Key Takeaways
- Asset quality deliberately worsened, then improved: Gross Stage-3 rose to 1.5% because management let delinquent accounts roll to 90+ days to enable SARFAESI action. November was the peak; December and January show improvement.
- Growth guidance held at ~30%: AUM growth target of 29–31% for FY26 and ~30–31% for FY27, with the ₹30,000 crore by 2030 ambition reaffirmed. Disbursement yield slipped to 14.6% as rate cuts were passed on, but spreads stayed above 6%.
- Stress is a ticket-size story, not a geography or product story: Customers in the sub-₹5 lakh bucket drive delinquency, while MP actually improved 5–7 bps. Home loans and LAP show near-identical NPA ratios.
- Cost of funds tailwind is building: Cost of funds fell 50 bps to 8.3%, with another 10–15 bps expected in 3–6 months even without a rating upgrade; an upgrade would add a further 20 bps.
- What to watch next quarter: Whether the Q4 GNPA decline to 1.3–1.4% materialises, the pace of digital sourcing scale-up toward 10%, and any rating agency action after FY26 results.
Asset Quality: The Deliberate Deterioration and the Turn
- The mechanics of the Stage-3 spike: Management took a "harsh call" to stop holding delinquent accounts at 60–90 days and let them flow into the 90+ bucket, because SARFAESI enforcement requires accounts to cross that threshold. Without this action, Stage-3 would have risen only 10–15 bps; with it, the rise was 25–30 bps.
- The peak is behind us, per management: November was the worst month for delinquencies. December brought a slight improvement, and January continued the trend. Fresh collection efficiency has been improving sequentially from Q1 to Q3.
- Stress concentrates in the smallest tickets: The sub-₹5 lakh loan bucket shows materially higher delinquency than the ₹10–15 lakh average ticket size. Cohorts at 18–24 and 24–30 months show no meaningful difference in behaviour; the first 12 months look distinctly positive.
- No single geography or product is the culprit: MP's NPA ratio improved 5–7 bps versus 3–4 quarters ago. Every state has 3–4 troublesome branches where controls are being tightened. Home loan and LAP NPA ratios are within 1–2 bps of each other.
- The ECL model adds a mechanical headwind: As the interest rate cycle turns down, the ECL model's PD assumptions rise, feeding into higher provisions. Management flagged this as a historical pattern seen during demonetisation, COVID and GST cycles.
- Provision coverage ratio stays steady: The company will maintain PCR at ~24–25% for Stage-3, consistent with its historical approach.
Rupinder Singh: "Unless and until we move to the next bucket, the legal action doesn't come into force for us to enforce SARFAESI. So, this was not initial thought to implement but seeing that the bucket is bulging at that time and customer has not paid that, I think then there is some harsh call which has been taken to let it go to the next bucket."
Growth and the 2030 Ambition
- AUM growth held at ~30%: FY26 is expected to close around 29–31%, with January better than December and early February logins giving a positive signal. FY27 is guided to the same 30–31% range.
- The ₹30,000 crore by 2030 target is reaffirmed: Management sees no reason to deviate from the long-term ambition in the current environment.
- Disbursement growth stays range-bound: Disbursements will remain in the 30–31% band for FY27. The home loan to LAP mix stays at 60:40.
- Branch expansion pace unchanged: 35 branches opened so far this year against a 40–45 target; another 4–5 will come in Q4. The annual cadence of 40–45 branches continues.
- No new states on the horizon: The current 15-state footprint already covers over 90% of India's housing finance market, so medium-term expansion will be within existing geographies.
Rupinder Singh: "Our major prerogative is by 2030 to reach a Rs. 30,000 crores of AUM. So, that thing is very well in line. And we do not see that we have to deviate from that in the current set of scenario."
Margins, Spreads and the Cost-of-Funds Tailwind
- Spreads protected above 6%: Even as cost of funds fell from 8.8% to 8.3%, the company passed on 25–30 bps to new customers, taking the disbursement yield to 14.6%. The spread on fresh disbursements remains above 6%, and the medium-term commitment is to hold that floor.
- NII growth lagged AUM growth for a benign reason: Interest income grew 4% QoQ versus 6% AUM growth because rate cuts were passed on to new borrowers, not because of margin compression.
- Cost of funds has further to fall: Management expects 10–15 bps of improvement in the next 3–6 months even without a rating upgrade. A successful upgrade after FY26 results would unlock another 20 bps.
- BT-out rate is low and stable: At 4.3%, the balance transfer outflow is being managed through a data-driven retention approach, even in a declining rate environment.
- OPEX ratio improving: YTD OPEX growth of 26% is below managed asset growth, pulling the OPEX-to-managed-assets ratio down 20 bps YoY to 4% (ex-labour code impact). A further 15–20 bps reduction is expected as digital sourcing scales.
