Poonawalla's return-on-assets gain leans on falling credit costs its credit chief expects to level off

Return on assets rose, and credit cost explains most of it

Poonawalla Fincorp reported a return on assets of 2.18% for the quarter ended September 2026, up from 1.98% the quarter before. Profit after tax has risen in every quarterly step shown below.

The question is where the latest gain came from. Credit cost, the yearly cost of bad loans as a share of the loan book, fell to 2.19% from 2.40%. By our rough arithmetic, that fall is about three-quarters of the pre-tax gain that the changes in margin, credit cost and cost ratio add up to.

Credit cost fell as forecast, and may now level off

In July the chief executive, Arvind Kapil, forecast that credit cost would keep improving quarter on quarter for a couple of quarters. It fell again, by the largest quarterly step in the five quarters we checked.

Shriram Iyer, the chief credit and analytics officer, described what comes next as a levelling: "we expect that this will stabilize". Kapil, in his own answer, called the new level a first milestone that looks structural, and later said he saw scope to keep improving beyond that. That is a difference of emphasis, not a split. But it means the credit leg of the return may soon stop adding to it, on a book that is young: by our arithmetic over a third of today's book is net growth over the last twelve months, and an analyst on the call put loans under twelve months old at well over half of the book, so much of it has yet to be tested through a downturn.

Growth is outrunning a guide the company has not restated

In July 2025 the company guided loan-book growth for that financial year at around 35% to 40%. The book grew 69.4% that year. In January and May 2026 it repeated the range without naming a period, and on this call management did not restate it. By our arithmetic the book is up about 23% since March, over half of that range used in half a year.

Growth like this has to be funded; since March borrowings have grown more slowly than equity, after the April raise, by our arithmetic. The debt-to-equity ratio is now below March's level after the April raise but above the level the company reported for the quarter before.

The target now rests on margin and a lower cost ratio

The company's return target is for the June 2028 exit; the chief executive referred to that commitment again on this call. Separately, an analyst, Abhijit Tibrewal of Motilal Oswal, recalled management saying its new businesses are being built for the same return. By our arithmetic, reaching that range needs a gain of 0.12 to 0.19 points a quarter over the remaining quarters, against a pace over the last two quarters at the top of that range, one that leaned on credit cost.

Cost to assets rose, the first quarter-on-quarter rise in the five quarters we checked. In May the company said it aimed to end the year below the level of the previous March, with a structural step down every March, so the next March is the first test.

Rates are the other test. Asked how a 50 basis point rise in interest rates would affect its cost of borrowing, management gave no figure on the call. The chief executive appeared to say about a third of borrowings is fixed-rate (the transcript is garbled), and described a 25 basis point move, which he said had been passed through a week before the call. Another executive, Sanjay Miranka, said the rise in borrowing cost over the last three months of tightening was minimal. Interest expense rose slightly faster than the book, and the figures alone cannot separate a higher rate from a bigger book.

What we do not know yet, and what the record says

We do not know yet what credit cost does as the young book matures, or whether the growth range still applies. Two records are checkable: the loan-book guide for the last financial year was exceeded, and the July forecast that credit cost would keep improving has held so far. A third depends on the measure: on the call that gave that loan-book guide, the credit chief said the company maintained its guidance of 1.5% to 2%; the summary of that call names no credit cost measure for it. The overall credit cost we track ran above that range through the year, while the company put credit cost for its core products at 1.43% on that call and 1.51% on the next, inside or just below it.

Management has met its growth range by a wide margin; its credit-cost range reads as missed on the overall measure and held on the core one; a holder is left to judge which of its numbers are still live.