral coverage, and the workout status of large problem borrowers than on the gross outstanding balance.

Second, BEA’s low ROE should be treated not merely as a temporary factor caused by CRE provisioning, but as a low-profitability issue that could become a medium-term rating constraint. As already confirmed in the existing report, BEA is not a capital-deficient bank. Rather, it is a bank with a thick CET1 buffer but low ROE, and it has not yet been confirmed whether its ability to regenerate capital through earnings will sufficiently recover after problem asset resolution. The discussion stated that ROE recovery should be verified not only by whether credit costs decline, but also by PPOP growth, NIM stabilisation, cost-efficiency improvement, a return to profitability in Hong Kong wholesale banking, and the quality of non-interest income.

Third, BEA’s current management direction is centred on defensive restructuring, but the bank may be increasing its reliance on non-interest income, wealth, insurance, structured products, treasury, and cross-border businesses in order to offset low ROE. This may represent genuine revenue diversification that reduces credit risk, but if it is merely a shift into revenue dependent on market conditions, there is a risk of overestimating PPOP stability. It is necessary to separate recurring fee income, net inflows in wealth AUM, reliance on structured products / securities / bond sales, volatility in treasury income, and market-risk RWA, rather than looking only at the total amount of non-interest income.

Fourth, BEA’s funding risk may appear first not as deposit outflows or a short-term liquidity crisis, but as higher refinancing costs for loss-absorbing instruments such as LAC / Tier 2. Senior credit is supported by the high CET1 ratio, low loan-to-deposit ratio, strong LCR, and deposit base. By contrast, non-preferred LAC and Tier 2 are more sensitive to the prolonged resolution of CRE issues, low ROE, SACP / BCA, subordination, call decisions, and investor demand. Therefore, LAC / Tier 2 issue spreads, book depth, investor distribution, maturity / call schedules, and headroom against the LAC requirement should be managed separately from ordinary liquidity indicators.

Fifth, downgrade risk is more likely to rise under a combined scenario in which Hong Kong CRE stress, stagnant PPOP, NIM compression, reversal in non-interest income, low ROE, and higher LAC / Tier 2 funding costs overlap, rather than from a single CRE loss event. The senior rating itself is relatively well cushioned by the bank’s thick capital and liquidity, but deterioration may first appear in the rating outlook, SACP / BCA, valuation of subordinated and LAC instruments, and spreads.

3. Organisation of Q&A Content

3.1 Could Hong Kong property investment become the next deterioration trigger?

The first question was whether Hong Kong property investment, rather than mainland China property, could become the next major deterioration trigger in BEA’s credit downside scenario. The purpose of the question was to confirm the sequence in which continued declines in Hong Kong commercial property rents, lower occupancy, falling collateral valuations, and weaker refinancing conditions for developers and property investment companies would affect BEA’s impaired advances, provisions, profitability, and CET1 ratio.

The answer was that Hong Kong property investment could be a major deterioration trigger candidate, but there is insufficient basis to conclude that the centre of gravity of risk has fully shifted from mainland China property to Hong Kong property. More accurately, overall CRE, including both Hong Kong and mainland China, remains the key risk, and Hong Kong property investment remains a large component within that. The existing issuer_summary also treats the reduction of mainland China property risk positively, while identifying Hong Kong property investment and Hong Kong-side impairment / provisioning as important constraints.

The follow-up organised the likely sequence of deterioration. First, falling rents and rising vacancies lead to lower DSCR and refinancing difficulty, and the bank sees increases in watchlist, special mention, and Stage 2 migration. Next, lower collateral valuations worsen LGD and ECL assumptions, appearing in collective provisions. Further deterioration into arrears, restructuring, and individually impaired loans then increases specific provisions. This subsequently flows through to losses in Hong Kong wholesale banking and depressed group ROE. Finally, low ROE makes the quality of the CET1 buffer appear weaker.

