Key Takeaways
- Microfinance stress defines the quarter: Guardrail implementation and cautious lending drove a deliberate slowdown, with group loan degrowth and credit cost guidance raised to 2.3%-2.5% for FY25. Management sees no MFI revival before Q4.
- Secured pivot is accelerating: Affordable housing, gold loans, vehicle finance and MSME all grew, pushing the secured mix toward the 40% target by year-end. The trade-off is NIM compression to 8.6%-8.8% for the full year.
- Return ratios will soften before they recover: ROE will dip in FY25 as credit costs bite and secured yields dilute margins, but management targets 18%-20% ROE in FY26 once microfinance springs back.
- The collection engine is strained but not broken: Non-delinquent collection efficiency slipped to 99.23%, and slippages are running at ~0.8% per quarter, with 80% of them from microfinance. Management expects efficiency to improve from Q3.
- What to watch next quarter: Whether collection efficiency actually recovers in Q3, the trajectory of credit costs as Q1-Q2 delinquencies season into NPAs, and any signal on when new customer acquisition in MFI can resume.
Microfinance: The Core Problem
- Deliberate contraction underway: The bank restricted disbursements to new-to-credit customers, stopped new customer acquisition in stressed clusters, and acquired only 1.5 lakh new group/individual loan customers with better credit behavior in Q2. Repeat loans to good-quality existing customers now dominate the book.
- Guardrails bite: New-to-bank customers now face a cap of 3 existing MFI lenders and Rs. 1.5 lakh MFI exposure. Pincode-level and center-wise decisioning applies even for repeat and top-up loans. These norms are compressing volumes by design.
- Stressed branches on a leash: Across 5-6 states including Western UP, Kerala, Gujarat, Tamil Nadu, Punjab and Haryana, branches are barred from acquiring new customers or new-to-bank/new-to-credit customers. Customers showing delinquency signs get no repeat loans or top-ups.
- Ticket size is a repeat of post-COVID: With new customer acquisition shut, the blend shifted toward repeat borrowers, pushing the average group loan ticket size up. Management expects this to reverse once NCA reopens, exactly as it did after COVID.
- No timeline for normalization: Management declined to guide on group loan growth or degrowth, calling the situation "still ambiguous" and "fluid." Business revival is penciled only from Q4 FY25 at the earliest.
Nautiyal: "We believe that business revival on a reasonable estimate will only be visible from Q4 FY25 onwards at the earliest." [4-5]
Credit Cost & Asset Quality Mechanics
- The lag problem: Credit cost trails delinquency by 90-180 days. Borrowers who slipped in Q1 and Q2 will season into GNPA and provisioning buckets in H2, so higher credit cost is locked in for the second half even if collections improve now.
- Full-year credit cost raised to 2.3%-2.5%: The increase reflects the delinquency build-up already in the system, not a fresh deterioration forecast.
- Slippages are overwhelmingly MFI: About 80% of overall slippages come from microfinance. The remaining secured and other unsecured books are contributing far less stress.
- Collection efficiency dipped but not collapsed: Non-delinquent collection efficiency fell from 99.46% to 99.23%. Management expects the guardrails and policy changes to settle customer behavior, with efficiency improving from Q3.
- MSME: a tale of two books: The old ~Rs. 600 crore pandemic-era book carries delinquency, but the recalibrated product and policy book built over the last 18 months (now 40% of MSME) has zero 1 DPD+ cases.
- Micro mortgages are pristine: A book built over 18 months with 50% graduating individual loan customers, only 1% new-to-credit, and high bureau cutoffs shows 99.5% on-time repayment and near-100% full-month repayment. Only 9-10 customers out of 8,000 are delinquent.
- Floating provision is locked: The floating provision can only be used in extraordinary circumstances with prior RBI approval, taken for unforeseen events like a pandemic. It is not available for normal credit cost absorption and is currently used only for PCR and NNPA computation.
Goel: "Credit cost typically comes with a 90 to 180 days lag from the date when a customer has gone into delinquency."
Secured Portfolio: The Offset Story
- Growth is broad-based: Affordable housing grew on systematic group marketing, deeper penetration in higher per-capita-income markets, and a stronger ground team. Vehicle finance leveraged upgraded infrastructure. Gold loans saw over 40% of monthly business from referrals. MSME stabilized with all products, policies and systems in place.
- Mix shift is deliberate and fast: The secured book is on track to reach around 40% of the portfolio by year-end, growing 40%+ for the full year. This is the bank's answer to MFI concentration risk.
- The NIM trade-off: Yields declined marginally as the secured mix grew, and further compression is expected. Full-year NIM is guided to 8.6%-8.8%, a little below 9%.
- New product launches fill the stack: Elite LAP, bank guarantees, and working capital term loans now make the MSME offering comprehensive. The AD-1 license, received during the quarter, unlocks retail foreign currency deposits, remittances, currency exchange, and trade finance.[5-6]
- Profitability is segmented, not binary: Housing is already in profit. MSME is likely to break even by year-end. Gold loan, agri and vehicle finance have just started and need a 12-month period to show profit. There is no simple yes/no on overall secured profitability.
- OPEX reflects the pivot: Personal expenses rose 31% on salary corrections and ~2,700 additional employees, mainly for secured business and 123 new branches. Fixed costs increased on occupancy, rent, electricity, maintenance, IT and depreciation for those branches.
