Discount Bank 2025: Credit and fees held up, but core profitability is already stepping down

Company Overview

Discount Bank is no longer mainly a story about baseline stability versus credit risk, and it is not just a “gap to the big two” story either. The cleaner way to read it now is as a broad commercial bank with three operating engines that need to work together: a domestic banking core that is still growing well, Mercantile Bank which is supposed to become leaner after a management change and a retirement plan, and IDBNY which is supposed to move from a size-first posture to a profitability-first posture. Sitting above all of that is CAL, which no longer reads as a growth asset so much as a transaction that should free capital and reshape the group’s profile.

What is working is fairly clear. Credit is still growing, fees are rising at a healthy pace, credit quality remains benign, and wholesale market access does not look like the binding constraint. What is less clean is the earnings engine. A superficial read of net interest income, up 0.7% to NIS 10.0 billion, can miss that financing income from current operations actually fell to NIS 9.812 billion from NIS 10.064 billion. That is already a different message. It says the core earnings base entered 2026 with less tailwind than the headline suggests.

That is the active bottleneck. Discount’s current constraint is not credit quality, and it is not funding access. It sits in margin, efficiency, and the question of how long it will take to convert the CAL transaction from a capital promise into a credible profitability story. Q4 already exposed it: net income fell to NIS 856 million, ROE dropped to 10.2%, and the efficiency ratio rose to 59.3%. Even after adjusting for Mercantile’s retirement-plan cost, the quarter was still softer than the market would want to see.

At roughly NIS 39.3 billion of market cap, with short interest at just 0.21% of float and an SIR of 0.65, the market is not pricing a credit event or a liquidity squeeze. It is pricing a different question: can Discount sustain double-digit returns once CPI becomes less helpful, deposit costs rise, and CAL is still not closed? That is now a question about earnings quality, not survival.

The Group’s Profit Map

This chart matters because it shows that 2025 was not a breakdown. It was a step down. Discount still generates more than NIS 4 billion of net income, but the distance to management’s own 2030 targets remains meaningful. Those targets are net income above NIS 5.2 billion, an efficiency ratio below 43%, and ROE of 13% to 14%, with payout of up to 50% of net profit. In plain terms, 2025 was not an arrival year. It was a year that made the gap between current capability and the end-state more visible.

Events And Triggers

The first trigger: CAL is no longer a distant strategic option. It is a near-term event that explains a large part of the 2026 read. Total consideration can reach about NIS 2.873 billion, of which about NIS 2.694 billion is immediate cash at closing, plus contingent consideration of up to NIS 180 million in 2027 and 2028. Based on end-2025 estimates, the closing should generate about NIS 370 million of after-tax gain, with room for another NIS 125 million if the contingent leg is paid. More important than the accounting gain is the capital effect: the bank estimated that CAL’s risk-weighted assets exiting the group amount to roughly NIS 17.4 billion net, implying a 44 to 55 basis-point uplift in CET1, depending on the dividend decision.

The other side of the same event matters just as much. The closing deadline has already been pushed out by 30 days to April 19, 2026, and if all available extensions are used it can move as far as November 1, 2026. The bank also states explicitly that ROE may fall temporarily after closing until the released capital is redeployed into loan growth or distribution. So even if the deal is clearly positive for capital, it does not automatically solve the earnings question.

The second trigger: Q4 was the quarter in which management lost the luxury of a clean narrative. The gap between Q3 and Q4 did not come from bad credit quality. It came from a mix of weaker CPI support, transition costs at Mercantile, and unusual costs at IDBNY. Mercantile’s retirement plan added NIS 159 million to other expenses at group level, and the annual presentation makes clear that the plan had a NIS 104 million net negative effect in Q4. The market will tolerate that kind of cost once, but only if the savings show up quickly on the other side.

The third trigger: Mercantile and IDBNY are no longer side businesses. Mercantile finished 2025 with NIS 776 million of profit, down from NIS 843 million, and the annual presentation shows its Q4 ROE falling to just 7.0%. At IDBNY, net income fell to $76 million from $89 million, and ROE fell to 5.5% from 7.0%, with Q4 at only 3.6%. If these two areas do not improve, the domestic core will have to carry too much of the group on its own.

