The aggregate numbers on Indian housing finance look unremarkable in mid-2026. Total housing loans outstanding have grown around 11-12% year-on-year, a deceleration from the high teens we saw in 2023-24 but still respectable relative to overall credit growth. The headlines mostly write themselves.

Underneath the aggregate, three more interesting shifts are playing out. Anyone trying to forecast the housing finance segment, or to position a portfolio against it, needs to look past the topline number.

The composition shift between banks and HFCs

For most of the last five years, banks have grown housing loan books faster than housing finance companies (HFCs). The reasons were straightforward: lower cost of funds, easier access to LCR-eligible deposits, regulatory capital advantages on housing as a priority sector category.

Through the first half of 2026 that gap has narrowed. A few of the larger HFCs have grown disbursements faster than the bank average, particularly in the affordable housing and self-employed segments where their underwriting capability has historically been stronger. The reasons are partly cyclical (banks have tightened underwriting on certain self-employed segments after some sour vintages) and partly structural (HFCs have invested in segment-specific origination capability that banks have mostly not built).

Whether this gap continues to narrow depends on how the Reserve Bank of India handles HFC regulation through the rest of the year. The pendulum has swung back and forth multiple times. The current regulatory stance is broadly accommodative of well-capitalised HFCs, but the segment is one regulatory tightening away from a different growth picture.

The geography of where loans are actually growing

The mortgage book at any given Indian bank is heavily concentrated in metros and tier-1 cities. The growth in new disbursements through 2025 and the first quarter of 2026, though, has been disproportionately in tier-2 and tier-3 cities. That’s a real shift and it’s worth thinking about why.

A few drivers. Real estate construction in tier-2 cities has been more active relative to demand than in the metros, where new launches have slowed in the higher segments. Affordable and mid-segment housing in cities like Coimbatore, Lucknow, Indore, Bhubaneswar, Visakhapatnam has continued to find buyers. The remote and hybrid work patterns that emerged post-pandemic have created sustained demand for housing outside the major metros, particularly for buyers in their 30s and early 40s.

The risk profile of this growth deserves attention. Tier-2 and tier-3 housing markets have less developed price discovery, more volatility, and historically higher default rates per income bracket than metro markets. Banks and HFCs that are growing in these segments need to be doing it with eyes open.

Some institutions are. A few have built dedicated origination teams and risk models calibrated to the specific dynamics of the smaller cities they’re targeting. Others appear to be treating geographic expansion as a top-line growth lever without the matching credit infrastructure. The latter is what worries me when I look at the segment.

CRISIL and CARE Ratings have both published recent commentary on tier-2 mortgage growth that’s worth reading.

The self-employed underwriting question

The structural challenge in Indian housing finance has always been underwriting the self-employed. Roughly half the working population is informally employed or self-employed in a way that traditional documentation doesn’t capture well. Banks have historically managed this by either avoiding the segment or by charging spread premiums that limited their competitiveness against HFCs that built the capability to underwrite it.

Three things have shifted in 2025-26.

Account Aggregator-mediated bank statement analysis at scale. Several large lenders are now pulling 12-24 months of bank statement data, running it through cashflow analysis models, and using the output as a meaningful input into income verification for self-employed applicants. This is a real capability change.

GST data integration for businesses with a GSTIN. Where applicable, this provides a third-party verifiable signal that’s better than the older approach of relying on declared income and tax returns alone.

Property-level data. Several lenders have invested in property data products that triangulate the underwriting on the asset side, partly to compensate for income side uncertainty. Valuation discipline, title search quality, and occupancy verification have all improved, particularly at the bigger HFCs.

The combination has expanded the credit envelope for self-employed borrowers without obvious deterioration in credit quality so far. The cycle isn’t proven, though, and a year or two of meaningful economic stress would tell us whether the new underwriting is genuinely better or merely benefiting from a benign environment.

What the next 12 months might look like

Three things to watch.

Real estate market dynamics. Property price growth in metros has decelerated meaningfully. New launch supply in some segments has stalled. If buyer sentiment shifts further, mortgage volumes will move with it regardless of what banks and HFCs do on the supply side of lending. The relationship between property price expectations and mortgage demand is direct and not symmetric — buyers move slowly into a rising market but quickly out of an uncertain one.

Interest rate trajectory. The RBI’s policy stance through the rest of 2026 will shape funding costs for the lenders and effective EMIs for borrowers. Mortgage demand is rate-sensitive at the margin, particularly in the segments where the borrower is stretched on affordability.

Asset quality on the 2022-24 vintages. A meaningful portion of the housing book at major lenders is from the 2022-24 vintage when underwriting standards in some segments had loosened. Those vintages are now in the part of their life where late-stage defaults emerge if they’re going to. Watch the gross NPA numbers in the housing segment over the next two quarters; they’ll tell us whether the underwriting through that period held up.

Mint’s housing finance coverage has been consistently good on these themes and worth following.

The summary

Indian housing finance in mid-2026 is healthier than the deceleration in headline growth suggests, with real composition shifts underneath that the topline misses. The interesting work is happening in segment-specific underwriting, tier-2 city expansion done with discipline, and the gradual normalisation of self-employed lending through better data. The risks are real — vintage quality, market dynamics, the perennial regulatory uncertainty — but the segment isn’t broken. It’s transitioning, and the institutions investing in the underlying capability rather than relying on aggregate growth will come out of this transition in better shape than those that aren’t.