Export Investment 2025: Bank Jerusalem Improved, But Not All Of That Improvement Moves Upstream

Getting To Know The Company

At first glance, Export Investment looks like a simple way to own Bank Jerusalem. That is incomplete. Export is effectively a banking holding company with only 7 employees, and almost its entire reason to exist is its 86.32% stake in Bank Jerusalem. So 2025 has to be read in two separate layers: what actually improved at the bank itself, and how much of that improvement is really able to move up to the public holding-company layer.

What is working now is fairly clear. Bank Jerusalem ended 2025 with attributable net profit of NIS 194.6 million and a 12.6% return on equity. At the Export parent layer there is no sign of immediate funding stress: year-end cash, cash equivalents, and short-term investments were about NIS 117.8 million, gross financial debt was about NIS 207.6 million, and net financial debt was about NIS 89.8 million. Covenants are comfortably wide, and the new bond issued in July 2025 extended the liability profile while replacing an older financing layer.

The active bottleneck sits elsewhere. At Bank Jerusalem, profit rose, but net interest margin was not what did the heavy lifting. Net interest income was almost flat at NIS 703.9 million versus NIS 703.6 million, while credit margin fell to 2.5% from 2.9% and deposit-gathering margin fell to 0.9% from 1.0%. What carried the year was lower credit-loss expense, stronger fee income, and gains from credit portfolio sales. That makes the picture less clean than the headline suggests. And at the Export layer, even when the bank generates profit, that value still has to pass through dividends, parent-company debt, and the constraints of a banking holding company.

There is also a practical market constraint that should be flagged early. Based on a last share price of 8,122 agorot and 11.29 million shares outstanding, Export trades around a NIS 917 million market value. Against that, the latest daily turnover in the stock was only about NIS 13.4 thousand, and short interest is close to zero. So the issue here is not aggressive short positioning. It is near-zero liquidity. That is exactly the kind of setup where even a better filing can take a long time to translate into price discovery.

Four non-obvious points matter most at the start:

• Bank profit improved, but not because the banking core became materially cleaner. Spreads compressed, and profit growth came mainly from lower provisions, stronger fees, and credit-portfolio sales.
• What the bank built in 2025 is a real capital-release engine, not just a one-off gain. It sold NIS 1.07 billion of credit in 2025, and by February 2026 it had already signed another NIS 680 million transaction.
• At the Export layer there is no immediate debt drama. Net financial debt fell, Series B replaced older funding, and covenants remain wide.
• The holding-company discount will probably not disappear just because the bank is profitable. Value still has to move upstream through dividends, under control restrictions, parent-company debt, and weak trading liquidity.

This chart explains why the story attracts attention. Bank Jerusalem really does arrive in 2025 stronger. But anyone trying to understand Export has to ask not only whether profit rose, but how clean that profit is, how repeatable it is, and how much of it can truly move up to the shareholder layer.

Events And Triggers

The Bank Built A Real Capital-Release Engine In 2025

Trigger one: Bank Jerusalem is no longer talking about portfolio sales as a tactical tool only. Its strategy explicitly frames credit sales, securitization, risk-transfer tools, and even bringing partners into selected activities as part of improving capital efficiency. That matters because it means the bank is trying to grow not only through more capital and a larger balance sheet, but through a lighter model.

Trigger two: In 2025 that strategy became real rather than rhetorical. In February 2025 the bank sold 90% of a roughly NIS 350 million commercial-credit portfolio and recognized a pretax gain of about NIS 12.6 million. In June 2025 it sold 90% of a roughly NIS 470 million housing-loan portfolio to a special-purpose vehicle. In August 2025 it sold 90% of a roughly NIS 318 million real-estate-collateralized portfolio and recognized a pretax gain of about NIS 11.3 million. In total, credit sold in 2025 reached NIS 1.07 billion, with total gain of NIS 31.3 million, and by year-end the bank was still servicing sold portfolios totaling NIS 3.55 billion for the buyers.

