IDBNY at Discount: Is the shift from size to profitability actually starting to work?

The main article already framed 2026 as a proof year for Discount's core profitability. This follow-up isolates IDBNY because one of the group's more interesting contradictions sits there: the U.S. balance sheet grew, net interest income rose, and the consent-order chapter with the FDIC and the NYDFS closed in October 2025 and January 2026. Even so, Bancorp finished 2025 with just 5.5% ROE, and the fourth quarter dropped to 3.6%.

That matters because IDBNY is no longer just a geographic diversification line. Bancorp's contribution to Discount's business results fell to NIS 176 million from NIS 274 million, and Discount's annual presentation makes clear that the combination of CPI, Mercantile, and IDBNY reduced group ROE by 4.8 percentage points in the fourth quarter. In other words, the U.S. unit is no longer just an option on future upside. It is part of the reason the group-level read looked less clean at year-end.

The answer for now is cautious. There is a real change in direction, but the numbers still do not prove that the shift from size to profitability is working. What 2025 shows most clearly is a bank going through a managerial and regulatory reset while credit costs and transition costs are rising faster than profit.

Three Points To Hold In Mind

• Growth did not disappear. Net interest income rose to $374 million from $328 million, the net loan book increased to $9.875 billion from $9.238 billion, and deposits increased to $12.503 billion from $12.066 billion.
• Profitability still weakened. Net income fell to $76 million from $89 million, ROE dropped to 5.5% from 7.0%, and Q4 net income fell to $13 million with ROE of just 3.6%.
• The end of the compliance chapter has not yet created free shareholder value. Bancorp paid no dividend in 2023 through 2025, and Discount states explicitly that transferring liquidity from IDBNY to the parent is constrained by local regulation.

Growth Held, Profitability Did Not

If one looks only at balance-sheet growth, it is easy to see why management keeps talking about a reset. Bancorp ended 2025 with $14.43 billion of assets, versus $13.86 billion a year earlier. Net loans rose to $9.875 billion, securities rose to $2.632 billion, and deposits rose to $12.503 billion. Non-interest income also increased to $83 million from $77 million.

The problem is that this growth did not translate into better bottom-line economics. Bancorp earned less, not more. A 5.5% ROE sits far below the group's 12.6% ROE, and even further below the 14.4% ROE generated by the group's banking operations in Israel. That is no longer a side issue. It is large enough to explain why IDBNY still sits on the drag side of the group thesis rather than the support side.

That chart is the core of the story. There is no balance-sheet collapse here and no obvious growth bottleneck. What there is, instead, is a bank that can still expand but is struggling to turn that expansion into acceptable profitability. So the real question is no longer whether there is enough activity. It is whether that activity is starting to look more profitable, or just larger.

Why This Still Does Not Look Like A Profitability Shift

The first key line is credit cost. Credit loss expense rose to $32 million in 2025 from just $2 million in 2024. That figure should not be read too simplistically. Discount explains that part of the increase came from extending the lookback period in the provision model from 2011 back to 2007. So not every dollar of the increase points to a sharp deterioration in underlying credit quality.

But it is also not a line that can be dismissed as purely technical. Alongside the model change, Bancorp ties the increase to loan-book growth and to a net $10 million rise in specific provisions, partly offset by changes in macro components. Put differently, even after one gives management credit for the methodological effect, 2025 still ends with a meaningfully higher risk-cost base than 2024.

The second key line is the expense base. Operating and other expenses rose by $46.5 million, or 16.6%, driven mainly by a $28.2 million increase in salary and employee-benefit expenses, a $6 million increase in maintenance and depreciation, and a $5.9 million increase in professional consulting expenses. This is exactly where the profitability-reset thesis gets tested. If new leadership and a new strategic plan are supposed to create a leaner and more profitable bank, 2025 mostly shows the bill for getting there.

The annual presentation is unusually direct on that point. On the strategy slide, IDBNY is framed under three items: new and smaller management, a new strategy that focuses on profitability rather than size, and the end of the consent order. But on the fourth-quarter expense slide, Discount also makes clear that Q4 included sizable costs at IDBNY following the management change. That is the classic bridge-year problem: the strategic logic is visible, but the cost arrives before the benefit.

