Bank of Jerusalem: Is The Retail Core Really Profitable, Or Is Earnings Power Sitting In Financial Management?

Where earnings actually sit

The main article argued that Bank of Jerusalem's 2025 was driven less by classic spread expansion and more by capital release, risk transfer, and the fees that come with it. This follow-up isolates the next layer of that argument: where profit is actually recognized, and where the cost base actually sits.

That matters because the segment table creates a very sharp split. In 2025, the household segment moved to a NIS 5.9 million loss after a NIS 6.9 million profit in 2024. In the same year, financial management jumped to NIS 134.5 million of profit from NIS 78.4 million. This is not random dispersion across segments. It is a split between the place where the bank originates credit and the place where it recognizes profit from selling, syndicating, and managing that credit.

A superficial read can lead to two opposite but equally incomplete conclusions. One is that the retail core is simply loss-making. The other is that financial management has become a clean, effortless profit engine. Both are partial readings. What the bank is really showing is a two-step model: the friction and much of the cost base stay in the core segments, while a meaningful share of the monetization is booked later in the chain.

Why households could not carry the year

To see whether this is only accounting or also economics, you need to go one layer deeper. Within households, mortgages remained profitable at NIS 29.0 million in 2025 versus NIS 31.3 million in 2024. The problem sits in the rest of households, where the loss deepened to NIS 34.9 million from NIS 24.4 million. Mortgages are still holding up. The rest of the retail core is not.

That point matters because households did not swing to a loss on a spike in credit costs. Quite the opposite. Credit loss expense in the segment fell to NIS 23.3 million from NIS 62.5 million. Non-interest income still rose to NIS 100.8 million from NIS 79.9 million. And yet the segment moved into loss. That tells you the real pressure came from two other places: net interest income fell to NIS 408.2 million from NIS 451.1 million, and operating and other expenses rose to NIS 481.5 million from NIS 449.6 million.

That is already an operating signal, not just a classification issue. Even after a sharp improvement in credit costs, and even after better fee income, households still failed to produce profit. So the idea that the entire gap is just an allocation artifact misses part of the picture. There is also real pressure on the retail spread and on the core cost base.

Not all fee growth carries the same quality

One of the better-looking lines in 2025 was the NIS 69.7 million increase in fee income. But once you break it down, the picture is less uniform: NIS 28.6 million came from syndication fees tied to joint housing-loan origination, NIS 24.1 million came from prepaid-card activity, and NIS 9.5 million came from securities activity.

In other words, a meaningful part of the growth did not come from a quiet improvement in plain-vanilla retail spread economics. It came from activities tied more directly to syndication, prepaid cards, and securities. That is not necessarily a negative. It does mean the top line in fees does not, by itself, tell you how much clean earning power is left in the core franchise.

Even inside prepaid cards, which is a real growth engine, quality matters. Revenue from that activity rose to NIS 65.6 million from NIS 41.5 million, but the related expense base rose in parallel to NIS 45.7 million from NIS 33.3 million. Activity volume jumped to NIS 7.808 billion from NIS 4.591 billion, but this is not free fee growth. It comes with an obvious operating cost.

Why financial management looks so strong

This is the core of the continuation thesis. Financial management did not improve in 2025 merely because market conditions happened to be favorable. Profit rose to NIS 134.5 million from NIS 78.4 million, with a NIS 20.1 million increase in net interest income and a NIS 45.7 million increase in non-interest income. The drivers are explicit: more asset-liability management gains, more trading-book and position-management income, and, above all, more income from syndication, securitization, and portfolio sales.

The key number here is NIS 70.9 million. That is the amount of income from portfolio sales, securitizations, and syndication in 2025, versus NIS 48.4 million in 2024. A large part of the segment's earnings jump sits there.

But the most important line is the note behind the segment table: income from portfolio sales, securitizations, and syndication is recorded in financial management, while production costs are recorded in the household and business segments. That is not a technical footnote. It is the economic map of Bank of Jerusalem in 2025.

The implication is that the segment table is not really describing two fully separate businesses. It is describing a split value chain. The retail and business franchises originate assets, carry staff, systems, operating work, and allocated expenses. Financial management is where the bank sells part of the risk, manages the balance sheet, and recognizes a larger share of the monetization.

That is why the simplistic reading of "households lost money while financial management made money" misses the point. The core segments are not disconnected from financial management's profit. They simply do not keep that profit within their own segment line.

The cost base tells the same story

The issue is not only where income is booked. It is also how the cost base moved. Total operating and other expenses for the bank rose by NIS 71.1 million to NIS 662.0 million. NIS 31.0 million of that came from salaries and related expenses, NIS 17.6 million from depreciation and maintenance, mainly around software costs, and another NIS 22.5 million from the other-expenses line.

Once you go into that line, the same pattern reappears. The engines that pushed fee income also pushed costs. Prepaid-card issuance expense rose by NIS 12.2 million to NIS 45.5 million, commissions expense rose by NIS 4.5 million to NIS 10.9 million, computer expenses rose by NIS 6.3 million to NIS 71.0 million, and training expenses rose by NIS 3.8 million to NIS 7.8 million.

That matters because it is easy to look at strong fee growth and assume it reflects a better product or a cleaner franchise. In practice, some of these lines are also asking for more systems, more people, and more operating work. So the right question is not whether fees are growing. The right question is where that growth is being booked, how much of it stays behind as profit, and which segment is carrying the cost.

So is the retail core really profitable

The short answer is: not enough to carry the story on its own. In 2025, the retail core still generated activity, profitable mortgages, growth in prepaid cards, and a broad deposit base. But in segment-profit terms, it was not where the bank recognized the bulk of its earnings. The household segment moved to a loss, and the rest of households became more deeply loss-making.

The more precise answer is that Bank of Jerusalem's economic model no longer rests on a clean retail core that earns on its own and adds a little syndication on the side. The model now works like this: the core originates credit and activity, and financial management turns part of that origination into profit through risk transfer, securitization, syndication, and balance-sheet management. In that sense, earnings power is sitting more in financial management than in households today.

That can still be a capital-efficient model. But it forces the reader to judge the bank differently. Instead of asking only whether credit grew and fees increased, the better question is whether the core can hold spread and costs at a reasonable level even before the next syndication transaction, and whether financial management can keep producing profit without leaving behind a heavier and less profitable core franchise.

Bottom line

The key point in this continuation is not that households are weak and financial management is strong. The real point is that Bank of Jerusalem has built a model where the two depend on each other, but do not benefit from the result in the same place. Cost and friction sit in the core, while profit from sales, syndication, and financial management is recognized further down the chain.

So when a reader looks at 2025 and sees a jump in consolidated profit, the first question should be simple: is this the profit of a retail platform that genuinely improved, or the profit of a bank that has learned to move assets off balance sheet and monetize that well. In 2025, the answer leans clearly toward the second interpretation.