Beinleumi 2025: Growth Is Strong, but the Spread Is Cooling and Capital Is No Longer Endless

Getting To Know The Company

Beinleumi in 2025 is not just another bank that benefited from high rates and printed a good annual report. It is one of Israel's five largest banking groups, with three clear economic legs: commercial and retail lending, niche retail brands such as Otsar HaHayal, PAGI and UBank, and a very strong capital-markets and client-assets franchise. What is working now is working well: net loans to the public rose to NIS 146.4 billion, public deposits grew to NIS 238.5 billion, fees climbed to NIS 1.777 billion, and client assets jumped to NIS 1.161 trillion. Credit quality also remained unusually clean, with NPL at 0.46% and the total allowance ratio at 1.25%.

But anyone looking only at NIS 2.26 billion of net profit and 16.2% return on equity is missing the core point. Profitability is already normalizing. This is not happening through a collapse. It is happening through a gradual cooling in the main engine: recurring financing profit fell 1.3% to NIS 4.868 billion, net interest income to average assets fell from 1.30% to 1.17%, and fourth-quarter net profit dropped to NIS 512 million from NIS 573 million a year earlier. The normalization is already here, just partly masked by continued growth in volumes and capital-markets activity.

There is another point the market can easily miss on first read: growth is not coming from the classic retail places. The loan book grew mainly in large corporates, middle-market clients and institutions. On the liability side, money is coming more from institutions and large businesses, and less from households and small businesses. That adds scale, but it also changes the economics of the bank: more wholesale volume, more sensitivity to deposit spreads, more liquidity volatility, and less natural support from sticky retail deposits.

The third issue is capital. This is now the first filter. Common equity tier 1 fell to 11.10%, and the surplus above the regulatory minimum stood at 1.87 percentage points at year-end, down from 2.08 percentage points a year earlier. That is still comfortable, but it is no longer a cushion that can be treated as endless, especially when the same board meeting approved a NIS 522 million dividend for March 2026, a framework for additional distributions, and room for buybacks. That is why 2026 looks less like another peak year and more like a bridge year in which the bank has to prove it can keep growing, preserve asset quality, and still avoid stretching capital too far.

The short economic map looks like this:

Events And Triggers

Capital returns and buybacks

The first trigger: on March 9, 2026, the board approved a NIS 256 million dividend for fourth-quarter earnings, and on the same day also approved an additional NIS 266 million cash dividend out of retained earnings. In practical terms that means NIS 522 million to be paid in March 2026. This is not just another routine dividend. It is a capital statement. The bank is telling the market that the surplus still looks sufficient to return capital even as profitability moves off peak levels.

The second trigger: the dividend is only part of the message. The bank is also considering a framework for three additional distributions, one every six months, up to a cumulative NIS 1 billion. In parallel, it is examining buybacks at roughly 25% of quarterly net income over the next two years. It has already approved a small buyback plan of up to 32,815 shares, with an estimated cost of about NIS 10 million, mainly meant to offset dilution from executive options. This matters for two reasons. On one side, the bank is signaling confidence. On the other, it is moving capital from background data into the center of the thesis.

CAL: real optionality, but not cash in hand yet

The third trigger: CAL became a source of friction in 2025, not just a store of value. The bank's share of CAL profit before tax fell to NIS 33.4 million from NIS 87.9 million, a 62% decline. That was not driven by a collapse in core activity. It was mainly the result of two one-off items at CAL level: NIS 137 million net after tax related to the VAT judgment, and another NIS 75 million net after tax from the revaluation of the El Al phantom option.

The fourth trigger: at the same time, the same holding can again become accessible value, but only if the transaction actually closes. The bank exercised its tag-along right in November 2025 in the sale of Discount's CAL stake, and if the transaction is completed it expects to record up to about NIS 132 million of net gain. The closing deadline, however, was extended by 30 days to April 19, 2026, and under certain conditions can be pushed out to November 1, 2026. So CAL is now both potential monetization, a source of accounting noise, and a timing event the market has to follow closely.

FIBI and the control structure

The fifth trigger: on February 3, 2026, the bank received from its controlling shareholder FIBI an invitation to negotiate a tax-free merger through a share exchange, under which FIBI would cease to exist and all its assets and liabilities would move into the bank. The board appointed an independent committee to examine the proposal. This should not be overstated, because there is no certainty around terms, regulatory approvals or completion. But it should not be ignored either. If the process advances, it can change how the market thinks about the control layer and about accessible value.

