CRC Credit Research Compendium
Menu
日本語English

Issuer Credit Research

HDFC Bank Additional Discussion Report: SSC Discussion on Funding, Risk, Capital and Rating Transmission

Issuer: Hdfc Bank | Document: Additional Discussion | Date: 2026-09-22 | Event: Ssc Discussion

1. Purpose and Treatment

This additional discussion records the credit-analysis questions and answers presented in a project-specified SSC external discussion. It is a supplementary discussion record, not a verification of new facts and not a revision to the existing issuer summary or issuer flash. Numerical observations already confirmed in the Q1 FY2027 issuer flash are distinguished below from analytical thresholds, hypotheses and source claims made in the SSC discussion.

The central issue running through the discussion is whether HDFC Bank can complete its post-HDFC Ltd adjustment through lower-cost, durable funding and operating leverage while preserving its established underwriting and capital advantages. The discussion does not conclude that those advantages have weakened. Instead, it lays out observable combinations of funding, asset-quality, profitability, capital and sovereign indicators that would justify changing that view.

2. Confirmed Context and Discussion Takeaway

The existing Q1 FY2027 issuer flash confirms a mixed but still resilient starting point. Deposits grew 14.7% year on year, faster than AUM growth of 12.4%; CET1 was 17.4%, total capital adequacy was 19.6%, average LCR was 115% and NSFR was 119%. At the same time, CASA declined to 32.3%, cost of funds was 4.4% and NIM on total assets fell to 3.26%. GNPA excluding agriculture was 0.91% and credit cost was 0.40%. These data support a view of incomplete post-merger normalisation, not a near-term liquidity, solvency or broad asset-quality problem.

The SSC discussion develops five connected propositions:

  1. Deposit growth above AUM growth is constructive in quantity, but only lower funding cost, more granular deposits, lower legacy borrowing and eventual NIM recovery would demonstrate improved funding quality.
  2. A normal cyclical increase in credit costs should be absorbable from the present earnings, provision and capital base; the more consequential risk would be persistent, multi-segment deterioration that challenges underwriting differentiation.
  3. The capital question is not a single CET1 level. It is whether retained capital can continue to match growth in RWAs as product mix, margins and credit cost change.
  4. Weak NIM recovery would become strategically more concerning if management sought to restore returns through a sustained shift to higher-RWA, higher-risk lending while loosening borrower selection or persisting despite weak new-vintage performance.
  5. A sovereign-driven movement in international ratings or spreads should be separated from HDFC-specific deterioration in funding, underwriting, capital generation or management execution. The portfolio implication differs even when the headline senior rating is unchanged.

The thresholds below are discussion monitoring lines, not management guidance, regulatory minima, or rating-agency triggers. They should not be treated as forecasts.

3. Q&A Discussion Notes

3.1 Funding transition: quantity versus the quality of replacement funding

Question intent. The first exchange asked whether deposit growth running ahead of advances/AUM can demonstrate that the post-HDFC Ltd funding transition is structurally sustainable when CASA has fallen and term deposits have grown more quickly.

Answer points. The SSC response treated the transition as two separate processes: the runoff of legacy HDFC Ltd borrowing and the slower rebuilding of low-cost customer deposits. It reported management commentary that approximately INR40,000-50,000 crore of borrowings could mature over the next couple of years, and that replacing borrowing costing above 7% with retail term deposits at roughly 6%-plus could still improve economics before any CASA benefit. The response therefore cautioned against treating all term deposits as poor-quality funding. It also cited a stated management aspiration to improve CASA over a multi-year horizon rather than promising a rapid return to the approximately 38% merger-period level.

The discussion proposed a more complete confirmation set than CASA alone: continuing reduction in the borrowing share and in absolute legacy borrowing; average deposit growth ahead of average AUM growth; stabilisation then improvement in CASA; evidence that incremental term deposits remain granular; a stable or improving retail/wholesale deposit mix; a decline in cost of funds from 4.4%; and NIM stabilisation or recovery without a shift into riskier assets. Liquidity ratios remain useful resilience checks but do not by themselves prove that funding has become cheaper or more durable.

