Key Takeaways
- Disbursements rebounded 24% QoQ despite a deliberate strategic pivot away from sub-₹7 lakh loans and weather disruptions, with management confident of sustaining 25%+ AUM growth.
- Credit cost jumped to 50 bps from 20 bps, driven entirely by a conservative accounting policy change that accelerated write-offs, not by underlying asset quality deterioration.
- The customer profile is being deliberately upgraded: the minimum ticket size floor and a completed data scrub showing only 1.5% MFI overlap signal a push toward better-quality borrowers, even at the cost of near-term disbursement volume.
- Balance transfer runoff is a manageable nuisance, not a threat: ~5% of the 7.5% annualized BT-out rate is customers repaying from their own savings, verified by bureau checks, leaving only ~2.5% as actual competitive loss.
- What to watch next quarter: Whether the credit cost normalizes downward from the one-off 500-day catch-up charge, the pace of disbursement recovery now that the sub-₹7 lakh pipeline is gone, and any early delinquency signals in the non-home loan mix that is inching up.
The Sub-₹7 Lakh Pivot: Strategy Over Stress
- What changed on July 1: The company stopped accepting loan proposals below ₹7 lakhs, a conscious decision to move away from a microfinance-adjacent customer profile and toward better-quality borrowers. Management was explicit that credit delinquency was not the trigger.
- The logic is structural, not reactive: A 15-year home loan with a ₹5-6 lakh ticket produces a small EMI, high collection cost relative to the ticket, and a customer base that keeps growing in volume terms without contributing proportionately to AUM quality. The company wants its average login ticket to rise from ₹10 lakhs now to ₹11 lakhs next year and ₹12 lakhs the year after.
- The near-term cost: The policy change created a short-term drag on Q2 disbursement growth and will weigh on the second half as well, though October showed improvement and the team has adjusted. Management declined to quantify the proportion of sub-₹7 lakh disbursements in prior quarters.
- Customer book scrubbed: A completed scrub of all ~1,70,000 customers showed 55% have no additional loan, 20% have one more loan, and MFI overlap is just 1.5%, reinforcing that the existing book is not a microfinance book in disguise.
Anandan: "We are wanting to move to a better category, better class of customers. And also looking at the transaction size because we are also realizing that this is a home loan of 15 years, the effective tenure of 10 to 11 years. We don't want to get into INR5 lakhs, INR6 lakhs, where your EMI will be less, your collection cost will be more, and it will keep growing."
Credit Cost: A Policy Shock, Not a Portfolio Shock
- The 30 bps jump unpacked: Credit cost rose from 20 bps to 50 bps because the technical write-off threshold was shortened from 2 years to 500 days, with 100% provision required beyond 500 days versus only 25% under the old policy. The Q2 charge was inflated by a one-time catch-up on loans sitting in the 500-day-to-2-year bracket.
- The philosophy is pre-emptive, not remedial: Management has always priced its products assuming a 0.5% credit cost; the past 20-25 bps experience was an undershoot, and the policy change brings reported credit cost in line with the conservative pricing assumption already embedded in yields.
- Normalization is logical but unguided: Management acknowledged that, logically, credit cost should come down from Q2 levels now that the catch-up is done, but it refused to commit to a specific number or a drop back to 30 bps, saying only that it will not be substantially different from current levels.
- Non-home loan mix is a quiet contributor: The non-home loan percentage is slightly going up, and management flagged this as a factor in the credit cost increase, though it did not break out the segment-level impact.
Anandan: "We have consciously provided 0.5% as the credit cost in our pricing of the products. Only thing in the past, we were only actually getting a charge of 25 to 30 basis points. But then to be on the conservative side, we said we will take it up to 50."
Balance Transfer Runoff: Separating Signal from Noise
- The headline number: BT runoff increased from ₹373 crores to ₹463 crores QoQ, an annualized ~7-7.5% of the loan book. Management does not see BT as a major concern and frames the real challenge as enhancing disbursements, not stemming runoff.
- The forensic breakdown: Of the ~7.5% annualized BT-out rate, approximately 5% is customers repaying from their own sources (chits, business savings), verified by a credit bureau check six months later showing 99% of those customers had not taken a loan elsewhere. Only ~2.5% is actual loan transfer to banks or other HFCs, mainly on slightly better interest rates.
- Competitive intensity is contained: The small genuine BT-out slice and management's stated lack of pressure to cut lending rates suggest the company's 15.5-21% yield band is not under severe competitive threat from banks offering lower rates to its customer segment.