The outside signal

The sixth trigger: in early February 2026, Moody's raised the bank's outlook from negative to stable and affirmed the long-term rating at Baa1. That is a supportive external signal, but not one that cancels the capital question. If anything, it sharpens the gap between a bank that looks stable from the outside and a bank that now has to be judged on the tension between capital returns, funding mix and spread erosion.

Efficiency, Profitability And Competition

The central insight is that the spread is already cooling, and fees are masking part of that story. Anyone reading only the profit line could come away thinking Beinleumi simply moved from one strong year to another. That is not what is happening.

The spread engine is no longer providing the same push

Net interest income rose 1.7% to NIS 4.822 billion. On the surface that looks stable. But the better quality engine weakened: recurring financing profit fell 1.3% to NIS 4.868 billion from NIS 4.932 billion in 2024. The decline came from lower CPI contribution and narrower spreads in lending and deposit-taking, only partly offset by larger business volumes.

The fourth-quarter numbers are sharper still. Reported financing profit fell to NIS 1.125 billion from NIS 1.271 billion, and recurring financing profit fell to NIS 1.085 billion from NIS 1.167 billion. This is no longer theoretical normalization. It is already reaching the income statement.

The problem is not that loans are failing to grow. They are growing well. The problem is who is funding that growth and at what price. The shift in funding mix across activity segments says exactly that: in 2025, public deposits from institutions rose to NIS 93.4 billion from NIS 73.0 billion, and deposits from large businesses rose to NIS 32.9 billion from NIS 24.0 billion. At the same time, household deposits fell to NIS 68.3 billion from NIS 72.2 billion, and small-business deposits fell to NIS 25.2 billion from NIS 26.6 billion. The bank explicitly states that this shift reduced the financial spread from deposit-taking. That is the heart of the story. Growth is still there, but it is no longer coming with the same economic quality of cheap retail funding.

Loan growth is real, but increasingly wholesale in character

On the asset side, the picture looks similar. Gross credit to the public rose 12.9% to NIS 148.0 billion. But the internal breakdown is more revealing: credit to large businesses rose 22.8% to NIS 54.9 billion, middle-market lending rose 25.9% to NIS 9.2 billion, and institutional credit rose 89.8% to NIS 2.2 billion. By contrast, households excluding mortgages rose only 2.8% to NIS 23.8 billion, and mortgages rose 7.0% to NIS 39.0 billion.

That is not automatically negative. A bank that can grow in quality corporate lending and institutions can build a very good business on top of that. But it does change the risk shape and the spread structure. Lending like this usually comes with tighter pricing and more dependence on the market side of funding. That is why double-digit balance-sheet growth on its own is no longer enough to explain why 2026 should be a good year. The question is whether this growth is generating cleaner economics or simply preserving size.

Fees and capital markets are doing the heavy lifting

If the spread is cooling, what is holding up the result? Fees and capital-markets activity. Non-interest income rose to NIS 2.100 billion from NIS 2.006 billion, and fees themselves jumped 14.4% to NIS 1.777 billion. Client assets rose 38.4% to NIS 1.161 trillion, and market share in the system reached 16.1% versus 14.3% a year earlier. That is not cosmetic growth. It explains why the bank can still produce strong profitability even as the spread cools.

At segment level, the institutions segment delivered 23.7% revenue growth to NIS 543 million and 12.5% profit growth to NIS 162 million. Financial management revenue rose to NIS 217 million from NIS 130 million, and segment profit increased to NIS 113 million. In other words, the bank remains very strong exactly where it has brand, distribution and capital-markets capability. That is a real advantage. But it also means a bigger part of the thesis now depends on capital markets and client assets, not only on classic banking spreads.

Efficiency is solid, but not going where the headline suggests

Headcount continues to fall. Average positions dropped to 3,515 from 3,555, and the group employed 3,455 people at year-end versus 3,519 at the end of 2024. The strategic plan also talks about digital, operating centers, Gen AI, and continued branch and HQ efficiency moves. But as of year-end 2025, the efficiency ratio rose to 46.1% from 44.1%, so there was no additional step down here.

This is not because the bank lost control of cost. Operating and other expenses rose 7.2% to NIS 3.190 billion, while salary and related expense rose only 1.7% to NIS 1.769 billion. Some of the pressure is technological: IT expenses rose to NIS 599 million from NIS 550 million. In other words, the bank is more efficient than it was years ago, but in 2025 it chose to keep investing in digital and operating capability rather than harvesting the full efficiency benefit immediately. Revenue of roughly NIS 1.97 million per average position shows scale is working. The efficiency ratio shows growth investment is still taking part of the benefit.

