Key Takeaways
- Steady quarter, broad-based momentum: AUM grew 27% YoY to ₹18,010 crore, PAT rose 29.1% YoY, and ROE improved 40 bps sequentially to 14.9%, all while asset quality held stable.
- Spreads defended, not expanded: Spread held at 5.2% and management recommitted to a 5.0%-5.2% band, ruling out a return to post-COVID highs. A 20-25 bps ROA hit from insurance fee changes must be absorbed over time.
- Operating leverage is kicking in: Cost-to-income improved 190 bps QoQ to 30.8%, helped by a seasonal lull in campus hiring; the structural target is a 10 bps annual reduction in opex to assets.
- The turnaround story is intact: Last year's problem markets (Tamil Nadu, Gujarat) are recovering, team stability is bearing fruit, and disbursement growth has re-accelerated after external and internal headwinds faded.
- What to watch next quarter: Whether co-lending volumes rebound after the RBI-guideline reset, the pace of Tamil Nadu's recovery, and any borrowing-cost transmission that forces a PLR increase.
Spread, Margin & the Insurance Overhang
- Spread commitment hardened: Management explicitly committed to a 5.0%-5.2% spread band. The book is fully floating-rate, so any borrowing-cost increase will be passed through to customers to defend that band.
- No return to post-COVID highs: Spreads of 5.5%+ were described as a short-lived anomaly when borrowing costs fell sharply; the company had always flagged them as unsustainable and sees 5.0%-5.2% as the normalised range.
- Book yield dipped 10 bps QoQ: A residual effect from an earlier phase when rates were expected to fall and the company passed lower rates to new customers; described as quarter-on-quarter fluctuation, not a trend.
- Yield stability on larger tickets: Origination yield held at 13% despite a mix shift toward loans above ₹25 lakh, because the borrower profile is unchanged (higher incomes but the same access challenges); yield compression in the larger-ticket segment is marginal.
- Insurance fee impact quantified: The regulatory change on insurance distribution fees translates to a ~35 bps ROA headwind. Management is hopeful ~10 bps will eventually return via regulatory concession, leaving a 20-25 bps net gap to fill.
- Mitigation levers identified: The gap will be closed through a combination of increasing LAP share (currently low, within the 20% non-housing guardrail), increasing direct assignment income, and slight pricing or fee adjustments. No single lever is oversized.
- PLR increase is data-dependent: No PLR hike taken yet because borrowing cost has not moved post the repo rate change. Management needs 10-15 bps of actual borrowing-cost movement to trigger a PLR change; a 25 bps repo hike translates to ~10 bps transmission, so a cumulative 50 bps hike would drive ~20 bps over 3-6 months.
Viswanathan: "We are running a fully floating rate book, and we are committed to maintaining a 5.0% to 5.2% spread, which is what we have always articulated. If the borrowing cost moves up, we will pass it on to the customer eventually. Ultimately, we have to maintain the 500 to 520 basis points spread."
Asset Quality & Credit Performance
- Gross Stage 3 improved 10 bps QoQ to 1.7%: Early delinquency indicators were flat sequentially (1+ DPD at 4.7%, 30+ DPD at 3.2%), signalling stable portfolio quality with no new stress formation.
- Credit cost contained at 40 bps: Consistent with the company's target range and supportive of the ROE improvement story.
- Improvement driven by slower inflows into delinquency: The 3-4 quarter trend of improving asset quality reflects prolonged stability without major macro shocks; new customers are entering the delinquency funnel at a slower pace, not a dramatic cure of the existing stock.
- Drought impact absent in key markets: MP and Gujarat showed no asset-quality deterioration because the customer base is in manufacturing and services, not agriculture. Management framed regional shocks as normal in a diverse country.
- Multi-credit concerns dismissed for this book: Periodic portfolio scrubs over three years show the ratio of customers taking other products (gold loans, personal loans) is static and known at origination. The broader industry discussion about borrowers using one source to pay another likely relates to a personal-loan-dependent cohort, not the home-loan customer base.