The fourth trigger: On the funding side, the picture is much more supportive. In early January 2026, both Midroog and S&P Maalot kept their ratings at stable outlook, while the group’s issuance frameworks were expanded to as much as NIS 2.65 billion of bonds, NIS 620 million of CoCo capital, and NIS 1.35 billion of commercial paper. That is an important external signal. The wholesale market is not closing the door. If the thesis breaks, it is more likely to come from execution and pricing than from lost market access.

Efficiency, Profitability And Competition

The central insight is that credit is still working, but the engine that turns it into profitability is becoming less generous. Net loans to the public rose to NIS 288.8 billion from NIS 267.3 billion, and the investor presentation shows gross loans already reaching NIS 292.5 billion, up 7.9% year over year. Most of the expansion came from corporates, international activity, and mortgages. On the surface, that is exactly what a bank wants to show.

But growth quality matters more here than growth alone. Loan-loss expenses remained very low at NIS 450 million, or 0.16% of average loans. Yet the profitability reading tells a different story. This was not mainly a year of a stronger credit engine. It was a year of balance-sheet management in an inflation and rate environment that had already started to shift.

That chart tells a straightforward story. The bank kept growing while the problematic-debt ratio kept improving. So anyone looking for Discount’s main problem in credit quality is probably looking in the wrong place. The cleaner place to look is margin and efficiency.

Financing income makes that point clearly. Net interest income rose 0.7% to NIS 10.001 billion, but total net financing income fell to NIS 11.082 billion from NIS 11.292 billion. After stripping out non-core items, financing income from current operations fell to NIS 9.812 billion from NIS 10.064 billion. That matters because it shows the underlying earnings base had already weakened before the CAL sale closed.

Q4 made that visible. Net interest income fell by NIS 380 million versus Q3, down 14.3%, largely because of CPI differences between quarters. The investor presentation also shows NIM falling to 2.16%, from 2.56% in Q3 and 2.45% in Q4 2024. Discount remains very profitable, but it no longer enjoys the same macro tailwind that supported margins in earlier periods.

The other side of the picture is fees, and here the news is much better. Fee income rose 12.2% to NIS 2.156 billion. That is not cosmetic. For a bank trying to move toward a less CPI-sensitive and rate-sensitive earnings mix, fee growth is evidence that the franchise is working and that a more durable revenue layer exists. The fact that the fee-to-assets ratio stayed at 0.50% also suggests the bank has not lost its ability to monetize activity.

The problem is that fees alone do not offset everything being lost on the funding and cost side. Operating and other expenses rose 5.7% to NIS 6.556 billion. The main deterioration came from other expenses, up 19.8% to NIS 1.747 billion, driven primarily by Mercantile’s NIS 159 million retirement plan and NIS 40 million of higher computer costs. That is why the annual efficiency ratio rose to 49.2%, still far from the sub-43% target management set for 2030.

This chart explains the pressure better than a long paragraph can. Public deposits rose to NIS 359.5 billion, but the faster growth came from the more expensive floating-rate bucket. So the funding base grew, but it also became costlier. The bank itself writes that the spread from taking public deposits fell to 1.2% in Q4 from 1.5% a year earlier, and to about 0.8% after excluding non-interest-bearing current accounts. That is no longer a number that allows for too much complacency.

That is also where competition enters the story. Discount explicitly says the spread from public deposits is likely to shrink further if customers continue shifting current-account balances into interest-bearing deposits. That is an important admission. It means competition for funding, especially among households and small businesses, is no longer a theoretical backdrop. It is a real driver of reported earnings.

That chart sharpens one more point. Mercantile’s retirement-plan cost did weigh on Q4, but even after excluding it the quarter does not simply return to the profitability level seen in mid-2025. Anyone telling themselves that all of late-2025 weakness is purely one-off is simplifying the picture too much.

Capital, Funding And Balance-Sheet Structure

For a bank, the right way to read “cash flow” is through capital consumption, liquidity, funding mix, and the ability to keep growing without putting pressure on regulatory ratios. On that reading, Discount is not a stressed bank, but it is also not yet a bank with excess capital fully sitting on the shelf waiting for distribution.