Cash Flow, Debt And Capital Structure

In a bank, the right cash picture is not a classic industrial-style free cash flow bridge. The relevant bridge is the capital and liquidity bridge. The right question is how much flexibility remains after RWA growth, funding-mix change, capital distributions, and rating and regulatory sensitivity.

The capital test: still strong, but already less generous

CET1 fell to 11.10% from 11.31%, and total capital fell to 13.19% from 13.55%. At the same time, risk-weighted assets rose 10.6% to NIS 133.6 billion, mainly because of the increase in loans to the public. Total capital itself still rose to NIS 17.628 billion, but the rate of capital growth no longer comfortably outruns the growth in risk.

The bank also provides the most important sensitivity estimate here: a two-notch downgrade of Israel by S&P would reduce tier 1 capital ratio by 0.47 percentage points and total capital ratio by 0.56 percentage points. When the CET1 surplus above the minimum is only 1.87 percentage points, this is no longer a theoretical footnote. It is a reminder that the capital buffer is real, but not endless.

There is an offsetting force. The expected adoption of Basel IV operating-risk rules should, by the bank's initial estimate, increase capital adequacy by about 0.14 percentage points. But against the scale of the planned distributions, that is a modest relief. The true message of 2025 is that Beinleumi is moving from a world in which capital is a static fact to one in which capital becomes an active part of the thesis debate.

The liquidity test: the issue is not scarcity, it is funding quality

This is not a weak-liquidity story. Liquidity coverage stood at 129%, net stable funding at 127%, leverage at 5.04%, and the bank held roughly NIS 79 billion in cash and deposits at the Bank of Israel. The public-deposit-to-net-loan ratio also remained very high at 162.9%. It is hard to argue that Beinleumi enters 2026 under classic funding pressure.

But here too, the headline does not tell the full story. LCR fell from 165% to 129%, and the bank explains that this was mainly driven by higher lending, more short-term deposits from financial clients, and lower deposits from retail and small-business customers. In addition, the share of deposits held by the three largest depositor groups rose to 9.7% of public deposits from 6.5% a year earlier. Even if the levels remain comfortable, the quality of the funding layer has clearly changed.

The bank is trying to balance that through the capital markets. In 2025 it issued around NIS 4.1 billion of debt with one- to three-year maturities, and bonds and subordinated notes rose to NIS 6.791 billion, up 51.6%. That makes strategic sense: more loans require longer and more stable funding. But the move cuts both ways. Longer debt adds stability, yet it is also more expensive than cheap retail funding. So what looks like structural strengthening is also a reminder that funding cost is not moving lower.

Credit quality remains unusually strong

The strongest side of the report remains credit quality. Total problematic credit risk fell 14.1% to NIS 1.656 billion, the problematic-credit-risk ratio fell to 0.8% from 1.1%, and the ratio of non-accrual loans or loans more than 90 days past due fell to 0.46% from 0.53%. Coverage of non-accrual credit rose to 251.5% from 244.6%.

There is also one important detail in the mortgage book, especially in the current Israeli market context: contractor-subsidized mortgages. Here the bank explicitly says its activity in that area is limited and that the share of such loans in the mortgage portfolio is negligible. In addition, 67% of the mortgage book was granted at LTV of up to 60%, and around 98% at up to 75%. In other words, this does not look like growth bought through weak underwriting. That is critical, because it is one of the reasons the bank can still return capital without looking reckless.

Forward View And Outlook

Before turning to interpretation, there are four findings the 2025 numbers force the reader to hold:

That also defines the year ahead. 2026 looks like a bridge year, not a breakout year. The bank did not publish numerical earnings guidance, but it did approve a new 2026-2028 strategic plan with emphasis on return on equity, efficiency, market share and revenue growth, digital, Gen AI, capital-markets leadership, and more activity in large, medium and small businesses. So the direction is clear. What is still missing is proof that the next earnings model can work when rates are lower, CPI is less helpful, and deposit mix is less favorable.

What has to happen for the read to improve? First, the spread needs to stabilize. That does not mean returning to 2024 levels, but it does mean preventing another meaningful erosion from an even faster move into institutional and business funding. Second, business and institutional lending growth has to continue without a renewed rise in problematic credit. Third, the non-core events need to clear. A successful CAL closing could return accessible value and remove noise. If it fails or slips again, the holding remains an overhang. Fourth, if the FIBI process advances, the market will want to see a credible path rather than a loose strategic idea.

What could break the thesis? A mix of further spread erosion, negative regulation, and overly aggressive capital returns. The regulatory backdrop matters here. On the table are a retail-fee reform, an advanced competition-authority concentration process for the major banks, and a January 2026 bill proposing a special bank-profit tax for 2026 through 2030. The bank itself says it cannot yet estimate whether the initiatives will materialize or what their effect would be. That is exactly the kind of backdrop that can cap valuation even when a bank still looks strong on current numbers.

