FULL PUBLISHED RESEARCHEarnings and management commentary analysis
16 November 2026 | Reference price ₹781.45 on 13 November 2026
Investment thesis: A mature high-quality mortgage franchise with temporarily disrupted origination; the question is whether full technology migration restores growth while maintaining returns.
1. Investment case
Key takeaways
- Can Fin Homes is an established mortgage lender with high return on equity and relatively low reported impaired assets.
- Q2 loan book ₹43,539 crore versus ₹39,657 crore last September (+9.8%) and ₹42,961 crore June (+1.3%).
- Disbursements ₹2,373 crore fell 6.8% YoY and 9.0% QoQ; PAT ₹275 crore rose 9.6% YoY.
This quarter's main investment question is whether origination weakness is transitory as Project Tejas technology migration concludes. Management says all branches went live by 1 October, with disruption in rollout months, and retains FY27 disbursement guidance ₹13,000 crore. This is a testable management assertion, not an achieved recovery. At ₹781.45 on 13 November, reported Q2 EPS ₹20.62 implies 9.47x annualized run-rate; the valuation is sensitive to whether book growth accelerates without losing underwriting. [1,2]
2. Financial bridge
Key takeaways
- YoY operating profit +₹20 crore, provision +₹5 crore, PBT +₹15 crore; PAT +₹24 crore, implying lower tax or other rounding/scope effects.
- Sequential operating profit +₹3 crore and credit expense down ₹5 crore, PBT +₹8 crore.
- NII +1% QoQ despite lower origination because interest is earned on the existing book.
ROA 2.42% vs 2.39% June and 2.46% prior year; ROE 17.12% vs 17.15% and 18.41%. NIM 3.89% vs 3.81% June, 4.02% prior year. Thus profitability is stable rather than accelerating; ROE has declined YoY despite higher PAT. Distinguish stock, flow and average earning assets. [1, p24–25]
3. Technology transition and growth
Key takeaways
- Management describes Project Tejas rollout: 11 branches July, 89 August, 77 September, remaining 98 from 1 October, with LOS/LMS migration fully live.
- Implementation months reduced branch disbursements; the legacy data migration delayed monthly handover.
- A state registration verification process in Telangana also delayed property documentation; the lender changed process to verify registrability upfront.
These explanations are plausible, but the next quarter's branch data must establish actual throughput improvement. Karnataka monthly disbursement ~₹230–250 crore per call. H1 disbursements ₹4,982 crore (₹2,609 + ₹2,373), leaving ₹8,018 crore to meet FY27 guidance, or ₹4,009 crore per H2 quarter—~69% above Q2. This is a material acceleration requirement. The annual target should not be treated as base certainty.
March 2028 roadmap aims to shift incremental origination outside South to 40% vs 34% September, non-housing AUM to 20% vs 16%, self-employed/nonprofessional to 35% vs 33%. These are mix targets, not outcomes.
Diversification can grow TAM and yield, but borrower cash-flow and title verification differ. A large established salaried portfolio is a credit-quality advantage; faster non-salaried growth needs vintage monitoring.
Technology should improve turnaround and auditability, not weaken approval controls. [1,2]
4. Credit and vintage
Key takeaways
- Gross NPA ₹392 crore vs ₹376 crore June and ₹373 crore prior year; ratio 0.90% vs 0.87% and 0.94%.
- The absolute impaired stock increased ₹19 crore YoY while ratio improved 4 bps, showing denominator effects.
- Net NPA excluding management overlay ₹185 crore vs ₹171 crore June and ₹191 crore prior year; ratio 0.42% vs 0.40% and 0.48%.
Q2 ECL provision ₹8 crore, down from ₹13 crore June. Gross and net ratios have distinct overlay basis; do not infer direct coverage without the provision schedule. A 50-bp annual loss stress on ₹43,539 crore book is ~₹218 crore before tax, 20% of annualized Q2 PAT before tax; actual average asset/credit definitions differ. A mortgage lender's low loss history cannot eliminate collateral, documentation or affordability risk. Track absolute arrears, vintage loss, registration rejection and collections, particularly in newer regions and self-employed borrowers. [1]
5. Funding and returns
Key takeaways
- Yield 9.84% vs 9.81% June and 10.08% prior year; NIM improved sequentially but lower YoY.