Ashish Gupta: "We feel that basis our FY26 financial results, we will be approaching to the rating agencies like for a rating upgrade. And we feel that if we get a rating upgrade, there will be a scope of further reduction in the cost of fund from here by about 20 basis points."
The Sub-₹5 Lakh Customer: High Yield, Higher Risk
- A 200 bps yield advantage that management will not walk away from: Sub-₹5 lakh customers deliver significantly higher yields than the ₹10–15 lakh average ticket. The strategy is not to exit but to tighten underwriting.
- The fix is scorecard and BRE tightening: Management will be "more choosy" in this segment, applying stricter scorecards and Business Rule Engine filters rather than shrinking the book.
- Login-to-sanction ratio tells the story: The ratio fell to ~45% as the BRE curtailed customers with weak bureau scores. It is now recovering toward ~50% as customer behaviour improves and December–January trends turn positive.
- Microfinance overlap is small and shrinking: Exposure to microfinance customers is around 4–5% and coming down, though management did not quantify the specific overlap with LAP and home loan borrowers.
Rupinder Singh: "This is a customer which gives a very good yield also. This customer bring almost 200 bps of advantage compared to the... It is not that we are going to leave this customer."
Operations: Digital Sourcing, PMAY and Collections Build-Out
- Digital sourcing is live and scaling: An in-house team of 40–50 people now generates 4–5% of disbursements through social media and alternate channels. The target is 9–10% within a couple of quarters.
- PMAY 2.0 is gaining traction: Over 2,000 customers have received the subsidy as of December 25. The scheme picked up from end-Q1 after initial system issues were resolved; Q2 and Q3 each saw 500–600 customers disbursed. MoHUA is committed to fixing remaining technical issues.
- Collections headcount expanded preemptively: ~400 employees were added in Q3, roughly 200 on the collections side, an action started in Q2 ahead of the stress. Overall collection efficiency remains at 98–99%.
- Branch attrition runs at ~35%: Vintage branches have lower attrition, but management did not provide quarterly trend data or name specific high/low geographies.
Guidance & Commitments
| Commitment | Specifics | Timeline |
|---|---|---|
| AUM growth | 29–31% | FY26 |
| AUM growth | ~30–31% | FY27 |
| Gross Stage-3 | Decline to 1.3–1.4% | Q4 FY26 |
| Gross Stage-3 target | ~1.2–1.25% | FY27 |
| Credit cost | 40–50 bps | FY26 and FY27 |
| Spreads | Maintain above 6% | Medium-term |
| Branch additions | 40–45 per year (35 opened YTD; 4–5 more in Q4) | FY26 and ongoing |
| Cost of funds improvement | 10–15 bps reduction | Next 3–6 months |
| Cost of funds improvement (if rating upgrade) | Additional 20 bps reduction | After FY26 results |
| Digital sourcing share | Reach 9–10% of disbursements (from 4–5%) | Within a couple of quarters |
| OPEX to AUM ratio | Reduce 15–20 bps YoY | From here |
| Home loan to LAP mix | Maintain 60:40 | Ongoing |
| Provision coverage ratio (Stage-3) | Maintain ~24–25% | Ongoing |
| Long-term AUM | ₹30,000 crore | By 2030 |
Key Metrics
| Metric | Q3 FY26 |
|---|---|
| Gross managed assets | ₹10,365 cr, +31% YoY and +7% QoQ |
| Disbursements | ₹977 cr, +11% YoY |
| Portfolio yield | 14.9%, stable YoY |
| Disbursement yield | 14.6% |
| Net interest income | +31% YoY |
| Profit after tax | ₹128 cr, +33% YoY and +5% QoQ |
| Return on equity | 17.1%, up 200 bps YoY |
| Return on assets | 5.8%, up 30 bps YoY |
| Net worth | ₹3,048 cr, crossing ₹3,000 cr mark |
| Bucket cost of funds | 8.3%, down 20 bps QoQ and 50 bps YoY |
| Marginal cost of funds | 8.1%, down 70 bps YoY |
| Spread on fresh disbursements | ~6.5% |
| OPEX to managed assets (excl. labor code impact) | 4.0%, down 20 bps YoY |
| Gross Stage-3 | 1.5% as of 31 Dec 2025 |
| Net Stage-3 | 1.2% as of 31 Dec 2025 |
| Credit cost (9M FY26) | 0.5% |
| Provision coverage ratio (Stage-3) | 25%, stable |
| 1+ DPD | 8.7%–8.8% |
| BT-out rate | Down by 4.3% |
| Branch attrition | ~35%, range-bound |
| Branches added in Q3 | 2, taking YTD total to 35; 4–5 more planned in Q4 |
| Digital sourcing share of disbursements | 4%–5%, target to reach 10% |
| PMAY 2.0 customers receiving subsidy | More than 2,000 as of Dec 2025 |
| Liquidity | ₹486 cr plus undrawn sanctions of ₹1,300 cr+ |
| AUM target by 2030 | ₹30,000 cr |