Operations & Infrastructure Build-Out
- Branch expansion on track: 20 of the planned 40 new branches for the year were added largely in Q2, with 4 more in Maharashtra and Odisha during October, bringing the total in those two new states to 15. The expansion is a core plank of the 25%+ AUM growth ambition.
- ZIVA LOS stabilized: The new loan origination system, which caused some disruption early in the quarter, is now fully onboarded across branches and is delivering stronger process control, improved data accuracy and better monitoring.
- Collection structure reorganized: Responsibilities are now clearly delineated: branch teams handle up to 12 EMIs, a dedicated collections team takes over at 90-180 days, and an independent legal recovery team is added after 180 days. Middle management has been strengthened at the state level.
- Opex discipline held: Opex as a percentage of AUM was maintained at 2.7% despite increased investments in branches and IT, through consistent productivity improvement across all functions.
- Direct assignment program launched: ₹170 crores executed in Q2, ₹300 crores in H1, cumulative ~₹320 crores (~3% of loan book). The board has authorized up to 10%; management is targeting 6-7% going forward.
Yield Management & Margin Levers
- Minor rate optimization, not a trend: Home extension loans were reclassified to the home loan rate (15.5% from 17.5%), and the SME rate was optimized from 22% to 21%. These tweaks did not materially impact yield.
- No pressure to cut lending rates: Management stated explicitly that it sees no pressure or compulsion to reduce lending rates, implying the yield profile is stable even as the customer profile shifts upward.
- ROE ambition intact: Management expressed high confidence in not just maintaining but improving the 20% ROE, supported by operating leverage from branch expansion, stable yields, and credit cost that is now aligned with pricing assumptions.
Anandan: "We are very confident of not maintaining, improving the 20% ROE. We are very confident."
Guidance & Commitments
| Commitment | Specifics | Timeline |
|---|---|---|
| AUM growth run rate | Sustain 25%+ growth; confident of "reasonably good Q-on-Q growth" from Q3 onward | Q3 FY26 onward |
| Medium-term AUM target | Reach ₹25,000 crores AUM (at 25% CAGR, crosses ~₹26,000 crores in 3 years) | Medium term (~3 years) |
| Branch expansion | Add 40 branches this year; 20 added largely in Q2 | FY26 |
| Average ticket size | Increase from ₹10 lakhs now to ₹11 lakhs next year, ₹12 lakhs year after | Annual, over 2 years |
| Direct assignment mix | Target 6-7% of loan book, not beyond board-authorized 10% | Ongoing |
| ROE | Improve from current 20% level | No date (confident stance) |
| Credit cost | Not substantially different from current ~50 bps level; may be lower but no commitment to a significant reduction | Ongoing run rate |
Key Metrics
| Metric | Q2 FY26 |
|---|---|
| Disbursements | ₹963 cr, +24% QoQ |
| Disbursements growth (H1 FY26) | ~8% YoY |
| AUM growth | 22% YoY, 4% QoQ |
| Total income | ₹554 cr, ~27% YoY growth |
| Net interest margin | ₹389 cr, +27% growth |
| Operating profit | ~₹312 cr, ~27% YoY growth |
| Profit after tax | ~₹227 cr, ~24% growth |
| Profit after tax (H1 FY26) | ~₹446 cr, +26% YoY |
| ROA | 7.9% |
| ROE | 20% |
| Opex as % of average AUM | 2.7%, stable |
| Credit cost (H1 FY26) | 50 bps, up from 20 bps due to accelerated write-off policy change to 500 days from 2 years |
| GNPA | 1.55%, up ~6 bps QoQ |
| Net NPA | 1.17% |
| Cost of borrowing | 8.42%, improved 20 bps QoQ from 8.62%; latest borrowing at ~7.9% |
| Direct assignment executed | ₹170 cr in Q2, ₹300 cr in H1 |
| Liquidity | ~₹1,700 cr, including ~₹1,100 cr undrawn bank sanctions |
| Branch count (Maharashtra and Odisha) | 15 as of 31 Oct 2025, including 4 added in the month |
| BT-out rate | ~7–7.5% annualized; ~5% from customer own sources, ~2.5% actual loan transfer |
| Average login ticket size | ~₹10 lakhs; average sanction disbursement size ~₹9 lakhs |
| Logging productivity | 3 to 3.1 files per sales staff per month at branch level |
| Loan book | ₹11,767 cr |