The First International Bank of Israel reported its second quarter results this morning, 12 August 2026.
Errors or inaccuracies are possible. Spotted something that looks wrong? Write to me and I will correct it.
Three Banks in Three Days
This is the third bank to report this week, and for the first time this year we have three different answers to exactly the same question.
Hapoalim reported on Monday a 2.1% fall in quarterly profit, with the financing margin eroding from 2.90% to 2.70%. Leumi reported this morning an 8.5% rise in profit, with its interest income up just 0.7% - meaning frozen.
And FIBI reports this morning an 8.5% fall in profit, and its interest income did not freeze - it fell.
The Quarter
| NIS millions | The quarter | A year ago | Change |
|---|---|---|---|
| Net interest income | 1,192 | 1,290 | -7.6% |
| Credit loss income (expense) | 38 | 16 | |
| Non-interest income | 571 | 551 | +3.6% |
| of which: fees | 473 | 434 | +9.0% |
| Operating and other expenses | 812 | 793 | +2.4% |
| of which: salaries and related | 450 | 449 | +0.2% |
| Net profit to the bank's shareholders | 583 | 637 | -8.5% |
| Basic EPS, agorot | 58.2 | 63.5 |
And for the first half:
| NIS millions | The half | A year ago | Change |
|---|---|---|---|
| Net interest income | 2,282 | 2,444 | -6.6% |
| Non-interest income | 1,129 | 1,065 | +6.0% |
| of which: fees | 937 | 859 | +9.1% |
| Operating and other expenses | 1,626 | 1,585 | +2.6% |
| Net profit to the bank's shareholders | 1,063 | 1,167 | -8.9% |
What Actually Happened Here
Net interest income fell 7.6%.
This is the figure that separates FIBI from the other two banks. At Leumi this line rose 0.7% - frozen, but not negative. At FIBI it genuinely fell.
And comparing it against the volume of credit makes it sharper: credit to the public grew 20.1% over the year - the fastest of the three. The bank lent a fifth more money, and earned less on it.
And it shows up in a ratio the bank publishes itself: net interest income against average assets fell from 2.0% to 1.7% in the quarter. Three tenths of a percentage point, on an asset base that grew.
In other words: the erosion here was not offset by volume. It was swallowed by it.
And What Did Work
Fees rose 9.0% in the quarter and 9.1% in the half.
That is a good pace, and it is consistent between the quarter and the half - meaning this is not a one-off event.
But the order of magnitude is completely different from Leumi's. At Leumi, non-interest income jumped 39.4% and contributed an extra NIS 570 million. At FIBI it rose 3.6% and contributed NIS 20 million.
And against a NIS 98 million fall in interest income, that is not enough.
And the customer asset portfolio grew 20.8%, to NIS 1,227 billion. This is the base from which fees are derived over the long run, and it is growing faster than the income it currently generates.
An Expense Line Worth Pausing On
Operating expenses rose 2.4% in the quarter. Salaries within them rose just 0.2%.
That is good cost control in its own right - and particularly alongside credit that grew 20%. The bank grew without growing its payroll.
And still, the efficiency ratio deteriorated:
| The quarter | A year ago | |
|---|---|---|
| Efficiency ratio | 46.1% | 43.1% |
| Efficiency ratio, half | 47.7% | 45.2% |
And the reason is not the expenses - it is the income. The efficiency ratio is expenses divided by income, and when income falls faster than expenses rise, the ratio deteriorates even if the expenses themselves behaved well.
And this is precisely the opposite of Hapoalim, where the efficiency ratio actually improved, from 32.8% to 30.6%.
The Credit Line, and Here There Is a Question
The credit loss line recorded income of NIS 38 million in the quarter, against income of NIS 16 million a year ago. Meaning the bank released provisions rather than making them.
And as a rate: the credit loss expense against the average balance stood at minus 0.10% in the quarter, against minus 0.05% a year ago.
And the quality metrics support it: the rate of credit not accruing or 90 days or more past due fell to 0.40% from 0.46%. The allowance rate fell to 0.96% from 1.19%.
These are good numbers. Quality metrics improving, not deteriorating.
And still I place a question mark here: releasing provisions increases reported profit, and it is happening on a book that grew 20.1% in a year. New credit has not yet had time to be tested. The allowance rate is falling exactly when the denominator - the credit balance - is growing fastest.
That does not mean anything is wrong. It means this metric is less informative than usual this year.
Capital and the Balance Sheet
| 30.6.2026 | 30.6.2025 | 31.12.2025 | |
|---|---|---|---|
| Tier 1 equity ratio | 10.87% | 11.54% | 11.10% |
| Leverage ratio | 4.85% | 5.26% | 5.04% |
| Liquidity coverage ratio | 127% | 134% | 129% |
| Net stable funding ratio | 122% | 125% | 127% |
And shareholders' equity stands at NIS 14,899 million.
All four ratios fell together, and that follows: the bank grew credit 20.1% and deposits only 11.7%. When the risk side of the balance sheet grows at twice the pace of the side funding it, capital and liquidity ratios erode on their own.
All the ratios remain above regulatory requirements, and the board's capital target is 9.50%. But the direction of all of them is the same.
Return on Equity
| The quarter | A year ago | The half | A year ago | |
|---|---|---|---|---|
| Return on equity | 16.0% | 18.3% | 14.5% | 17.1% |
| Adjusted return | 20.0% | 18.4% |
The adjusted return neutralises two things, per the bank's own definition: the excess of the tier 1 ratio over the board's target, and the special tax levy applying to the bank in 2026.
The gap between 16.0% and 20.0% is four percentage points, and that is an order of magnitude worth noting - but it is worth remembering this is a theoretical return. The investor receives the 16.0%.
הזווית שלי
דעה אישית של אילן אברמוב - לא ייעוץ ולא המלצה
All three banks reported the same problem, and each answered it differently. This is the most instructive comparison we received this week.
Hapoalim answered on expenses. Its efficiency ratio improved from 32.8% to 30.6% in a single quarter, which is an impressive move. Its profit still fell 2.1%.
Leumi answered on other income. Plus 39.4% in non-interest income. Its profit rose 8.5%.
FIBI answered on volume - and that answer did not hold. Credit grew 20.1%, the fastest of the three, and interest income still fell 7.6%.
And the conclusion I take from it: volume is the weakest of the three answers.
The reason is simple. When the margin erodes, growing volume multiplies a smaller number by a larger one. That works only if the growth rate in volume exceeds the erosion rate in the margin - and here it did not. The bank grew the book by a fifth and still earned less on interest.
And there is a further cost that does not appear in the profit line. Credit growing 20% a year, against deposits that grew 11.7%, erodes the capital and liquidity ratios in parallel. All four ratios the bank publishes fell together. This is not a dangerous position - it is a position with a limit.
And what I do value in this report is two things.
The first - cost control that is genuinely exceptional. Salaries rising 0.2% on a book that grew 20% is a real operational achievement, even if the efficiency ratio conceals it because income fell.
And the second - 20.8% growth in the customer asset portfolio, to NIS 1,227 billion. This is the base from which fees are derived, and fees are indeed rising at a consistent 9%. If that trend continues, it builds FIBI precisely the alternative income engine that saved Leumi's quarter - only several years later.
And what I put an asterisk on is the provision release. The bank recorded NIS 38 million of income in the credit loss line, meaning it released provisions, on a book that grew 20.1%. The quality metrics genuinely improved, and I am not claiming anything is being concealed. I am claiming that credit one year old has not yet told its story, and that when the denominator grows this fast, every ratio divided by it looks better than it is. This line is worth checking again in four quarters, not today.