The credit implication is that this is likely to appear earlier in the market valuation of LAC / Tier 2 than in short-term liquidity or senior repayment capacity. Because CET1 and liquidity are thick, a scenario in which Hong Kong CRE stress immediately undermines senior credit is not the base case. However, if CRE issues are recognised not as single-year provisioning but as a multi-year problem of low profitability and weak internal capital generation, the impact may appear first in non-preferred LAC / Tier 2 spreads and in rating-outlook caution.

Unverified items include the breakdown of Hong Kong property investment by use, LTV distribution, collateral revaluation frequency, concentration in large problem borrowers, and provisioning sensitivity to collateral value declines. During the discussion, it was also noted that certain figures for Hong Kong impaired advances may differ between the existing issuer_summary and external verification values because of possible definitional differences, and numerical reconfirmation is a point for attention in the next update.

3.2 Has the quality of the remaining portfolio deteriorated after balance reduction?

The next question was how to assess the possibility that, even if BEA is actively reducing CRE exposure, stronger exposures are being repaid, sold, or refinanced first, leaving the heavier problem exposures behind. The purpose of the question was to confirm whether a simple reduction in balances can be treated as credit improvement, or whether rising risk density in the remaining portfolio should be suspected.

The answer was that this hypothesis is worth testing, but cannot be concluded at this point. A surface-level reduction in balances can be confirmed, but the attributes of the remaining exposures are not sufficiently disclosed. Therefore, greater emphasis should be placed on arrears, Stage 2, collective provisions, specific provisions, collateral coverage, and realised disposal of large problem borrowers than on the total outstanding balance.

The follow-up point explored was the possibility that balance reduction and lower credit costs may not coincide. Even if balances decline, if collective provisions or arrears do not improve, the bank may not have “reduced bad assets” but may instead be holding a portfolio after “good assets have exited.” In that case, the risk shifts from the balance amount to the loss rate on the remaining book, recovery period, and difficulty of bilateral negotiations.

The credit implication is that BEA’s Hong Kong CRE risk needs to be reframed from “how much remains” to “what is the quality of what remains.” The existing issuer_summary treated Hong Kong property investment as a major constraint, but the current discussion clarified that, from the next update onward, balance reduction itself should not be treated as an improvement signal; it is necessary to confirm whether asset-quality indicators are improving at the same time as balances are declining.

Unverified items are the composition of the remaining portfolio by use, borrower, sponsor, and collateral. In particular, the breakdown among offices, retail, hotels, mixed-use commercial properties, listed developers, unlisted property investment companies, and family-owned holding companies remains unconfirmed. On public materials alone, it cannot be stated that “good loans have exited first.”

3.3 Is low ROE a temporary provisioning issue or a medium-term rating constraint?

The third question was whether BEA’s low ROE is a temporary issue caused by property provisions, or a structural decline in earnings capacity that could become a medium-term rating constraint. The purpose of the question was not to assess whether current capital ratios are high, but to confirm whether BEA remains a bank that can regenerate capital through its own earnings in the future.

The answer was that low ROE should not be explained only as a temporary factor caused by CRE provisioning, but should be treated as a low-profitability profile that could become a medium-term rating constraint. However, it cannot yet be said that recovery is structurally impossible. As of 2025, the situation was characterised by heavy property provisions combined with NIM compression, low growth, asset portfolio restructuring, cost investment, and losses in Hong Kong wholesale banking.

The important follow-up point was that PPOP has not collapsed, but strong growth capacity has also not been confirmed. CRE credit costs account for a large part of the direct cause of low ROE, so ROE could mechanically improve if provisions fall away. On the other hand, if NIM compression, weak net interest income, a deteriorating cost-to-income ratio, and losses in Hong Kong wholesale banking remain, ROE may not recover to a high level even after credit costs normalise.

The credit implication is that BEA should be viewed not as a capital-deficient bank, but as a bank with thick capital whose capital regeneration capacity after problem asset resolution has not yet been proven because of low ROE. For senior credit, CET1, deposits, and LCR provide near-term support. For LAC / Tier 2, however, future capital regeneration capacity, earnings stability, problem asset resolution costs, and the regulatory loss-absorption hierarchy carry greater weight.