Follow-up implication. The follow-up asked when persistent low-3.2% NIM would cease to look like a temporary lag. The SSC answer proposed a 12-18 month, or roughly four-to-six-quarter, test after further legacy-borrowing runoff. A sustained combination of cost of funds around 4.4% or higher, CASA at or below the low-30s and still declining, a persistently elevated wholesale/non-granular share, and NIM at or below roughly 3.25% would support an analytical conclusion of structurally weaker funding economics. The discussion then linked funding to credit only when profitability and retained capital generation begin lagging RWA growth, with sustained RoA below roughly 1.5%-1.6% and CET1 trending toward 15%-16% as more material portfolio-warning conditions.

Uncertainty. The SSC material did not establish marginal pricing of new deposits, the FY2027 ticket-size composition of incremental deposits, the detailed legacy-borrowing maturity profile, a firm CASA target, or a numerical NIM floor. Those missing data prevent a conclusion that the funding franchise has either normalised or structurally weakened.

3.2 Credit-cycle resilience: flow indicators and the breadth of stress

Question intent. The second exchange asked how a slower Indian credit cycle and prolonged high real borrowing costs could affect asset quality, particularly in rate-sensitive retail, unsecured and SME exposures.

Answer points. The SSC answer began from the Q1 FY2027 baseline already recorded in the issuer flash: GNPA of 1.17%, NNPA of 0.41%, GNPA excluding agriculture of 0.91% and credit cost of 0.40%. It argued that the more useful early evidence lies in flow data and recent-vintage performance, not a small movement in headline GNPA. The discussion noted Q1 slippages of INR80bn, including INR61bn excluding agriculture, and described personal loans, cards, Business Banking/SME and commercial transportation as the higher-priority portfolios for follow-up.

The response regarded a normal increase in credit cost toward a high-single-digit-basis-point range as potentially absorbable given the reported provision stock, RoA and CET1. It distinguished this from a structural deterioration, which would require persistence and breadth: worsening 30+/60+/90+ day delinquency or SMA migration, weaker cures, repeated high ex-agriculture slippages, and stress spreading from unsecured retail into Business Banking/SME or another material portfolio. The SSC discussion identified credit cost around or above 1% for several quarters, ex-agriculture GNPA moving materially above the existing baseline, and RoA falling toward 1.5% as analytical warnings rather than formal downgrade thresholds.

Follow-up implication. The follow-up tested when risk selection, rather than the cycle, should be considered weaker. The SSC answer set out a staged sequence. Two consecutive quarters of broadening early delinquency, including movement into Business Banking/SME and weaker cure or vintage performance, would justify tightening limits or stopping incremental risk. A stronger base-case change would require roughly three-to-four quarters of multi-portfolio deterioration, credit cost around or above 1% for several quarters, RoA at or below about 1.5%, and a CET1 decline of roughly 100-150bp toward 15%-16% without an unrelated capital action. A broad, persistent combination with credit cost around 1.25%-1.5% or more would be a materially more adverse issuer-credit outcome.

Uncertainty. The relevant public leading indicators were not available in the material: product-level DPD/SMA migration, cures and roll rates, 3/6/12-month vintages, collections, underwriting-score trends, restructuring and segment-level credit cost. The discussion explicitly does not infer deterioration from the absence of these disclosures.

3.3 Self-funded growth: retained earnings versus RWA consumption

Question intent. The third exchange asked whether strategic expansion can remain self-funded if profitability stays below pre-merger levels and business mix increases RWA intensity.

Answer points. The SSC response noted that, based on the discussed Q1 data, RWAs had grown about 12.4% year on year while CET1 remained 17.4%. It treated that as evidence that current growth had not yet eroded capital ratios. It also highlighted a tension in the discussion material: adjusted earnings growth was said to have lagged balance-sheet growth. The analysis consequently focused on retained-capital growth relative to RWA growth, not on CET1 in isolation.