Deposits, Fees & Other Levers
- CASA traction through value-add: Maxima SB book is growing, and the Non-Resident Maxima variant launched in Q2. The Navratna Family Program targets HNI customers and their families. The strategy emphasizes wallet share of existing customers alongside quality acquisition.
- MAB charges gave a one-time spike: New minimum account balance charges introduced in Q1 resulted in a Rs. 12 crore QoQ spike in miscellaneous income. This will slowly decline as customers start maintaining balances, but will persist every quarter at a lower level.
- Fee income pipeline building: Mutual fund distribution and ASBA facility will be added during this financial year, along with other insurance product offerings. The AD-1 license opens foreign exchange and trade finance fee streams.[5-6]
- Bad debt recovery target: The bank aims to collect over Rs. 100 crores in bad debt recovery this year. Separately, Rs. 50 crores of Assam portfolio (NPA since 2019) was written off during Q2.
Guidance & Commitments
| Commitment | Specifics | Timeline |
|---|---|---|
| Secured book mix | Reach around 40% of portfolio | End of FY25 |
| Secured portfolio growth | Grow 40%+ for the full year | FY25 |
| Individual loan growth (MFI) | Grow 17% for the full year | FY25 |
| Full-year credit cost | 2.3% to 2.5% | FY25 |
| Full-year NIM | 8.6% to 8.8% | FY25 |
| Slippages per quarter | ~0.8% for remaining two quarters | Q3 and Q4 FY25 |
| Collection efficiency recovery | Improve from Q3 onwards | Q3 FY25 |
| PCR floor | Remain above 70% | Ongoing |
| Bad debt recovery | Collect over Rs. 100 crores | FY25 |
| Gold loan branch coverage | Cover over 200 branches | End of Q3 FY25 |
| MSME break-even | Likely to break even | End of FY25 |
| ROE target | 18% to 20% | FY26 |
| Universal Banking License | Under active consideration, likely this financial year | FY25 |
| Mutual fund distribution and ASBA | Launch during this financial year | FY25 |
| Cost-to-income ratio | Keep in control in upcoming quarters | Ongoing |
| PCR computation | Floating provision utilized only for PCR and NNPA computation | Ongoing |
Key Metrics
| Metric | Q2 FY25 |
|---|---|
| Total disbursements | ₹5,376 cr, up 2% QoQ, down 6% YoY |
| Loan book | ₹30,344 cr, up 14% YoY and 1% QoQ |
| Total deposits | ₹34,070 cr, up 17% YoY and 5% QoQ |
| CASA book | ₹8,832 cr, 25.9% of total deposits, up 26% YoY and 6% QoQ |
| Retail deposits | ₹24,746 cr, 73% of total deposits, up 32% over Sep'23 |
| Net interest margin | 9.2%, full year expected around 8.6% to 8.8% |
| Cost of funds | 7.5%, similar to last quarter |
| Cost to income ratio | 60% |
| PAT | ₹233 cr |
| ROA | 2.2% |
| ROE | 15.7%, FY26 target 18% to 20% |
| Secured portfolio share | 34.9% of total book, expected to reach around 40% by FY25 end |
| GNPA / NNPA | 2.5% / 0.6% |
| PAR (overall) | 5.1% vs 4.2% in Jun'24 |
| PAR 0 (group loan portfolio) | 5.5% vs 4.1% in Jun'24 |
| Slippages | ₹243 cr vs ₹192 cr in Q1 FY25; microfinance slippages at 0.99%, secured at 0.40% |
| Credit cost (quarter) | ₹151 cr vs ₹110 cr in Q1 FY25 |
| Full year credit cost guidance | 2.3% to 2.5% |
| Write-offs | ₹140 cr during Q2, including ₹50 cr of legacy Assam portfolio |
| Bad debt recovery | ₹25 cr in Q2 FY25, full year target over ₹100 cr |
| Non-delinquent portfolio collection efficiency | 99.23%, dipped from 99.46% |
| Overall collection efficiency (due vs collection) | 97% |
| Borrowers with 4+ lenders | 7% of borrowers, contributing about 10% of slippages |
| Individual loan GNPA | Below 2% |
| Affordable housing and micro-mortgages book | ₹5,784 cr, up 43% YoY, adding ₹585 cr vs Jun'24 |
| Micro-mortgages book | ~₹400 cr vs ₹76 cr in Sep'23 |
| MSME book | ₹1,514 cr, up 5% YoY and 7% QoQ |
| FIG book | ₹2,042 cr, 7% of total asset book, up 57% YoY and 13% QoQ |
| Vehicle finance book | ₹262 cr, up 105% YoY and 20% QoQ |
| Gold loan book | ₹62 cr, growing over 100% QoQ |
| Secured portfolio growth guidance | 40% plus for full year |
| Individual loan (microfinance) growth guidance | 17% for full year |
| Collection team strength | 2,200+, added over 300 personnel during Q2 |
| PCR | 78%, approximately 97% including unutilized floating provision |
| Floating provision | ₹250 cr total, ₹100 cr unutilized |
| Dividend paid | ₹290 cr, causing dip in capital adequacy ratio |