Unverified items include normalised ROE if CRE-related provisions normalise, whether losses in Hong Kong wholesale banking are temporary or structural, the sustainability of fee income, and scope for ROE improvement through capital allocation review.

3.4 Should PPOP growth conditions be set for ROE recovery?

The next question was whether, in assessing BEA’s ROE recovery, it is sufficient to wait for CRE credit costs to normalise, or whether explicit recovery conditions should be set for PPOP growth itself. The purpose of the question was to distinguish between a bank that will naturally recover once credit costs decline and a bank in which low ROE remains a structural constraint unless PPOP grows.

The answer was that explicit recovery conditions should be set for PPOP growth itself. BEA in 2025 was not only facing high credit costs; PPOP was barely growing. Therefore, even if CRE provisions decline to some extent, credit profile improvement will be limited if PPOP is flat or declining while NIM and cost efficiency deteriorate.

The follow-up identified the conditions for ROE recovery as a decline in CRE credit cost / PPOP, sustained positive year-on-year PPOP growth, NIM stabilisation, improvement or cessation of deterioration in the cost-to-income ratio, sustained post-impairment profitability in Hong Kong wholesale banking, non-interest income sustained over multiple periods, and a clear improvement from the 3% range in ROE. The particularly important indicators are year-on-year PPOP growth and a return to profitability in Hong Kong wholesale banking.

The credit implication is that even if ROE rises because CRE provisions decline, that only means “losses have fallen”; it does not necessarily mean “earning capacity has recovered.” To improve the view on BEA, both lower credit costs and PPOP growth need to be confirmed at the same time.

Unverified items include the sustainability of PPOP growth, stability of non-interest income, effectiveness of cost-efficiency improvement, and the conditions for profitability in Hong Kong wholesale banking. These need to be verified together with segment results in the next interim and annual results.

3.5 Is management strategy defensive restructuring or renewed risk-taking?

The next question was whether BEA’s management strategy is defensive restructuring that reduces credit risk, or a move to increase new risk-taking in order to offset low profitability. The purpose of the question was to assess, when CRE exposure reduction, capital preservation, cost reduction, stronger non-interest income, wealth / insurance / digital investment, mainland China business, and shareholder returns are viewed together, whether the credit profile is becoming more conservative or revenue volatility is increasing.

The answer was that, at present, defensive restructuring is the main theme. CRE exposure reduction, maintenance of capital and liquidity, a lower loan-to-deposit ratio, cost-efficiency initiatives, and RWA discipline all point in the direction of lower credit risk. However, at the same time, it is also clear that the bank is increasing its reliance on non-interest income, wealth, insurance, structured products, trading, and cross-border business in order to offset low ROE. Therefore, BEA should not be viewed simply as a bank engaged only in de-risking.

The follow-up organised the issue as follows: BEA is reducing CRE risk while trying to supplement capital efficiency and profitability through non-interest income, wealth, digital initiatives, and mainland China connectivity. This is positive for senior credit in terms of loan-risk reduction, but if the quality of earnings shifts toward dependence on market conditions, PPOP volatility and conduct risk become new monitoring items.

The credit implication is that BEA’s management strategy should not be viewed through a binary lens of “defensive or offensive.” CRE resolution, capital preservation, and liquidity maintenance support rating stability. On the other hand, if reliance on non-interest and market-related revenue increases in order to offset low ROE, the credit focus broadens beyond CRE credit-loss risk to revenue volatility and market-condition dependence.

Unverified items are the breakdown between stable and market-dependent components of non-interest income, risk of renewed expansion in mainland China business, the relationship between ROE improvement pressure and dividends / buybacks, and whether the decline in RWA reflects genuine de-risking or regulatory calculation effects.

3.6 Can non-interest income be assessed as stable fee income?

The next question was whether, if BEA increases non-interest income, wealth, structured product, and treasury revenue, that revenue should be viewed as stable fee income or discounted as market-condition-dependent revenue used to compensate for low ROE.