So the 2025 read is not that IDBNY is failing. It is that IDBNY has not yet proven that the reset works economically. Net interest income improved. Non-interest income improved. The balance sheet grew. Yet the efficiency ratio still worsened to 71.6%, and to 76.4% in Q4. That is not what a business looks like once it has already come out the other side of a restructuring effort.

The End Of The Consent Order Is A Starting Condition, Not The End Of The Story

Important things did happen at the managerial and regulatory level in 2025 and early 2026. Avner Mendelson became CEO and board member of Bancorp and IDBNY on May 1, 2025. During 2025, IDBNY also formulated a new strategic plan aimed at improving operating efficiency and profitability, alongside an explicit commitment to simplify processes and activities and reduce costs relative to revenue.

The compliance chapter also moved forward in a meaningful way. IDBNY had signed parallel consent orders in May 2023 with the FDIC and the NYDFS. The FDIC order was terminated on October 10, 2025, and the NYDFS order was terminated on January 15, 2026. That is a real change, not just a headline. It removes one regulatory overhang that had hung above the U.S. operation for a long time.

But this is exactly where one should stop and be careful. Closing the consent-order chapter does not mean IDBNY has already turned into accessible value for Discount shareholders. It mainly removes an external constraint. It does not yet create a new profitability floor.

This may be the most important point in the whole continuation. In its group liquidity disclosure, Discount says explicitly that recognition of IDBNY's excess liquidity in the group model is limited because transferring liquidity from the U.S. subsidiary to the parent is constrained by local regulation. At the same time, IDBNY does carry an LCR above the regulatory minimum, and it has meaningful assets and loans pledged to the Federal Reserve Bank of New York and the Federal Home Loan Bank in support of its funding. That means the U.S. balance sheet is not under obvious stress, but it is also not truly free from a group-shareholder perspective.

That changes how the end of the consent-order chapter should be read. At the headline level, it is an achievement. At the shareholder-economics level, it is only a prerequisite. For the shift to start working in a meaningful way, IDBNY does not just need to be a cleaner bank. It needs to become a more profitable bank, with a calmer cost base and profits that feel less trapped.

What Has To Happen Next

The first checkpoint is ROE. After 5.5% in 2025 and 3.6% in the fourth quarter, the market will need to see a meaningful move away from those levels. Not because they imply crisis, but because they are too low for a subsidiary that is supposed to support group profitability rather than dilute it.

The second checkpoint is credit cost. Management has already given a credible explanation for part of the 2025 jump, but that explanation will carry less weight from here. If credit-loss expense stays elevated in 2026 even after the model reset, the read on the U.S. book will become tougher.

The third checkpoint is transition cost. 2025 only deserves to be called a reset year if 2026 shows a clear decline in salary, consulting, and other transition expenses linked to the management change and efficiency plan. Without that, "profitability reset" remains a story of intention rather than delivered economics.

The fourth checkpoint is value accessibility to the group. The market does not need to see an immediate Bancorp dividend to improve the read, but it does need to see that U.S. earnings stop feeling trapped. As long as there is no dividend, and as long as excess liquidity remains constrained in the group model, investors are likely to keep applying some discount to the quality of those earnings.

Conclusions

The current thesis in one line: IDBNY is showing a real change in direction, but so far investors are seeing the cost of the transition more clearly than its benefits.

What changed versus the simpler version of the story? The question is no longer whether the U.S. bank can grow. It is whether that growth can generate acceptable returns on equity and improve group economics. In 2025, the answer is still only partial. The end of the consent-order chapter, a new CEO, and a new strategic plan are all constructive starting points. Against that, 5.5% ROE, 3.6% in Q4, higher risk cost, and heavy transition expenses all say the shift has not yet been proven.

Counter-thesis: it is possible to argue that this reading is too harsh. Net interest income did grow, the balance sheet did expand, the compliance chapter did close, and transition costs do not have to become a permanent base. Under that reading, 2025 is not a new earnings floor but a bridge year.

What can change the market read over the short to medium term? One or two reports showing a lower efficiency ratio, calmer credit costs, and ROE moving back toward a more reasonable level. On the other hand, if Bancorp keeps showing growth without profitability, or if transition costs prove sticky, the market will keep reading IDBNY as a drag rather than a support layer.

Why this matters is straightforward. In a banking group, capital and liquidity inside a subsidiary are not automatically the same thing as shareholder value at the parent. IDBNY still has to prove not only that it is cleaner from a compliance perspective, but that it is also more profitable and economically less trapped.