Bank of Jerusalem: How Much Capital Do Loan Sales, Syndication, and Securitization Really Release?

The main article argued that capital release has become one of Bank of Jerusalem's core earnings engines. This follow-up isolates the actual machine: how much balance-sheet exposure really leaves the bank, where the income is booked, and what still stays with the bank even after a 90% sale headline.

Four sharp points up front:

• The 70.9 million shekel headline in 2025 is not a vague one-line gain. It can be reconstructed through 31.3 million shekels of loan-sale gains plus 39.6 million shekels of syndication fees.
• The economics are broader than that headline, because there is also a separate 17.6 million shekel net servicing-income line. The filings do not isolate how much of that line comes only from sold or jointly originated portfolios.
• The mortgage book the bank manages is already materially larger than what the balance sheet alone shows: by the end of 2025 the managed mortgage book stood at 16.529 billion shekels, while gross on-balance-sheet mortgage credit stood at 12.116 billion. The gap, 4.413 billion shekels, is sold portfolios still serviced for others.
• Risk transfer is real, but not absolute. The bank stresses true sale on the portion that was sold, and by year-end 2025 it no longer had securitization exposures. Even so, deal structure and regulation still leave it with at least 10% of each loan and with an ongoing operating role.

What Actually Left the Balance Sheet

The right way to read Bank of Jerusalem is no longer through reported credit balances alone. The bank is increasingly acting as an originator, distributor, and servicer. The investor presentation shows this clearly: the managed mortgage book has doubled since 2018 and reached 16.529 billion shekels by the end of 2025.

The real tell is in the footnote to that slide: of the managed book, roughly 12.116 billion shekels are gross on-balance-sheet credit, while about 4.413 billion shekels are portfolios sold and still managed for others. In other words, roughly a quarter of the mortgage book the bank manages now sits outside its own balance sheet. This is no longer a side activity. It is an architecture.

The presentation's cumulative loan-sales and syndication slide makes the same point. Cumulative sold portfolios and syndications reached 6.952 billion shekels by the end of 2025, while cumulative income from those moves reached 405 million shekels, excluding operating fees.

Four major stops in that machine can be identified in 2025 and just after year-end:

Syndication has a slightly different economic logic. Here the bank does not even have to warehouse the full exposure and then sell it later. Under the agreement, the bank funds 10% of each mortgage and the institutional partner funds 90%. During 2025, mortgages originated under that framework totaled 790.6 million shekels, and the outstanding balance reached 956 million by year-end. That means part of the capital-release engine already works at origination, not only after origination.

Where the Money Is Actually Made

The easy read is to treat the 70.9 million shekels as proof that selling portfolios "does the job." The more useful read is internal. The filings allow a fairly precise bridge:

• Loan-sale gains in 2025: 31.3 million shekels.
• Syndication fees in 2025: 39.6 million shekels.

Those two numbers add up exactly to 70.9 million shekels. In other words, the headline around sales, securitization, and syndication is really made up of two very different earnings layers: disposal gains on the one hand, and syndication fees on the other.

And that still is not the whole picture. There is a separate servicing line: net income from loan-portfolio servicing reached 17.6 million shekels in 2025, up from 14.7 million in 2024. Of that, 16.1 million sat in the financial-management segment. That matters because it shows the economics do not stop on the sale date. The bank continues to earn from ongoing portfolio administration.

But this is also where caution is required. The bank explains that it provides servicing and operations for two kinds of loans: portfolios sold or jointly originated, and government-backed budget loans. So the full 17.6 million shekels cannot be attributed cleanly only to loan sales, syndication, and securitization. What can be said with confidence is that the model already generates a recurring servicing layer, and not only disposal gains.

The cash-flow statement adds a third angle. In 2025 the bank received 1.101 billion shekels of proceeds from selling loan portfolios, while the net change in credit to the public was a negative 1.637 billion shekels. This is not just an accounting P&L story. The sale channel funded a substantial share of the year's credit growth. So when talking about "capital release," the right frame is not only reported profit, but also cash proceeds and the ability to recycle that cash back into new growth.

How Clean Is the Risk Transfer, Really?

This is where the continuation becomes more nuanced. On one side, the bank states clearly that it has no credit risk on the sold portion of the loans, and describes that as true sale. In the June 2025 mortgage securitization it adds another important layer: by year-end 2025 it no longer had securitization exposures, and the Pillar 3 risk report says explicitly that the bank does not have exposure arising from that activity. That suggests the 2025 executed securitization was relatively clean in terms of the structured tail left behind.

On the other side, "clean" does not mean "fully gone." Bank of Israel rules prohibit cherry picking, require the loans to stay on the bank's books for at least 12 months before sale, and require the bank to keep at least 10% of each loan in a sale transaction. The post-balance-sheet February 2026 deal states the point without any gloss: 68 million shekels, 10% of the portfolio, stay with the bank, pari passu with the SPV's rights. At the same time, the bank remains the servicer and operator of the sold piece.

So the risk transfer is real at the 90% level, but it is not a full exit from the economics of the loans. The bank still holds a direct 10% tail, and it remains operationally and economically tied to the portfolios it sold.

That distinction matters for the question in the title, how much capital is really released. The filings do not provide a direct shekel-by-shekel CET1 release bridge for each deal. They provide a bridge of transferred assets, gains, fees, and the regulatory envelope within which the model operates. That is not the same thing. The practical read therefore has to rely on two proxies: how much credit is moved out, and how much room is still left under the regulatory limits.

By the end of 2025, sold mortgage portfolios were running at about 22% of the mortgage book, against a 25% limit. Including syndication, the ratio reached about 29% against a 35% limit. That means two things at once. Yes, there is still runway. But no, this is no longer an open-ended channel. The model works, and it will probably keep working, but it is now part of day-to-day capital allocation rather than a one-off tactical tool.

What This Means for 2026

Q1 2026 will open with a clear tailwind from the roughly 38 million shekels of pre-tax gain expected from the February private securitization. That number will attract immediate attention, and for good reason. But the right way to read the quarter is to separate three layers:

• one-off sale gains,
• syndication fees, which still depend on origination pace and deal structure,
• servicing and operating income, which is the closest thing here to recurring franchise income.

If those three layers continue to be presented as one headline, the market may overstate earnings quality. If readers focus only on sale gains, they may understate the fact that Bank of Jerusalem has already built a real distribution-and-servicing platform. That tension is exactly the point of this continuation.

Bottom line: Bank of Jerusalem no longer operates only as a hold-to-maturity lender. It increasingly acts as an originator, distributor, and servicer of credit. That does release capital, and it does create meaningful income and fees. But the filings also show that the release is not a full exit from risk, and that the precise amount of capital released per deal is still not disclosed directly. So in 2026 the key question is no longer whether the model exists. The question is how much of it is truly repeatable, and how much regulatory room is still left to keep turning the wheel.