The answer was that not all non-interest income should be treated as stable recurring fee income; it should be separated into stable fees and market-condition-dependent revenue, and the latter should be discounted in the assessment. Insurance, MPF, account administration, payments, corporate cross-border fees, and part of AUM-based wealth revenue are relatively stable. By contrast, mutual fund, bond, and securities sales, structured product sales, and treasury / trading income are more likely to be affected by market conditions, client risk appetite, and product sales volumes.

The follow-up stated that the focus should not be total non-interest income, but the recurring fee income ratio, net inflows in wealth AUM and the contribution from market valuation gains, reliance on structured products / securities / bond sales, quarterly volatility in trading / treasury income, linkage between non-interest income growth and market-risk RWA growth, customer complaints and conduct-related disclosures, and the actual contribution to PPOP.

The credit implication is that growth in non-interest income is positive in itself, but if market-related revenue reverses while CRE provisions remain, ROE and internal capital generation capacity may look weak again. For LAC / Tier 2 in particular, it is important not to overestimate PPOP stability.

Unverified items are that BEA does not sufficiently disaggregate non-interest income in public materials, net inflows in wealth AUM are unclear, it is unclear whether revenue from structured products / bonds / securities sales is temporary or recurring, and signs of conduct risk cannot be identified from the materials reviewed in this discussion alone.

3.7 Will funding risk appear in LAC / Tier 2 refinancing rather than deposit outflows?

The next question was whether BEA’s funding and liquidity risk may surface first as higher refinancing costs for capital-like and loss-absorbing instruments such as LAC / Tier 2, rather than as deposit outflows. The purpose of the question was to review BEA not as a normal liquidity-risk case, but as a refinancing-risk case for loss-absorbing capital.

The answer was that this hypothesis is reasonable. BEA’s deposit and liquidity indicators are strong, and there is little need to set a conventional bank-run type risk as the main scenario. By contrast, LAC / Tier 2 is sensitive to CRE issues, low ROE, rating agencies’ standalone assessments, and investor risk tolerance because of its capital-like nature, subordination, and position in the loss-absorption hierarchy.

The follow-up stated that LCR and the loan-to-deposit ratio alone are insufficient, and that it is necessary to track LAC / Tier 2 issuance feasibility, issue spreads, reoffer yields, secondary spreads, investor distribution, book depth, redemption / call decisions, the balance trends of Tier 2 capital and AT1 capital, headroom against the LAC requirement, and changes in SACP / BCA.

The credit implication is that senior credit and loss-absorbing instruments need to be assessed separately. Senior credit is supported by high CET1, LCR, and the deposit base, but LAC / Tier 2 may price in future losses, capital erosion, call risk, refinancing cost, and weak internal capital generation earlier.

Unverified items are the precise maturity, call, and refinancing schedule for LAC / Tier 2 over the next several years, issue spreads, investor demand, intention to reissue AT1, and headroom against the LAC requirement.

3.8 Can BEA adjust LAC / Tier 2 issuance in a high-cost environment?

The next question was whether, if BEA faces high-cost LAC / Tier 2 refinancing, it would continue issuance to maintain market access, or whether it has room to adjust issuance volumes and redemption policy by using its thick CET1 buffer. The purpose of the question was to confirm how much deterioration in the LAC / Tier 2 market can be absorbed through capital-policy flexibility.

The answer was that BEA has some room to avoid high-cost LAC / Tier 2 issuance, but cannot completely avoid market access. Because CET1 is thick and liquidity is strong, the likelihood of being forced into high-cost issuance in the near term is limited. On the other hand, under Hong Kong LAC regulations, the bank needs to maintain loss-absorption and recapitalisation resources, and whenever maturities, redemptions, or calls arise, it will need to adjust using a combination of CET1, RWA, dividends, and LAC-eligible debt.