- Bounce rate optics explained: October bounce was higher than September, but that is a calendar effect (the 4th falling on a weekend); year-on-year, this October's bounce was lower than last October's.
- ECL coverage on Stage 3 ticked up slightly: A routine quarter-to-quarter movement from updating probability-of-default and expected-loss models, now recalibrated every quarter instead of less frequently.
Viswanathan: "This discussion which is going on around customers borrowing from one source to pay off another could be mostly a different cohort of customers, customers who are mostly dependent on personal loans, etc, and may not be directly related to home loan."
Growth Recovery & Geographic Mix
- Disbursement growth re-accelerated to 32.2% YoY: The recovery is a tale of two halves. Three quarters ago, external issues (tariffs impacting Gujarat and Tamil Nadu) and internal leadership challenges in some markets weighed on growth. Both resolved around last October, and ground-level demand is now good.
- Tamil Nadu returned to growth: 12% increase in Q2 after being marginally down YoY in H1, reflecting early signs from a strengthened team. Southern markets are self-construction-driven, so turnaround takes a couple of quarters to show full benefit versus apartment-driven markets where sale and disbursal are quicker.
- Core markets performing strongly: Maharashtra, Gujarat, and Madhya Pradesh delivered strong growth versus H1 last year with healthy portfolio quality. UP, Rajasthan, NCR, and Karnataka recorded strong YoY growth.
- Branch network expanded to 181: Six branches added in Q2; 10-15 more planned in H2 FY27. A couple of underperforming branches in Rajasthan were closed, described as routine annual pruning.
- Team stability cited as the growth engine: Stability across most states over the last 12-18 months is bearing fruit; the earlier leadership challenges in some markets were addressed a few quarters back.
- UP scale-up on track: Plans to strengthen the team and take the state to the next growth level are progressing.
Viswanathan: "That is the nature of a diverse country, where we are operating in multiple states. There is always a drought in some place, there is some rain in some place. Those things keep happening."
Operating Efficiency & the AI Build
- Cost-to-income improved 190 bps QoQ to 30.8%: Opex to assets improved to 2.6% from 2.8% QoQ. The sequential drop was partly seasonal: Q2 and Q3 see a lull in new employee joining after a large campus batch joins in Q1, so employee cost fell ~7% QoQ.
- Structural target is 10 bps annual reduction in opex to assets: This year's target is 2.6%, and the company is on track. Q4 and Q1 will see the next campus batches join, so employee cost will rise again in those periods.
- Four in-house AI capabilities now live: Cue (records and analyses 95% of customer calls), ClariFi (single case page for underwriters on 100% of loan files), BankLens (bank statement to cash-flow view in ~1 minute), and DocIQ (reads all underwriting and operational documents). These are point solutions today.
- AI roadmap targets a unified platform: Over the next 18 months, the plan is to move from point AI solutions to one shared data and decision platform, extending AI into sourcing, collections, and servicing in phases.
- AI impact on opex is aspirational, not committed: Management is working toward the 10 bps annual reduction; if AI delivers on its promise and can be better leveraged, there could be downside to that number, but no separate AI-driven target is set.
- Balance transfer outflows reduced: BT outflows fell from 5%-6% to ~4.5% through a structured intervention: engage the customer to understand reasons, explain the cost of transferring, offer a top-up if needed, then offer a rate break. Expected to remain in the 4%-6% range.
Co-lending, Assignments & the Liability Side
- Co-lending volumes collapsed this quarter: Transactions fell to ₹1.8 crore from ₹46 crore QoQ because working through new RBI guidelines with partners took more time than anticipated.
- Recovery expected in Q3: The company is almost at the stage where the guideline issues are resolved; an uptick in numbers is expected in the coming quarter. Infrastructure and operational capabilities are being strengthened with partners.
- Direct assignments executed at ₹281 crore: This channel provides additional income and is one of the levers identified to offset the insurance fee impact.