Trigger three: After the balance-sheet date, the next step already arrived. On February 23, 2026 the bank sold 90% of a roughly NIS 680 million real-estate-collateralized loan portfolio, and it expects to recognize a pretax gain of about NIS 38 million in the first quarter of 2026. For the market, that is a critical number. It should make the next quarter look strong, but it will also force investors to separate sale-driven profit from underlying banking profit.

This chart sits at the heart of the 2026 thesis. If this model repeats without weakening credit quality, Bank Jerusalem gets a growth engine that frees capital and creates servicing and fee income. If it turns out to be a timing tool that temporarily flatters earnings without really improving underlying returns, the story looks very different.

At Export Itself, 2025 Was A Year Of Reorganization Rather Than Stress

Trigger four: Export received an ilA rating affirmation with stable outlook in May 2025. In July 2025 it issued Series B bonds with NIS 100 million par and a 5.03% annual coupon, and in October 2025 it fully redeemed Series A. That did not create new value, but it did improve the quality of the parent liability stack: less near-term pressure, longer duration, and less reason to treat the holding-company layer as an immediate weak point.

Trigger five: Parent covenants tell a much calmer story than the phrase "banking holding company with debt" might suggest. Export's equity-to-assets ratio stood at 85.9% against a 35% minimum. Liabilities-to-equity stood at 16.4%, far below the relevant ceilings. And the ratio between Export's share in Bank Jerusalem's equity and Export's financial debt stood at 14.96 against a 2.5 minimum. In other words, the debt has not disappeared, but it is not the core problem.

Dividends Are Moving Upward, But Not At The Same Pace The Consolidated P&L Suggests

Trigger six: In 2025 Bank Jerusalem paid NIS 20.43 million for second-half 2024 earnings and NIS 44.36 million for first-half 2025 earnings. At the Export layer, actual dividends received from the bank during 2025 were NIS 55.9 million, while Export itself paid only NIS 9 million to its shareholders through two NIS 4.5 million distributions. After the balance sheet, the bank approved another NIS 33.48 million dividend for second-half 2025, and Export approved a NIS 9 million distribution of its own.

The message is simple. Value is already moving upward, but not at the same speed at which the bank reports profit. That is exactly why an article about Export has to stay focused on value transmission between layers, not just on subsidiary earnings.

Efficiency, Profitability, And Competition

The central insight is that Bank Jerusalem ended 2025 stronger, but not in exactly the way a classic bank would want to frame it. A reader looking for clean improvement in core spreads did not really get it. A reader looking for a bank that is broadening its income model, reducing capital intensity, and earning more from the same asset base got a much more interesting story.

What Really Drove Profit

Net interest income was almost unchanged at NIS 703.9 million versus NIS 703.6 million. By contrast, credit-loss expense fell to NIS 52.4 million from NIS 79.7 million, and non-interest income rose to NIS 295.4 million from NIS 211.4 million. Within that, non-interest financing income rose to NIS 64.3 million, while fee income jumped to NIS 224.9 million from NIS 155.2 million. On the cost side, operating and other expenses rose by NIS 71.1 million to NIS 662.0 million.

That means 2025 was not a year in which Bank Jerusalem simply charged more on credit. It was a year in which the bank combined three sources of improvement: lower provisions, stronger fees, and more activity in portfolio sales, syndication, and treasury/financial management. That is a perfectly legitimate business model, but it requires a much more precise read of earnings quality.

This chart matters because it decomposes the year into its actual moving parts. What holds up 2025 is not one clean engine. It is a combination of engines, some cleaner and some more sensitive to timing, provisioning assumptions, and non-spread activity.