The follow-up identified available tools as using the CET1 buffer, suppressing dividends, reducing RWA, issuing LAC / Tier 2 in smaller and more diversified tranches, not calling or redeeming, and limiting capital returns. However, none of these is cost-free. Maintaining a thick CET1 ratio makes ROE look lower, reducing RWA may weaken PPOP growth, and skipping a call could adversely affect investor confidence and the terms of the next issuance.

The credit implication is that BEA is not a bank that would mechanically continue issuing LAC / Tier 2 in a high-cost environment, but regulatory requirements and market practice mean it cannot completely avoid market access for loss-absorbing instruments. Therefore, if high-cost issuance continues, it should be managed as a risk that affects PPOP, ROE, and LAC / Tier 2 spreads first, and only later flows indirectly through to senior credit.

Unverified items include headroom against the LAC requirement, the full LAC / Tier 2 maturity and call schedule, issue spreads, reliance on small RMB / CNH issuance, call decisions, willingness to adjust the dividend payout ratio downward, and the substance of RWA reduction.

3.9 Where would rating risk deteriorate first under combined stress?

The final question was whether downgrade risk for BEA is more likely to rise under a combined scenario involving Hong Kong macro deterioration, a low-rate environment, low profitability, and weaker capital-market valuation, rather than from a standalone CRE loss event. The purpose of the question was to confirm the sequence for monitoring asset quality, PPOP, ROE, SACP / BCA, the LAC issuance environment, and capital return policy.

The answer was that downgrade risk is more likely to increase under a combined scenario than from a standalone CRE loss event. Because BEA has thick capital and liquidity, it is not the type of bank whose senior rating would immediately break from a one-off property loss. On the other hand, if CRE-related provisions, low ROE, stagnant PPOP, NIM compression, market-dependent non-interest income, and weaker LAC / Tier 2 market valuation overlap, caution may first appear in rating agency commentary, standalone assessments, and spreads on subordinated / LAC instruments.

The follow-up indicated the deterioration sequence as follows: deterioration in Hong Kong CRE asset quality, pressure on PPOP from credit costs, NIM compression and slowing non-interest income, lower ROE and internal capital generation, worsening SACP / BCA and outlook, deterioration in LAC / Tier 2 market valuation, and finally spillover to senior ratings.

The credit implication is that BEA’s boundary line should not be set only at absolute declines in CET1 or LCR. Rather, if persistently high CRE credit cost / PPOP, continuing losses in Hong Kong wholesale banking, no recovery from ROE in the 3% range, stagnant PPOP, NIM compression, slowing non-interest income, widening LAC / Tier 2 spreads, and worsening rating agency commentary overlap, the view on BEA needs to shift from “a bank still working through CRE issues” to “a medium-term deterioration in credit profile.”

Unverified items include the quantitative levels at which rating agencies would change the outlook or lower SACP / BCA, the Stage 2 ratio of Hong Kong CRE and LTV after collateral revaluation, expected recovery by large problem borrower, the breakdown of non-interest income, and secondary spreads and investor distribution for LAC / Tier 2.

4. Distinction Between Confirmed Points, Discussion Hypotheses, and Unverified Items

5. Continuing Follow-Up and Candidates for Transfer to issuer_notes

From the current discussion, the candidates that should be considered for transfer to the “Follow-up on management strategy, investment plans, and financial policy” section of issuer_notes.md in future updates are narrowed down to the items below. This does not update issuer_notes.md as of the date of this report; these are candidates that should be managed continuously in future report updates.

6. Unverified Items and Materials to Review Next

In this discussion, many items remained “hypotheses to be verified” rather than “confirmed facts.” In the next issuer_summary / issuer_flash / issuer_notes update, the following should be prioritised for verification.

7. Reference Context

This report was prepared based on the following existing context.

• Bank of East Asia issuer summary dated 2026-05-11
• Bank of East Asia issuer notes, knowledge snapshot, and source registry
• discussion generated on 2026-05-29

The main sources referenced within the discussion were BEA annual, interim, and regulatory disclosures, rating agency materials, HKMA-related information, and the public issuer page. However, this report is an organisation of the relevant Q&A, and the discussion hypotheses listed here should not all be treated as verified facts.