Risk-weighted assets rose to NIS 332.6 billion from NIS 303.2 billion. At the same time, CET1 fell to 10.38% from 10.66%, the total capital ratio fell to 13.07% from 13.60%, and the leverage ratio fell to 6.5% from 6.7%. These are still comfortable levels above regulatory requirements, but before CAL closes the bank is only 118 basis points above its 9.20% CET1 minimum and just 57 basis points above its 12.5% total-capital minimum.

That has a double meaning. It does not signal distress. But it also means the “excess capital” story is not fully in hand yet. Loan growth consumes capital, and until CAL is closed, the room for maneuver is narrower than a simple capital-strength headline might imply.

The picture is reassuring, but not loose. LCR at 120.8% and NSFR at 117.2% point to solid liquidity, and the bank also benefits from a diversified domestic funding base, with 57% of local public deposits coming from retail sources. On the other hand, these same ratios show that some of the liquidity margin narrowed during 2025. There is no liquidity problem here, but there is more funding cost and more active balance-sheet management.

Capital allocation matters just as much. In 2025 the bank returned NIS 1.751 billion in cash dividends and another NIS 215 million in buybacks, for total shareholder payout of NIS 1.966 billion, or 47% of net profit. In addition, under the buyback program the bank repurchased 15.5 million shares, around 1.26% of capital, for about NIS 433.7 million. That is a positive signal from management that it trusts the bank’s capital-generation ability. But it also creates a higher bar: if the capital released by CAL is not turned into efficient growth or convincing distribution, those decisions will look less smart in hindsight.

As an outside signal, capital markets are still giving Discount credit. Both rating agencies kept stable outlooks in January 2026, and in the same move the group’s funding frameworks were expanded. That says the wholesale funding front is open. If the thesis gets challenged, it is more likely to come from core earnings, CAL timing, or a longer-than-expected cost transition, not from loss of access to markets.

Outlook

Finding one: 2026 looks like a transition year with a proof component, not a breakout year. The gap between 2025 actuals and management’s 2030 targets is still wide: NIS 4.14 billion of net income versus a target above NIS 5.2 billion, an efficiency ratio of 49.2% versus a target below 43%, and ROE of 12.6% versus a 13% to 14% target. The market will need to see a credible execution bridge, not just ambition.

Finding two: CAL will help capital faster than it helps earnings. The bank itself says ROE may fall temporarily after closing until the released capital is redeployed. So the test does not end when the deal closes. It simply moves to a different question: whether Discount knows how to turn that capital into growth, payout, or a sensible mix of both.

Finding three: Mercantile and IDBNY are now part of the 2026 core read. If Mercantile does not turn the retirement cost into a lower run-rate cost base, and if IDBNY does not move back toward acceptable profitability after the management reset and the end of the consent-order chapter, the group will remain too dependent on the domestic bank and on capital actions.

Finding four: Good credit quality buys time, but it does not replace margin. As long as the cost of credit stays around 0.16%, Discount can absorb some spread pressure without breaking. But if the environment shifts from supportive CPI and rates toward more competitive deposit pricing, profitability will be tested even without any credit event.

Management is building the forward plan against a backdrop of expected economic rebound, but that is only the backdrop. In the market, the three items that will decide how the next reports are read are different: CAL completion, the pace of spread compression, and proof of cost savings. Those are the variables that will decide whether 2026 is viewed as a build year toward 2030 or as a year in which the long-term targets start to look a bit too optimistic.

That chart highlights why 2026 is a proof year. The domestic bank still carries most of the story, but the surrounding pieces are no longer negligible. Mercantile and IDBNY do not just need to be “fine”. They need to become visibly cleaner contributors again, or any slippage in the domestic core will stand out much more.

What has to happen over the next 2 to 4 quarters for the thesis to improve? First, CAL needs to close without another meaningful delay. Second, margin pressure needs to stop broadening into a wider annual erosion story. Third, Mercantile needs to show that the retirement cost was the price of real savings. Fourth, IDBNY needs to show that the new organizational setup is translating into profitability rather than more transition expense.