Spreads Compressed While The Bank Became More Fee Driven

Another easy miss: credit margin fell to 2.5% from 2.9%, and deposit-gathering margin fell to 0.9% from 1.0%. At the same time, the efficiency ratio worsened to 66.2% from 64.6%. So even after a good year, Bank Jerusalem still does not look like a particularly lean bank enjoying classic operating leverage. It looks more like a bank trying to widen its income base through adjacent activities, even at the cost of a higher expense base.

That fits the strategic picture the bank itself is laying out: expanding BaaS, growing prepaid-card activity, deepening partnerships with non-bank institutions, widening construction finance, and using credit sales and securitization to improve capital efficiency. This is not a story of the bank becoming a mini-version of the large banks. It is a story of a bank trying to define a complementary model, earning less purely through spread and more through fees, service income, and capital turnover.

What A Superficial Reader Could Miss

The fourth quarter already hints that the full year was stronger than the run-rate exiting it. In Q4, attributable net profit rose only to NIS 24.3 million from NIS 22.6 million, and return on equity fell to 6.1% from 6.4%. Credit-loss expense in the fourth quarter rose to NIS 19.1 million from NIS 16.9 million. So even if full-year 2025 looks good, the year-end pace was less impressive than the annual headline suggests.

The risk of misreading is straightforward. It would be easy to assume the bank has already reached a new, clean base-profit level. A better read is that 2025 was a year in which several things worked at once, but some of them still need to prove their repeatability.

Cash Flow, Debt, And Capital Structure

The framing has to be explicit here. At the bank layer, the key questions are capital generation, liquidity, and growth capacity. At the Export layer, the more relevant frame is all-in cash flexibility, meaning how much room remains after actual cash uses, parent-company debt, and distributions. You cannot mix those two layers as if bank profit were the same thing as free cash at the holding-company level.

The Bank Layer: High Stability, But Not Unlimited Free Capital

Bank Jerusalem ended 2025 with a NIS 23.1 billion balance sheet, NIS 16.17 billion of net loans, NIS 18.25 billion of public deposits, and NIS 1.65 billion of equity. CET1 stood at 10.8%, total capital ratio at 13.4%, leverage at 6.8%, LCR at 181%, and NSFR at 133%.

These are good numbers, but the softer signal matters too. The regulatory CET1 minimum at the reporting date was 9.4%, while in March 2026 the board approved raising the internal required capital ratio to 10.25% effective April 1, 2026. So even if there is excess capital, it is not unlimited. If the bank wants to keep growing, keep paying dividends, and keep selling portfolios without harming flexibility, the capital machine still has to keep working.

Liquidity is another supportive point. The bank itself notes that after neutralizing institutional deposits up to one month, average quarterly LCR in Q4 2025 would have been 277%. That is a useful signal because it suggests liquidity is less fragile than the headline 181% ratio alone might imply.

At Export, Flexibility Exists, But It Is Not Detached From The Bank

At Export itself, the picture looks fairly comfortable. The company had NIS 117.8 million of cash, cash equivalents, and short-term investments at end-2025, about NIS 7.4 million of unused credit lines, and NIS 207.6 million of gross financial debt. After deducting cash, deposits, and marketable securities, net financial debt stood at about NIS 89.8 million. That is not a perfect capital structure, but it is a long way from real pressure.

There are also genuine sensitivities. A 10% decline in the value of the company's securities and deposit holdings would reduce value by about NIS 11.5 million, and a 1% increase in prime would raise annual interest expense by NIS 276 thousand on the prime-linked loans. Those numbers matter, but they do not alter the main conclusion: at Export itself, 2025 is not a liquidity-crisis story.

This chart makes clear why the stock looks cheap on the surface. Taking the market value of Export's stake in Bank Jerusalem, about NIS 1.409 billion, and subtracting roughly NIS 89.8 million of net financial debt produces a rough asset value of about NIS 1.319 billion. Against a market value of about NIS 917 million, that looks like a discount of roughly 30%. But this is exactly where the easy reading has to stop.