- A 25-bp annual margin shock on closing loan stock ~₹109 crore pretax, material relative to annualized PBT ₹1,388 crore.
- Actual repricing and funding mix must be independently reconciled.
Long-tenor mortgages face competitive refinancing and reset risk; a technology project can increase short-term expense before productivity gains. Current ROE 17.12% and ROA 2.42% indicate an established return engine; the goal is to preserve these while growing faster.
The prior five-year trend shows AUM ₹42,209 crore FY26 and PAT ₹1,086 crore, compared with FY22 AUM ₹26,711 crore and PAT ₹470 crore; historical growth has been healthy but should not be extrapolated from the fastest years. At current quarterly PAT, annualized ₹1,100 crore only slightly above FY26. Thus growth acceleration is essential to a higher earnings trajectory. [1]
6. Valuation framework
Key takeaways
- Reference ₹781.45, Q2 EPS ₹20.62, annualized ₹82.48, indicative P/E 9.47x.
- Bear assumes 10% lower normalized EPS at 8x; base 12% growth at 11x; bull 23% growth at 14x.
- These are illustrative analyst scenarios, not observed peer multiples or formal targets.
The base needs origination recovery and stable credit; the bull requires conversion toward the ₹13,000 crore FY27 aspiration and durable productivity. A slower H2 disbursement recovery would constrain future interest-earning assets even if Q3 PAT remains supported by the existing book. A valuation cross-check should use audited BVPS, retention, ROE and cost of equity; no unverified current book multiple is invented. A mature lender at a modest earnings multiple can compound if returns stay high, but low growth and competitive margin pressure could justify a persistent discount.
Institutional deep dive — quantify technology recovery rather than presume it
The disbursement hurdle and book-growth lag. FY27 guidance ₹13,000 crore after H1 ₹4,982 crore requires ₹8,018 crore H2; the average quarterly requirement ₹4,009 crore is 69% above Q2 and 54% above Q1. Build an explicit recovery test rather than simply label the quarter transitory.
These are analyst paths, not management forecasts. Gross disbursement is not net book growth: Q2 loan stock increased ₹578 crore despite ₹2,373 crore gross flow, a gap of ₹1,795 crore from scheduled repayment, prepayment, transfers and other movements.
If the same gross-to-net difference were repeated, a ₹4,009 crore quarter could add roughly ₹2,214 crore net stock, but repayment rises with book and timing varies. Do not forecast using that fixed gap.
The key insight is that even strong H2 approvals can take several quarters to materially lift NII because disbursed balances earn over time. A December quarter recovery in origination should be observable before assuming a full earnings acceleration.
Approval quality and workflow throughput. Q2 approvals ₹2,604 crore and disbursement ₹2,373 crore imply a same-quarter flow ratio 91.1%; Q1 2,609/2,795 ≈93.3%. This is not a matched application conversion rate because sanctions and disbursements can occur in different quarters.
The correct operational measure tracks cohorts from lead, title/registration checks and credit approval to final funding. Project Tejas was intended to improve end-to-end visibility, but a higher apparent approval-to-disbursement ratio can be caused by clearing old backlog rather than new acquisition.
Demand monthly new applications, funded cohort timing, backlog aged by branch and source, registration rejection and post-funding documentation completeness. A strong test is sustained new origination after the October 1 full migration rather than a one-time clearing of September pending files.
Branch and geographic economics. The migration sequence 11+89+77+98 = 275 branches is a rollout count per call, not proof of current total footprint or independently measured productivity. On that denominator, Q2 gross disbursement ~₹8.6 crore per rollout branch; the guidance catch-up average ~₹14.6 crore per quarter.
This illustrates the substantial throughput change needed even if the existing network is fully operational. Geographic mix outside South targeted at 40% of incremental origination by March 2028 from 34% September.
A diversified origination base reduces concentration, but expansion should include independent title and legal diligence in unfamiliar markets. The self-employed and non-housing mix targets increase risk-adjusted yield opportunity while changing borrower cash-flow volatility.
Compare product contribution after funding, expected loss, legal cost and collections, not merely yield or share of disbursement.
