We had given an ‘accumulate’ call on the stock of CSB Bank in February 2025 (after Q3 FY25 results), when the stock was trading at ₹306 or 1.2x its book value. The GNPA ratio then was 1.6 per cent. RoA (return on assets) for Q3 FY25 and 9M FY25 came in at 1.5 per cent and 1.4 per cent respectively.

Since our call, the stock made a 52-week high of ₹575 in early January 2026, reflecting strong growth in business— advances grew 30 per cent in FY25 and 29 per cent in H1 FY26. At this price, the P/B (price-to-book value) multiple works out to 2x. This, despite the bank’s GNPA ratio rising to 1.8 per cent. However, the stock started correcting soon, more steeply so towards the end of January, when gold price dropped sharply (gold loans account for half of the loan book). With Q3 earnings release coinciding, it only added to the selling pressure. The stock is now at a 30-per cent discount to its 52-week high.

The bank reported a 29 per cent and 21 per cent rise in advances and deposits in Q3 FY26. Net interest and fee income grew well, helping the bank post a 32 per cent growth in pre-provisioning operating profit (34 per cent in 9M FY26). However, the bad news came from asset quality. Provisionsfor NPAs rose to ₹77 crore from a mere ₹7 crore a year ago and ₹55 crore a quarter ago. Credit cost reached 0.7 per cent, which had been around 0.5 per cent in Q1 and Q2. GNPA ratio worsened to 2 per cent. Net profit growth was flat for the quarter and just about 7 per cent for 9M FY26. RoA for 9M FY26 slipped to 1.2 per cent. All these largely had to do with the bank’s SME portfolio, which accounts for about 12 per cent of gross advances. Yet, the bank is reasonably comfortable for the reasons included below.

Also, CSB is in the process of scaling up into a full-service bank by FY30 (discussed below) from being a gold-loan focused financier (current loan-book is just over ₹37,000 crore). Tech/ system upgrades to support this are largely done, and the bank would enter the ‘scale’ phase from FY27. That said, FY27 would still be the first year in this phase and there may be initial hiccups in execution. Benefits are likely to accrue from FY28. CSB does have a fairly large footprint with about 850 branches to support the scale-up, comparable with Karur Vysya Bank’s 900 branches (loan book of ₹97,000 crore).

The two-week ceasefire helps one hope the energy crisis may not prolong to cause large-scale disruption to businesses, especially for SMEs. This coupled with the facts on SME book stated below put the bank in a reasonably-comfortable spot. However, this may not hold true if the crisis prolongs. This along with any execution hiccups in FY27 may cause some corrections in the share price, opening up buying opportunities.

The stock currently trades at 1.5x its book value. The five-year average P/B ratio is 1.6x. With the worst of credit cost behind, the management expects RoA to return to 1.5 per cent by FY28. The transformed and scaled-up bank by FY30 should be more resilient with a larger but granular franchise, no longer heavily dependent on gold loans. Therefore, we suggest investors with a horizon of three years or more to accumulate the stock on dips of 10 per cent or more, as margin of safety is more comfortable at those levels or lower.

SME trouble

Gold loans, corporate loans, SME loans and retail loans share the ₹37,000-crore-odd loan-book in a 51:24:12:13 ratio. Gold and corporate portfolios are performing all right. But SME and retail aren’t and thus dominate the slippages. Some portions of the retail book such as two-wheeler, agri and microfinance loans experience stress and so the bank is consciously not growing them. Slippages from older accounts continue to show up though. Rest of the retail book appears to do well.

SME is the bigger problem. Slippages are primarily from those borrowers whose businesses were adversely affected by Trump’s tariffs. The management has cut back on disbursements—from an average growth of 47 per cent year-on-year in the quarters of FY25 to an average -7 per cent in 9M FY26. However, the management is optimistic about the book as they have seen some accounts being upgraded as ‘performing’ in Q4 itself and the expectation is that about half of the slipped accounts would have been upgraded by March-end. Further, about 80 per cent of the SME book is collateralised.

Otherwise too, the bank has buffers from accelerated provisions and contingency provisions. Capital adequacy ratio is healthy at 19.4 per cent.

Gold loan LTV

Since late January, gold has corrected about 10 per cent. This is negative for CSB as it would move the LTV (loan-to-value) ratio of gold loans higher. If the ratio breaches the regulatory threshold (75-85 per cent based on the loan amount), it will lead to margin calls. LTV ratio stands at 63 per cent as of Q3 FY26.

The management has stated that it does not take price risk with gold loans—meaning it does not take advantage of higher gold prices to lend more. Working with current gold price (13 per cent up from December 31, 2026) and accrued interest for Q4 FY26, the LTV works out to 57-59 per cent. Our estimates suggest that gold should fall at least 20 per cent from the current level for the LTV ratio to reach 75 per cent.

Vision 2030

A key reason why the bank has not been able to expand into a full-service bank is the absence of systems to support diverse products. For instance, a new product like LAP (loan against property) would need its own suite of systems including the underwriting engine. CSB has invested in the same and the upgrades are largely in place now. By FY30, it envisages a loan-book mix of gold loans 25-30 per cent, corporate 30 per cent+, SME 18-20 per cent and the rest from retail.

Once the products are in place, the bank can raise low-cost deposits unlike now, where 40 per cent of deposits are bulk deposits and low-cost CASA balances are just 21 per cent. In the above example, a LAP loan can typically spawn a current account. This could bring cost of funds down and aid higher RoA. However, with non-gold loans dominating the book, yield on advances would largely come to track benchmarks unlike now, where about 60 per cent of the book (mainly due to gold) is based on fixed interest rates. Cost-to-income ratio falling to 50 per cent with scale (63 per cent for 9M FY26) can help offset this.

Published on April 11, 2026