The most important point is that the bank is not entering 2026 weak. It is entering 2026 less simple. That is a material difference. If 2024 was a clear peak-profit year, 2025 was already a year mixing structural strength with normalization in spread earnings. 2026 will test whether Beinleumi can replace the tailwind from rates with a tailwind from fees, client assets, strong niches and capital discipline.

Risks

The main risk is not a credit collapse. The main risk is that three forces start working together: spread erosion, heavier regulation, and capital being distributed too quickly.

Risk one: spread erosion through funding mix

The bank already shows that the move toward institutional and business funding is pressuring the deposit spread. If that trend continues, and rates do not sharply reverse higher, it will be hard to reproduce 2024 spread profitability. The decline in LCR and the lower share of households and small businesses in public deposits reinforce the point. This is not a liquidity-stress risk. It is a funding-quality risk.

Risk two: capital return versus capital flexibility

The bank wants to return capital, and the current numbers still allow it. But the level of comfort is not endless. A 1.87 percentage-point surplus above the CET1 minimum is still comfortable, yet not large enough to ignore the combined effect of RWA growth, delays around CAL, sovereign-rating sensitivity or adverse regulation. The 0.47 percentage-point sensitivity to a two-notch sovereign downgrade is not trivial in that context.

Risk three: regulation and taxation

A retail-fee reform, the concentration-group process, a special bank-profit tax, and regulatory tension around control of credit-card groups are all more than theoretical. These are open processes hanging over the sector. Even if none of them fully materializes, the uncertainty itself can weigh on multiples and on how much credit investors are willing to give the excess capital story.

Risk four: strategic events that do not close

CAL and FIBI could both prove to be opportunities, but as long as they do not close they are also friction. CAL can return cash and create a gain. It can also keep getting pushed out. The FIBI merger could simplify the structure. It could also remain a long negotiation without outcome. The market usually does not give full value to events like this before certainty exists.

Short Read

The short-interest layer is interesting here. Short float is still not extreme, but it is no longer negligible. By the end of March 2026 it stood at 1.36%, after rising from a low of just 0.35% at the end of January and peaking at 1.62% in mid March. SIR, or days to cover, stood at 3.37 at the end of the period, matching the sector high shown in the comparison data.

That does not mean the market is betting on disaster. Far from it. It does mean the story is no longer being read as a clean case of a strong bank with clean credit and surplus capital. Part of the market is now testing the counter-read: a cooling spread, a shrinking capital surplus, and several strategic and regulatory threads that are still unresolved. As long as short interest stays in the moderate range, it is not powerful enough to change the thesis on its own. But it is a useful sign that the market is now looking for cracks, not just strengths.

Conclusions

Beinleumi ends 2025 as a larger bank, a very clean-credit bank, and still a strong bank in capital markets and client assets. But the main bottleneck is no longer whether the bank knows how to grow. It is whether it can preserve spread, keep liquidity comfortable and leave enough capital surplus while also returning capital and dealing with CAL, FIBI and regulatory change. This is exactly the kind of report where a reader who does not dig deeper can come away too optimistic.

Current thesis: Beinleumi remains a high-quality bank with strong growth in loans, fees and client assets, but 2025 already showed that spread profitability is normalizing and capital return is turning excess capital into the main question for 2026.

What changed versus the earlier read: the story shifted from a bank riding an unusually favorable rate backdrop to a bank that now has to produce better-quality earnings through fees, client assets and capital discipline while spread support fades.

Counter thesis: the concern may be overstated because the bank still posts high ROE, one of the cleanest credit books in the system, a strong capital-markets franchise, high liquidity, and enough capital to support both growth and distributions.

What could change the market read in the short to medium term: spread stabilization in the first half of 2026, a successful CAL closing, or, on the negative side, evidence that capital returns are continuing too aggressively while surplus capital keeps shrinking.

Why this matters: in banking, earnings quality is defined not only by how much is earned, but by where it comes from. As the spread cools, the question becomes how quickly fees, client assets and capital discipline can replace the tailwind from rates.

The test for the next 2 to 4 quarters is fairly clear. The spread has to stabilize, funding mix needs to stop deteriorating, credit quality has to remain clean even as business lending keeps growing, and the bank needs to show that capital returns do not come at the expense of real flexibility. If that happens, the read on Beinleumi can improve meaningfully. If not, 2025 may be remembered as the year the bank still looked very strong on paper, but had already started to lose some of the ease the market was willing to grant it.