Not All Value Is Accessible Value

Export is not a free-form holding company. It is required to hold control of the bank through the company, its stake in the bank cannot fall below 50.1%, its bank-control shares cannot be pledged, and it may not engage in deposit-taking, lending as a business, or any activity that competes with the bank. On top of that, any transfer of control itself requires regulatory approval. So the discount is not just a market mistake. Part of it is the price of regulated value, a partially leveraged parent layer, and a stock that barely trades.

Outlook

2026 currently looks like a proof year, not an easy harvest year. Before getting into the details, four points have to be held together:

1. The first quarter of 2026 is likely to look very strong, but part of that strength comes from the NIS 680 million sale rather than from the banking core alone.
2. If the portfolio-sale engine repeats in 2026, Bank Jerusalem can grow faster than its capital base. If it slows down, pressure goes straight back to capital and dividend capacity.
3. At Export, the thesis strengthens only if dividends keep moving up at a pace that justifies the holding-company discount, not merely if the bank prints another profitable quarter.
4. The rise in non-accrual and 90+ days past-due credit, together with a softer Q4 exit rate, means the market will care more about the quality of 2026 than about the headline of 2025.

What Has To Happen At The Bank

The bank enters 2026 with a fairly clear strategy: expand its customer base in mortgages and targeted segments, widen business credit and construction finance, improve consumer-credit profitability, deepen BaaS and prepaid-card activity, and improve capital efficiency through sales, securitization, and risk transfer. The challenge is that not all of those goals move together.

On the positive side, there is real evidence that the model works. The bank explicitly states that excluding the NIS 1.07 billion of sold credit portfolios and NIS 711 million of jointly originated housing loans, the credit book would have grown by about 16%. That is a strong number. But if that growth relies on transactions that generate sale gains and fee income, the market will still want proof that the bank has a core earnings base that can hold up without leaning too heavily on those mechanisms.

What Has To Happen At Export

At the Export parent layer, the story is more straightforward. The company has stated in its dividend policy that it intends to distribute 80% of dividends actually received from the bank, net of company expenses. That policy makes it clear why the real focus is not consolidated profit, but actual dividend cash flow. If dividends from the bank keep rising, Export debt remains controlled, and Export progressively lifts its own payout, the discount should look less justified. If not, value on paper can remain value on paper for a long time.

What The Market Will Measure In The Next Reports

The first test will be simple: whether the next bank quarter still looks strong after separating the roughly NIS 38 million sale gain from underlying banking performance. The second test is credit quality. Non-accrual and 90+ days past-due credit has already climbed to 1.96%, and Q4 credit-loss expense already moved higher. If those two indicators keep worsening, the market will reopen its read of 2025.

The third test is capital versus growth. The board has already raised the internal capital target, so continued dividends and continued book growth have to coexist without eating into the cushion. The fourth test is the parent layer itself: how much of the bank's improvement actually turns into dividends that reach Export, rather than just a cleaner-looking consolidated P&L.

Risks

2025 Profit May Look Cleaner Than It Really Is

The first risk is earnings quality. Credit-loss expense fell to 0.32%, but non-accrual and 90+ days past-due credit rose to 1.96%, and the fourth-quarter exit rate was already softer. If provisions move back up and the bank cannot maintain similar fee momentum, part of the 2025 improvement may prove less durable than it first appears.

This is one of the most important charts in the piece. On the surface, provisioning improved. In reality, problem-credit indicators also rose. The question for 2026 is not which one is "right." It is which one is the earlier signal of where the book is actually going.

The Shift In Loan Mix Also Changes The Risk Profile

Bank Jerusalem is no longer only a mortgage bank. In 2025 the housing-loan book fell to NIS 9.314 billion from NIS 9.693 billion, while construction credit rose to NIS 2.004 billion from NIS 1.515 billion, and other business credit rose to NIS 2.615 billion from NIS 1.644 billion. That is a material shift in mix.