What breaks the thesis? Another long CAL delay, more NIM pressure without compensation from fees, or a situation in which spending at IDBNY and Mercantile stays elevated even after management has already framed the restructuring moves as the way forward.

Risks

The first risk is timing and capital-use risk around CAL. The transaction should improve capital, but the closing date has already slipped and the final consideration remains subject to adjustments and contingent payments. Even after closing, there is a risk of “idle capital”, meaning a period in which capital is released but not yet turned into returns. That is a less dramatic risk than capital stress, but it matters a great deal for how the stock will be valued.

The second risk is margin erosion. The deposit spread already fell in Q4, and the bank itself assumes that continued migration from non-interest-bearing current accounts into interest-bearing deposits will put further pressure on it. This is exactly the type of risk that hides behind a healthy loan-growth number, which is why it deserves to be made explicit. If competition for funding intensifies, loan growth alone will not be enough.

The third risk is execution risk at the subsidiaries. Mercantile still has to prove that the new plan produces lasting net savings, and IDBNY still has to prove that it can move back toward better profitability after the management change. At IDBNY, 2025 also included a sharp rise in credit-loss expense, driven by a longer lookback period in the model and higher specific provisions. That is not a crisis, but it is not noise either.

The fourth risk is legal and regulatory. The external auditors specifically pointed to a lawsuit that cannot currently be estimated, involving the claim that mortgage borrowers were not informed that there is no legal obligation to insure the pledged property under a building-insurance policy. In February 2026, another class-action request was filed against Discount and Mercantile that combines several claims around fees, legal-treatment accounts, dormant accounts, and liens, with claimed damage above NIS 2.5 million. None of these cases changes the thesis on its own, but the fact that the auditors highlighted one of them means the legal layer cannot simply be dismissed as immaterial background noise.

The fifth risk is the operating environment itself. The risk report stresses that Discount still operates in a high-risk geopolitical setting, and that the full consequences of recent security developments for the economy cannot yet be assessed. In a bank where interest margins, business activity, and credit quality are all sensitive to the pace of the local economy, that is not a throwaway macro paragraph. It is part of the reason the market will still demand proof before granting full credit for the 2030 plan.

Conclusions

Discount exits 2025 as a stronger bank than a weak Q4 alone would imply, but also as a less clean story than the full-year profit figure alone suggests. Credit, fees, and liquidity still support the case, but margin is compressing, capital is still waiting for CAL, and Mercantile plus IDBNY have not yet turned back into full support acts. In the short to medium term, the market’s reading is likely to depend far more on NIM, CAL, and costs than on credit growth itself.

Current thesis in one line: Discount remains a solid, good-quality bank, but 2026 will be a test of core earnings power and capital deployment rather than a test of loan growth.

What has really changed relative to the simpler read? The focus has shifted from “is the bank stable enough?” to “is the earnings engine strong enough without major CPI tailwind and while CAL is still unresolved?” That is an important shift because it moves the debate away from capital and credit and toward margin, efficiency, and capital allocation.

Counter-thesis: It is possible the market is being too harsh. Discount still shows loan growth, strong fees, good credit quality, solid liquidity, and open capital-markets access. If CAL closes on a reasonable timetable and Mercantile’s savings show up quickly, even a roughly 12% to 13% return profile may still look good enough.

What could change the market’s reading in the short to medium term? Any regulatory progress on CAL, one quarter in which margin looks stable rather than weaker, and clear evidence that transition costs at Mercantile and IDBNY are starting to fade. On the other hand, another CAL delay or another weak margin quarter would likely weigh on the story faster than any single credit datapoint.

Why does this matter? Because banks do not only lose quality through credit losses. They can also lose quality through slow erosion in the earnings engine. That is exactly Discount’s test now.

Over the next 2 to 4 quarters, what strengthens the thesis is CAL closing, spread pressure stabilizing, and restructuring turning into real savings. What weakens it is more delay on the transaction, more deposit-side pressure, or continued underperformance at Mercantile and IDBNY.