The response described moderate capital use as potentially deliberate. It cited management commentary in the SSC material that approximately 60-70bp of annual capital consumption had historically been compatible with steady growth. It contrasted that with a sustained mismatch in which RWA growth exceeds underlying earnings or retained-capital growth by about five percentage points or more; CET1 declines by roughly 75-100bp annually; RoA drifts toward 1.5%; and RWA density rises. In the SSC framework, a gradual move from 17.4% toward the high-16% area could be a normal use of surplus capital if returns remained adequate, while a move toward or below 16% combined with weaker RoA and higher RWA intensity would indicate deteriorating self-funded growth.

Follow-up implication. The follow-up made the RWA-intensity test more specific. The SSC answer proposed watching whether RWA/assets moved persistently from the stated 69% baseline toward 71%-72%, and whether reported risk-weight density moved from 65% toward 67%-68% or higher, while separating regulatory changes from management choices. For subordinated risk, it recommended an earlier action than for senior debt: two-to-three quarters of a persistent earnings/RWA gap, falling CET1 and RoA below roughly 1.6% could justify tighter AT1/Tier 2 limits. CET1 near 16%, RoA at or below 1.5%, rising RWA density and no improvement in the earnings/RWA gap would be a clearer reduction trigger.

Uncertainty. The exact internal CET1 floor, incremental RWA and economic return by product, payout effects on retained CET1, and final April 2027 expected-credit-loss transition impact were not established. The discussion also noted that regulatory CET1 and rating-agency RAC are different measures and should not be mechanically equated.

3.4 Profitability restoration: testing for reach for yield before NPAs rise

Question intent. The fourth exchange asked whether post-merger returns can be restored through funding, operating leverage and deeper relationships, rather than through higher-risk lending.

Answer points. The SSC response described a benign route through lower funding cost, improving branch and technology productivity, mortgage-customer deposit relationships and cross-sell. It reported that the discussion material viewed these as credible return levers and found no verified evidence that current underwriting standards had been relaxed. It also noted that Business Banking and small/mid-market lending were growing faster than the overall book and that unsecured disbursements were increasing, which makes the next vintages more important.

The response distinguished high growth from reach for yield. A concerning pattern would require weak funding/NIM normalisation, higher-yield portfolios persistently growing much faster than mortgages and the total book, rising RWA intensity, a loosening of borrower-selection or acquisition boundaries, worsening new-vintage performance, and continued management acceleration despite those signals. It suggested that an increase in risk-weight density toward approximately 68% or higher, after starting near 65%, would be more meaningful when accompanied by a broader mix shift and not explained by regulation.

Follow-up implication. The follow-up established an early action sequence. The SSC answer proposed stopping incremental subordinated risk after roughly two quarters if weak NIM, higher RWA intensity and either a credible policy/acquisition easing or worse vintages occur while high-risk growth continues. It proposed tighter issuer and subordinated limits after cumulative evidence across several quarters, including persistent differential growth, multi-cohort deterioration and at least two indicators of weaker selection. The discussion would reduce subordinated exposure before credit cost reaches 1% if management continued accelerating after those signals; broader senior exposure would require the later financial confirmation of sustained credit cost near or above 1%, RoA near or below 1.5%, falling CET1 and weak internal capital generation.

Uncertainty. The public material did not provide borrower-score distributions, FOIR/debt-service data, approval and override rates, new-to-bank sourcing, collateral/LTV movement, risk-adjusted return on capital or product-level vintage curves. Business Banking growth also requires composition analysis before it can be interpreted as a shift down the credit curve.

3.5 Rating and spread transmission: sovereign ceiling versus standalone weakening

Question intent. The fifth exchange asked which downgrade pathways are plausible over the next 12-24 months and how a sovereign-driven move differs from a deterioration in HDFC Bank's standalone profile.

