Discount in the first quarter: profit recovered, but the margin and capital test remains open

Discount opened 2026 with NIS 930 million of net profit, up from the fourth quarter but down 10.2% from the comparable quarter. Excluding one-offs, mainly the special bank tax and settlement effects, profit reached NIS 1.04 billion and return on equity (ROE) rose to 12.2%, so the quarter is not as weak as the reported line suggests. Still, the more important number is that financing income from regular operations fell to NIS 2.365 billion and continued declining against every quarter in 2025. That happened while credit grew, so the issue is not volume but pricing: loan and deposit spreads are under pressure. At the same time, credit-loss expenses rose to NIS 182 million, almost entirely from collective provisions, while problematic debt declined. The CAL sale can add 0.41% to 0.50% to Common Equity Tier 1 (CET1), but until closing the bank is paying out 50% of profit and operating with a narrower regulatory buffer: 10.24% versus a 9.20% requirement. This also continues the checkpoint from the prior annual analysis, which framed 2026 as a year in which earnings quality matters more than reported profit. The next few quarters need to show stabilization in the spread, completion of the CAL sale, and use of the released capital without a prolonged hit to ROE.

Profit recovered, but the financing engine weakened

The bank entered the quarter with four engines that need to be read together: the Israeli banking business, Mercantile, IDBNY in the US, and CAL, now presented as a discontinued operation ahead of sale. Net profit alone blurs the story. Continuing operations generated NIS 863 million, CAL added NIS 75 million, and subsidiaries remained meaningful contributors. But the recurring profit line that matters most for a commercial bank is financing income, and the direction there is less comfortable.

Net interest income fell to NIS 2.309 billion, down 4.5% from the comparable quarter, even though net credit rose 2.2% from year-end 2025. The internal bridge explains the gap: price effect reduced income by NIS 240 million, while volume added NIS 132 million. In business terms, the bank is extending more credit, but the economics of its sources and uses no longer look like they did during 2025.

That gap matters precisely because there were also positive points in the quarter. Fees rose 13.7% to NIS 589 million, mainly from credit-handling fees and financing-business fees. Operating expenses fell 4.8% to NIS 1.509 billion, and salary expenses fell 7.8%. The efficiency ratio was 48.3%, much better than the fourth quarter and well below the 59.3% recorded then. The quarter therefore shows progress on expenses and fees, but it also shows that efficiency alone does not offset banking-spread erosion.

The provision rose without a credit-crisis signal

The line that may look too negative on a quick read is credit losses. Credit-loss expenses rose to NIS 182 million, from NIS 52 million in the comparable quarter and NIS 141 million in the fourth quarter. The cost of risk was 0.25%, not an extreme level for a bank, but much higher than in the comparable quarter.

The important detail is where the provision came from. The quarter included NIS 4 million of net income from individually assessed borrowers, while the expense came almost entirely from NIS 186 million of collective provisions. That means this is not currently a signal of a sharp wave of large borrower deterioration, but a mix of credit growth, accounting write-offs and weaker macro parameters. At the same time, problematic balance-sheet credit to the public declined to NIS 4.993 billion from NIS 5.406 billion at the end of 2025, and non-accrual debt declined slightly.

This is a delicate point for a bank that is still growing credit. Credit to mid-sized businesses rose 17.5% during the quarter, and credit to large businesses rose 4.3%, while household credit grew more moderately. Construction, real estate activity and mortgages remain a major part of the credit book. As long as problematic debt is declining, the provision looks more like collective conservatism than a credit event. But as the bank keeps growing credit while spreads are lower, the market is likely to treat provisions less as quarterly noise and more as a test of growth quality.

CAL should release capital, but the current buffer narrowed

The largest event around the bank remains the CAL sale. Immediate consideration is expected at about NIS 2.694 billion in cash, out of total consideration that can reach about NIS 2.873 billion including a contingent component. As of March 31, 2026, completion is expected to generate about NIS 307 million of after-tax gain, with a possible additional after-tax gain of up to about NIS 125 million if the contingent consideration conditions are met. More important than the accounting gain, CAL's risk-weighted assets are expected to exit the group, lowering RWA by about NIS 17.122 billion and lifting CET1 by 0.41% to 0.50%, depending on the dividend distributed from the sale gain.

That is a meaningful capital addition, but it is not in hand yet. The Competition Authority extended its review to May 28, 2026, and the contractual closing date has been extended to June 18, 2026, with possible further extensions. Until then, the current capital picture is less wide than the CAL headline suggests. CET1 fell to 10.24%, from 10.38% at year-end 2025 and 10.54% in the comparable quarter. The buffer above the regulatory requirement fell to 1.04 percentage points, after 1.18 at year-end and 1.35 a year earlier.

The bank is also continuing to return cash to shareholders. The board approved a dividend of about NIS 465 million, equal to 50% of first-quarter profit. That payout signals confidence, but it also sharpens the tension: before CAL, capital is growing more slowly than risk-weighted assets. After CAL, the bank will need to show that released capital does not remain a temporary surplus that depresses ROE. The bank itself notes that ROE may decline temporarily after closing until excess capital is either invested in the bank's business or distributed.

The operating moves across the group feed the same test. Management presents a target to reduce about 450 positions at the bank itself by the end of 2026, Mercantile already has more than 100 employees signed to an early-retirement plan, and IDBNY reduced about 70 positions over the last two quarters. That should help expenses, but the quarter already showed the limitation: expense savings can improve efficiency, but they do not replace a stable financing spread.

Conclusions: a proof year for the spread, not only for capital

The first quarter does not weaken the bank, but it does not close the debate either. Excluding tax and specific settlement effects, profitability looks better than the reported number. Credit quality is not showing acute pressure, and efficiency work is beginning to show in the income statement. On the other side, recurring financing income is still falling, provisions are moving back to the center of the story, and the capital buffer above the requirement narrowed before the CAL sale has closed.

From here, 2026 looks less like a breakout year and more like a proof year. A cleaner improvement requires stabilization in the spread, continued stability or decline in problematic debt, completion of the CAL sale, and a capital decision that shows what the bank does with the proceeds: credit growth, distribution, buyback, or a combination. The counter-thesis is strong enough: adjusted profit still implies double-digit ROE, expenses are falling, and CAL can release capital quickly. But until the spread stabilizes and CAL capital becomes active or distributed capital, the market is looking at a stable bank whose profitability test is still unfinished.