Aadhar Housing Finance — a metronome lending at the bottom of the pyramid

The Pulse

Aadhar Housing Finance is India’s largest low-income housing lender — average loan about ₹11 lakh, to a customer banks mostly ignore — and it just finished FY26 exactly the way it said it would: AUM crossed the ₹30,000 crore milestone (up 20%), profit grew 20% to ₹1,096 crores, gross NPAs printed 1.08%, and every start-of-year guidance number was met. The franchise is almost eerily metronomic — thirteen straight quarters of sequential revenue growth, tax at 22% every period, NPAs that rise and clean up on the same seasonal schedule annually. The two things that actually changed this year are on the ownership and capital side, not operations: promoter Blackstone cut its stake from 75% to 65% in the March quarter, and the balance sheet remains drastically overcapitalised at 42% Tier 1 — a lending engine deliberately running at half throttle. At 2.7 times book for a 16% ROE growing 20% a year, the price is full but not silly.

The Business

The model is a spread business with a twist: lend small and far from the competition. Aadhar writes home loans (73% of the book) and micro loans-against-property (27%) at an average ticket of ₹10.9 lakh and a loan-to-value of 60%, fully retail and fully secured — no developer or corporate exposure at all. The book yields about 13.5% against a cost of funds of 7.7%, an exit spread of 5.8% that would make most lenders blush. The customer is the low-income borrower — 55% salaried, heavy on first-time homebuyers, the segment PMAY subsidies are designed for (Aadhar claims to be the scheme’s largest player, with ~₹50 crores of subsidy routed to 15–16,000 customers).

The defensible part is distribution and underwriting, not the loan itself. Of 626 branches across 22 states, more than 450 sit in “emerging” and deeper locations — small-town India, reached through a four-tier branch format that runs from a 50-square-foot ultra-micro outpost to a full urban branch, each sized to its catchment. Yields out there run 14.5–16% versus 12–12.5% in cities, customers are stickier, and management says new entrants show little traction in the sub-₹15-lakh segment. Underwriting runs on two tracks: salaried files processed centrally for speed and cost, self-employed files judged at the branch by credit managers who can read an informal income — the hard problem of this segment, solved with people plus a TCS-built tech stack. The result is the segment’s paradox: lending to “economically weaker sections” with gross NPAs that have held in a 1.0–1.5% band for years, 60% collateral cover, and collection efficiency above 99%. Geography is the stated risk framework — no state exceeds 15% of AUM, which is why regional scares (Karnataka e-Khata, tariff-hit textile towns, MFI ordinances) keep washing through peers without touching Aadhar’s numbers.

Ownership is the live story. Blackstone (via BCP Topco) took the company public in May 2024 and has been gliding out since — a slow drip to 75.19%, then a 10-point block sale in the March 2026 quarter that took promoters to 64.9%, absorbed largely by public and institutional buyers. None of the four earnings calls addressed it; the exit path of a private-equity promoter is the one material question management hasn’t been asked on record.

How Management Thinks

MD & CEO Rishi Anand and CFO Rajesh Viswanathan run on consistency as a brand. The FY26 closing boast was not a growth number but a delivery record: “On all the 3 critical parameters, we have stood by our guidance and commitment to the market.” The record backs it — the 20–22% AUM, ~25 bps credit cost, ~50 bps cost-to-income improvement and 1.10–1.15% year-end GNPA promises made in July 2025 all landed, the last one beaten at 1.08%.

The risk reflexes are the most distinctive thing about them. Twice during the year — first on US tariffs hitting textile and gems towns, then on the West Asia war — the instinct was to pre-emptively throttle the higher-yield LAP book and put collections on alert before any stress appeared, treating bounce rates as the tripwire. LAP is explicitly framed as a stored option: 400 basis points of extra yield they can dial up “if push comes to shove” to defend spreads, kept below the 30% regulatory cap by choice. They drove 1+ DPD as a managed KPI for eighteen months with ~1,800 collection staff tracking it daily, and refuse rate wars on principle — “we are not in the race of undercutting… we will play on speed.” Candour is decent: one-offs volunteered unprompted, weak geographies (East India, Kerala, Punjab) named without being asked, and a genuinely unusual disclosure that only ~₹28 crores of nine-month profit came from co-lending and assignment upfronting — management quantifying its own earnings quality.