- Borrowing cost flat QoQ: No transmission of the repo rate change has yet flowed through to the company's cost of funds, which is why no PLR increase has been taken.
Guidance & Commitments
| Commitment | Specifics | Timeline |
|---|---|---|
| Spread band | Maintain 5.0% to 5.2% spread; pass through borrowing-cost increases to customers | Ongoing |
| AUM growth | 25% growth to ₹20,000 crore; 1%-2% upside possible with tailwinds | March 2027 |
| ROE exit | 15% ROE exit (FY27); 16% exit (FY28 March); current momentum suggests likely to exceed | March 2027 / March 2028 |
| Opex to assets | Reduce by 10 bps annually; target 2.6% this year | FY27 (March 2027) |
| Credit cost | Around 40 bps; gross Stage 3 maintained around 1.7% | Ongoing |
| Branch expansion | 10-15 more branches added | H2 FY27 (by March 2027) |
| Tamil Nadu turnaround | Growth to strengthen further as teams bed in | Q3 and Q4 FY27 |
| Co-lending recovery | Uptick in volumes as RBI guideline issues are resolved | Q3 FY27 |
| PLR increase | Triggered when borrowing cost moves 10-15 bps; currently flat | Data-dependent |
| Insurance fee gap | Make up 20-25 bps ROA impact through LAP, assignments, pricing | Over time (no date) |
| AI platform | Move from point solutions to unified data and decision platform; extend to sourcing, collections, servicing | Next 18 months |
| CEO tenure | Manoj Viswanathan intends to remain available for reappointment beyond current term | Term ends July 2028 |
Key Metrics
| Metric | Q2 FY27 |
|---|---|
| AUM | ₹18,010 cr, up 27% YoY and 6.3% QoQ |
| Disbursements | ₹1,705 cr, up 32.2% YoY and 4.7% QoQ |
| Total income | ₹567 cr, up 18.4% YoY and 5.0% QoQ |
| Net interest income growth | 31.9% YoY |
| Profit after tax | ₹170 cr, up 29.1% YoY and 6.5% QoQ |
| Pre-provision operating profit | ₹238 cr, up 26.1% YoY |
| Return on assets | 4.2% |
| Return on equity | 14.9%, improved 40 bps QoQ |
| Spread (excl. co-lending) | 5.2% |
| Net interest margin | 6.0% |
| Cost-to-income ratio | 30.8%, improved from 32.7% in Q1 |
| Operating cost to assets | 2.6%, improved from 2.8% in Q1 |
| Cost of borrowing | ~7.8%; incremental borrowing cost at 7.5% |
| Origination yield (excl. co-lending) | 13% |
| Disbursement yield | 13% |
| Gross Stage 3 | 1.7%, improved 10 bps QoQ |
| 1+ DPD | 4.7%, unchanged QoQ |
| 30+ DPD | 3.2%, unchanged QoQ |
| Credit cost | 40 bps |
| Stage 3 provision coverage | 23.6%; total coverage including management overlays at 46.4% |
| Capital to risk-weighted assets | 39.8%; Tier 1 at 39.4% |
| Net worth | ₹4,665 cr |
| Book value per share | ₹446 |
| Branch network | 181 branches, 379 touch points; 6 branches added in Q2 |
| Individual housing loans share | 84% of portfolio |
| Co-lending portfolio | ₹593 cr, 3.3% of AUM; co-lending disbursements at ₹1.8 cr in Q2 |
| Direct assignment transactions | ₹281 cr executed in Q2 |
| Average ticket size (new originations) | ~₹15 lakhs to ₹16 lakhs |
| Balance transfer outflows | ~4.5%, down from 5–6% earlier; expected range 4–6% |
| Top-up contribution to disbursements | 2–3% |
| Active connectors | ~4,400 |
| AUM growth target | 25%, implying ₹20,000 cr by Mar 2027 |
| ROE exit target | 15% for FY27, 16% for FY28 |