Eight Finance Companies,
One Balance-Sheet Cycle

Every one of these eight lenders grew its loan book between 39% and 91% last year. None of them funded that growth primarily with deposits — and the difference between them is now written in their capital ratios, not their profits.

Unlock the full research   Read the free briefing

Four things the filings show

1. The funding gap is universal

Across the eight, loans grew about Rs 595bn while deposits grew about Rs 201bn. Borrowings rose roughly Rs 423bn to bridge it. The regulator-cited sector figures say the same thing: non-bank lending grew 56% in 2025 against deposit growth of 19.8%. The proximate cause is not mysterious — the five-year vehicle-import ban was lifted on 1 February 2025.

2. The dispersion is in capital, not profit

Every company reported higher profit and six of eight reported ROE above 17%. But Tier 1 headroom over each company's own disclosed minimum now ranges from 13.42 percentage points at LOLC Finance down to 0.55 points at Vallibel Finance. Vallibel Finance and Singer Finance have already gone to shareholders for equity; HNB Finance's board resolved in March 2026 to raise Rs 3.5bn.

3. Improving NPL ratios are partly an illusion of arithmetic

An NPL ratio is bad loans divided by total loans. If the denominator grows 70% while the numerator holds steady, the ratio falls about 40% with no change in underwriting at all. Newly written loans also cannot default for several months. Sector-wide, net Stage 3 fell from 6.5% to 3.2% — some of that is real recovery and some is the denominator. The premium report separates the two company by company.

4. Two companies publish no asset-quality ratios at all

HNB Finance and Mercantile Investments disclose neither an NPL ratio nor a coverage ratio in the filings reviewed. Under this report's rules an axis missing more than 30% of its inputs scores N/A rather than a default value — so neither carries an asset-quality score, and both carry the lowest data-confidence grade. That is a finding about the companies, not a gap in the analysis.

The sector map

One chart does most of the work. Horizontal axis: the sustainable return on equity this report uses for valuation, after normalising for one-off items. Vertical axis: Tier 1 headroom over each company's own disclosed minimum. Bubble size is total assets; colour is credit cost.

The useful reading is not who is highest, but who is isolated. Two companies sit alone in the upper right with both the return and the capital to keep compounding. Three sit low and left, where the return does not obviously justify the capital consumed to earn it.

Top three on operating quality

The Operating Quality Score weights asset quality (20%), profitability (15%), earnings quality (15%), capital strength (15%), funding and liquidity (15%), efficiency (10%), franchise (5%) and governance and disclosure (5%). Every input is a peer percentile rank on a disclosed metric — no score was adjusted to fit a conclusion.

The full eight-company ranking, both score systems, all sixteen radar charts and the complete submetric breakdown are in the premium edition.

Why the LOLC Finance repurchase deserves scrutiny

On 22 July 2026 LOLC Finance's board resolved on a second share repurchase, announced the following day: up to 2,299,122,556 shares at Rs 7.00, a maximum of Rs 16.09bn, on an entitlement of seven shares for every ninety held. The stated purpose is to optimise shareholder value, and the company states that central-bank approval is not required. It comes twelve months after a Rs 21.11bn repurchase at Rs 6.00.

Four facts frame it. The offer price is a 42.9% premium to the Rs 4.90 undisturbed market price and 1.33× book value, while the shares trade at 0.93× book. The controlling chain holds 96.29%, so on pro-rata participation roughly Rs 15.50bn of the Rs 16.09bn returns to one shareholder. The transaction is accretive to earnings per share and dilutive to net asset value per share — that is simply what buying back stock above book does. And LOLC Finance's liquid-asset coverage stood at 101.6% of the regulatory floor at 31 March 2026, down from 140.2%.

Sector risks

1. Seasoning. Books that grew 39–91% in twelve months have not been tested. The 2025–26 vintages report through FY27–FY28.
2. Capital. Four companies have headroom measured in fractions of a point, not points.
3. The Rs 100bn threshold. HNB Finance's audited annual report discloses that crossing a Rs 100bn asset base triggers classification as a Domestic Systemically Important Licensed Finance Company — and it raised capital in anticipation. Several peers are at or near that line. The governing direction was not obtained for this report, so the consequences are treated as a disclosed forward risk, not a confirmed shortfall.
4. Funding-cost repricing. Fixed-deposit-heavy funding against multi-year fixed-rate lease assets, with rates already turning up late in FY26.
5. Vehicle loan-to-value caps. From November 2025, 70% for commercial and 50% for private unregistered vehicles. No company in the set has quantified the FY27 effect.
6. Gold. Gold loan books grew fastest of any product at several companies, against a 70% maximum loan-to-value.

Method, in brief

All eight share a 31 March year-end, which makes a clean common-period comparison possible. Figures are company (standalone) basis unless labelled. Audited figures take priority where an audited annual report was read. Return on equity was recomputed independently from profit and average equity for all eight and ties to the disclosed figure at every company. Credit cost is recomputed rather than taken from any company's own presentation. Derived values are marked and their formulas shown. Facts, calculations, interpretations and open questions are labelled distinctly throughout.

Valuation uses residual income and justified price-to-book, not enterprise-value multiples or a standard free-cash-flow model — deposits are a lender's raw material, not corporate debt, and subtracting them from an enterprise value produces nonsense.

Limitations you should know before reading further

• Three of the eight are analysed on condensed interim figures because their audited annual reports were not reconciled in this build: CDB, Vallibel Finance and Mercantile Investments.
• The governing central-bank capital direction was not obtained. No capital shortfall is presented as fact anywhere in this research.
• Two companies publish no asset-quality ratios, so their asset-quality scores are N/A.
• Valuation outputs are illustrative ranges from stated assumptions. They are not price targets.