Capital allocation is conservative to a fault. Leverage sits at ~2.5x net worth against rating-agency comfort of 4.5x; the stated plan is to creep to 3x, slowly. Yet they raised another ₹1,000 crores of equity in May 2025 with Tier 1 already above 40%, have never paid a dividend in eight years, and ROE has drifted from 18% to 16% as the idle capital dilutes returns. The funding side is genuinely well run — 42 lender relationships, a CARE upgrade to AA+, cheap NHB money at 6.6–7.2%, a maiden $50 million ECB, NABFID added, and a 74–76% floating match on both sides of the balance sheet that let them pass rate cuts through without spread damage.

Where It’s Going

Guidance for FY27 is the same song: ~20% AUM growth, ~20% PAT growth, 17–18% disbursement growth, another ~50 bps off cost-to-income, and a new stake in the ground of ₹50,000 crores AUM in three years — roughly a re-doubling. The growth machine is branch-led and formulaic: 40–50 additions a year, weighted toward deeper “B and C” emerging locations, with small branches breaking even in 9–12 months; management decomposes its 20% growth as roughly 5–6% ticket-size inflation, 6% maturing-branch productivity, and 5–6% from branches under 18 months old.

The tailwinds are real and named: 125 bps of repo cuts through 2025, GST 2.0 cutting construction-material costs, PMAY 2.0 subsidies flowing with Aadhar first in line. The tensions are equally clear. The easy funding-cost gains are banked — banks have passed through 70–75 bps and the CFO says he’d be “very, very happy” just holding 7.7% — so spreads are guided to compress 8–10 bps a year, defended by the LAP lever and emerging-market mix. Competition is intensifying in urban markets where banks undercut on rate, which is precisely why the strategy keeps pushing down-market. And the ROE story now depends on leverage actually rising — at the current glacial pace, the 17–18% aspiration stays aspirational. Behind it all sits the Blackstone glide path: a promoter at 64.9% and falling means supply of stock, and eventually a new anchor owner, neither of which management has discussed.

The Four Checks

1. Quality and moat. A good business with a real but contestable edge. The moat is distribution depth where competitors are thin (450+ of 626 branches in small-town India), an underwriting machine for informal incomes that produces 13.5% yields at 1.1% GNPA, and a funding cost advantage (AA+, NHB access) smaller rivals can’t match. It is not a fortress — balance-transfer outflow of ~5.6% a year proves customers can be poached, and nothing stops a determined bank from following them down-market. Niche dominance, honestly held: 6.

2. Returns on incremental capital and runway. A retained rupee goes into book growth at ~16% ROE — and that understates the engine, because ROA of 4.4% is exceptional for a lender and ROE is suppressed only by leverage of 2.5x against headroom of 4.5x. At normal gearing this is a 20%+ ROE machine. The runway is long: India’s mortgage penetration is far below global norms, the sub-₹15-lakh segment is underserved, and the company’s own share of its peer pool is ~18%. Held back from an 8 only by the fact that the higher returns remain latent: 7.

3. Capital allocation for the stage. Mixed, leaning rational. The funding diversification, branch economics, tempered co-lending income and pre-emptive risk throttling are textbook. But raising ₹1,000 crores of fresh equity at 44% Tier 1, paying no dividend in eight years, and letting ROE drift down while capital idles is conservatism past the point of efficiency — surplus is being stored, not compounded. Deliberate, defensible for a recently-listed lender building rating credibility, but a real drag: 6.

4. Price. ₹20,504 crores of market cap is 18.5 times earnings and 2.7 times book for a 16% ROE compounding at 20% with best-in-class asset quality. That is full but defensible — cheaper than the premium affordable-housing names have historically traded, paying for the consistency without paying for perfection. The promoter overhang cuts both ways: supply pressure now, cleaner register later: 6.

Sources

• Earnings call transcripts: 25 Jul 2025 (Q1 FY26), 7 Nov 2025 (Q2 FY26), 30 Jan 2026 (Q3 FY26, filed Feb 2026), 5 May 2026 (Q4 FY26/full year) — all BSE filings via screener.in.
• Annual reports: FY24 and FY25 (FY23 was not available on screener). Both extracts were partial — strategy, risk and governance sections parsed; MD&A financial detail largely missing, so hard numbers lean on the concalls and snapshot.
• Screener.in consolidated snapshot, fetched 2026-06-10 (public, logged-out session).
• Gaps: a May 2026 concall row in screener’s documents list had no transcript link (likely a duplicate listing of the Q4 call); the March 2026 promoter stake sale is visible only in the shareholding data — no management commentary on it exists in any document read.
• Research dumps in vault/Sources/Earnings/Aadhar Housing Finance Ltd/ (not published).