Answer points. The SSC discussion stated that, in the cited S&P framework, HDFC Bank's BBB issuer rating is constrained by India while the assessed standalone credit profile is stronger. It therefore treated a sovereign downgrade as a possible transmission path even if HDFC's franchise, asset quality and capital remain strong. The discussion also stated that HDFC is a private-sector bank and should not be understood as receiving an assumed extraordinary sovereign-support uplift.

For a bank-specific path, the SSC answer required a connected chain rather than one weak metric: weaker underwriting or risk selection, broader losses, lower RoA, RWA growth outpacing retained capital, lower CET1/RAC, and potentially weaker funding or management execution. It identified a one-notch standalone assessment/BCA weakening, if reported, as an important portfolio signal even if a sovereign-capped senior rating did not change.

Follow-up implication. The follow-up on spread widening proposed a classification test. Widening is primarily country-risk repricing when there is an identifiable sovereign catalyst, Indian sovereign and bank peers widen together, HDFC does not materially underperform peers after duration/liquidity adjustment, and HDFC's funding, credit cost, RoA and CET1 remain resilient. A mixed or issuer-specific view becomes more appropriate if HDFC underperforms peers while funding normalisation remains weak, credit cost approaches or exceeds about 1% for several quarters, RoA approaches 1.5%, CET1 trends below about 16%, and asset-quality stress broadens. The response emphasised that required spread, country concentration and subordinated-risk appetite should react earlier than a formal senior-rating action.

Uncertainty. The SSC discussion did not independently retain a new full S&P scorecard following methodology changes or a complete Moody's mapping from BCA changes to every instrument rating. It also did not establish security-specific spread levels. Peer-relative spread interpretation requires matched maturity, ranking and liquidity data.

3.6 Management continuity as a cross-cutting monitor

The final extraction from the SSC discussion added CEO succession as an unconfirmed, cross-cutting monitor. The discussion did not treat succession itself as credit deterioration. It identified an orderly transition with continuity in underwriting, funding and capital discipline as the relevant confirmation, while prolonged uncertainty, key risk/control departures or a visible change in growth and risk appetite would require reassessment. This point should be read alongside, not instead of, funding, underwriting and capital evidence.

4. Candidate Items For issuer_notes.md

The following are candidates for a later approved update to issuer_notes.md; they are not updates to issuer memory.

No separate issuer-notes candidate is identified for sovereign-rating transmission because it is principally a ratings and portfolio-monitoring issue rather than a management-strategy item.

5. Monitoring and Next Check

The next quarterly results and earnings call should be assessed as a connected set rather than as individual targets. Priority disclosure requests are: (1) absolute and relative legacy-borrowing reduction, incremental deposit granularity, deposit mix, cost of funds and NIM; (2) early-delinquency, vintage, cure, SMA and segment credit-cost data for unsecured retail and Business Banking/SME; (3) RWA growth, RWA density, RoA, CET1 and retained-capital generation; and (4) product-mix, underwriting, acquisition and pricing commentary that can distinguish disciplined growth from a change in risk appetite.

For rating and market interpretation, future S&P, Moody's, Fitch and domestic-agency publications should be read separately from HDFC-versus-India-sovereign and HDFC-versus-peer spread performance. A sovereign-catalysed widening with intact HDFC fundamentals and peer-relative performance is a different outcome from HDFC underperformance accompanied by weakening standalone indicators.

6. Unverified and Pending Items

The SSC material contains external-research claims and analytical thresholds that were not independently re-verified for this report. In particular, the detailed timing and cost of legacy-borrowing runoff, granular composition and marginal pricing of incremental deposits, product-level delinquency migration, new-vintage performance, underwriting policy changes, incremental RWA by product, internal capital floors, current rating-agency scorecards and matched relative-spread data remain pending.

The cited time horizons and numeric warning lines, including NIM, RoA, credit-cost, CET1 and RWA-density ranges, are discussion constructs. They should guide subsequent evidence gathering and should not be represented as bank guidance, agency criteria or confirmed forecasts.

7. Reference Context