The implication cuts both ways. On one side, this widens the profit engine and reduces absolute dependence on mortgages. On the other, it adds more weight to business credit, construction, and real estate, exactly when the bank itself is flagging some increase in risk profile in those areas. The top 10 borrowers now amount to about 77% of equity, still within the bank's risk appetite, but clearly a figure that deserves monitoring.

Export's Debt Is Not The Problem, But It Keeps The Discount Alive

Export's financial debt looks manageable, but its existence still means not every shekel of bank dividend received becomes an equal shekel of dividend paid to shareholders. The company explicitly says its cash flow depends mainly on dividends from Bank Jerusalem, financing expense and debt service, debt raising, and dividends it pays onward. In other words, the Export parent is not just a passive conduit. It is a real layer with expenses, debt, and constraints.

Bank Value Is Not Necessarily Liquid Value For Shareholders

Control of the bank is regulated, the control shares cannot be pledged, and the company itself is barred from competing in banking activity. So even if Bank Jerusalem keeps improving, the value of that stake will not automatically turn into accessible cash. In ordinary holding companies, sometimes one asset sale is enough to narrow a discount quickly. Here that path is much more constrained.

The Market Is Extremely Thin

Daily turnover of about NIS 13.4 thousand means that even if the market understands the thesis better, the translation into price may remain slow and uneven. On the other side, the near absence of short positioning, only 47 shares short and an SIR of 0.06 in the latest week, means there is no aggressive bearish setup building under the surface. This is a thin stock, not a crowded short battleground.

Conclusions

Export ends 2025 in better shape than the phrase "banking holding company" might imply. Bank Jerusalem is stronger, the parent liability layer is more orderly, and dividends are moving upward at a better pace. That is the supportive side of the thesis. The central blocker is that the improvement at the bank is still not entirely clean in quality, and Export still has to prove that those better results can become accessible, durable value for its own shareholders.

So the right thesis for 2025 is not simply "Bank Jerusalem is better, therefore Export is cheap." The better thesis is that the gap between asset value and holding-company value remains large, but only part of that gap is opportunity. The other part is the price of parent debt, regulation, only-partial dividend transmission, and almost nonexistent trading liquidity.

Current thesis: Export benefits today from a better bank and a calmer parent debt layer, but the story still depends on proving that improvement at the bank can keep moving up through dividends rather than getting stuck at the holding-company level.

What changed: In 2025 Bank Jerusalem took another step from being a traditional niche bank toward a model that frees capital through credit sales, syndication, and securitization. At the same time, Export reorganized its own debt layer, so the issue is now less about financing stress and more about structure.

Counter-thesis: It is possible that this caution is too severe. If credit sales become a repeatable capital tool, capital stays comfortable, and the bank continues distributing 40% of earnings, then a meaningful part of the holding-company discount may look too wide even without a dramatic improvement in core spreads.

What could change the market's reading over the short to medium term: first and foremost, how Q1 2026 looks after separating the roughly NIS 38 million sale gain from the bank's recurring performance. After that, the key tests will be credit quality, capital headroom, and the pace at which dividends actually move up to Export.

Why this matters: in a banking holding company, shareholder value is not measured only by bank profit. It is measured by how much of that profit survives capital requirements, regulation, parent-company debt, and control constraints, and whether it really reaches the top as accessible cash. In 2025 Export looks better, but not yet clean enough to erase the full gap between asset value and market value.

What has to happen over the next 2-4 quarters: Bank Jerusalem has to show that portfolio sales remain a capital-efficiency tool rather than a substitute for core earnings, that problem-credit indicators stabilize even if provisioning stays low, and that capital ratios remain comfortable alongside growth and dividends. At Export itself, investors need to see continuing bank dividends and a gradual lift in distributions to shareholders. What would weaken the thesis is any combination of rising provisions, further pressure on core spreads, or a situation in which the bank keeps reporting higher profit while value still fails to move upstream.