Income and margin engine. Q2 NII ₹431 crore annualizes ₹1,724 crore; operating profit ₹355 annualizes ₹1,420 crore; PBT ₹347 annualizes ₹1,388 crore. A 25bp gross annual spread shock on ₹43,539 crore closing loan stock ≈₹109 crore pre-tax, 7.8% of current annualized PBT.
If an additional 25bp loss shock occurs, combined gross sensitivity ~₹218 crore, 15.7% of run-rate PBT, before tax and offsets. A faster-growing book may offset margin pressure in absolute income, but lower incremental spreads reduce economic return.
A stable NIM requires timely repricing and disciplined customer pricing; falling liability rates do not necessarily improve spread if attractive customers refinance elsewhere.
Q2 operating profit/NII 82.4% vs Q1 82.4%, broadly stable. The difference between these lines is not exact operating expense because other income can contribute. This suggests the operational transition has not yet established a step-change in reported profit conversion.
Project Tejas benefits should appear in faster sanction-to-funding, lower rework and fraud, lower cost per completed loan and better servicing. Require a specific technology expense bridge, implementation amortization, recurring savings and branch staffing.
Cutting processing time while compromising title/document checks could create future loss. The correct success metric is risk-adjusted loan contribution, not platform go-live.
Credit and overlays. The gross NPA stock ₹392 crore increased from ₹373 crore prior year despite ratio improvement. Implied gross NPA increment ₹19 crore is small relative to loan growth but underscores why ratio alone is insufficient.
Net NPA excluding management overlay ₹185 crore implies difference ₹207 crore, but that difference is not an exact stage-specific provision because net measure definitions and denominators can differ. Track actual stage-specific coverage, write-offs, recoveries and management overlay stock/flow.
A future release of overlay may improve accounting earnings without representing recurring risk-adjusted profit. The credit case remains attractive if low absolute slippage persists during origination acceleration; the risk is an increase in new-state/title or self-employed delinquency before aggregate GNPA reflects it.
Return and valuation hurdle. Historical FY26 PAT ₹1,086 crore vs Q2 annualized ₹1,100 crore shows limited improvement in current earnings run-rate despite strong historical expansion. At reference ₹781.45 and annualized Q2 EPS ₹82.48, a 10% decline in earnings with unchanged 9.47x run-rate multiple gives ~₹703; at an 8x multiple ~₹594 (existing bear case).
If sustainable EPS grows 12% but market multiple stays 9.47x, price ~₹875; the base valuation scenario of 11x additionally requires multiple expansion. Make that distinction explicit: earnings growth and re-rating are separate sources of return.
An illustrative sustainable ROE 14%/17%/19%, cost of equity 15%/14%/13%, long-run growth 5%/6%/7% yields terminal P/B ~0.9x/1.38x/2.0x. Actual book and share count remain necessary. The most credible catalyst is actual post-migration net book growth, not optimism about technology alone.
Investment decision checkpoints: assess Q3 gross flow versus ₹3,000 crore gradual-recovery path and ₹3,800 crore catch-up path, then track actual net asset addition, approval conversion, current NIM and NPA stock. A successful rollout should generate persistent operating benefit across cohorts.
If volumes remain near Q2 after all branches are live, reduce the growth assumption rather than shift the target repeatedly. If volume recovers at lower spread or deteriorating early arrears, assess the net return before crediting a re-rating.
7. Catalysts and evidence limits
Key takeaways
- Monitor October–December branch throughput after full migration, disbursement and approvals conversion, Telangana documentation cycle, monthly Karnataka, H2 trajectory versus ~₹8,018 crore required, NIM ~3.89%, gross NPA stock and ratio, actual technology expense savings and new mix.
- A strong positive signal is disbursement normalization within the first full post-migration quarter without a spike in early arrears.
- Falsifier: continued disbursement weakness attributed to migration after rollout completion, or growth achieved by looser self-employed credit or lower pricing.
Original investor presentation and transcript support report; independent branch/vintage dataset, full statutory funding/ALM and current book/share/peer valuation audit remain open. [1,2]
Original source ledger
[1] Q2 FY27 investor presentation (2026-10-14): original filing.
[2] Q2 FY27 earnings call (2026-10-12): original filing.
Key takeaways
- Prices: Sage PRO canonical close, 13 November 2026.
- All scenarios illustrative; no formal target or independently certified full QA.
- Open items: Branch productivity/vintages, audited funding/ALM and matched peer valuation.