Isracard: Credit Is Growing Fast, but What Is Happening To Price And Quality
Isracard expanded consumer and business credit rapidly in 2025, but average yields and spreads fell in both books. Business credit reached NIS 3.26 billion, yet segment net profit slipped to NIS 199 million, which means the 2026 question is no longer only how much credit was sold, but at what price and with what quality.
Starting point
The main article argued that capital had become tighter as Isracard pushed credit growth forward. This follow-up steps back one layer and isolates the question that comes before capital: is credit still being sold at the right price, and does portfolio quality really justify the pace of growth.
The 2025 answer is less clean than the headline growth story suggests. In both credit engines, consumer and business, balances grew quickly, but average interest yield and pricing spreads moved lower. That does not automatically prove weak underwriting. It does show that the company expanded at a lower price per unit of risk.
The sharper issue sits in business credit. This is the engine Isracard has been presenting as a core growth pillar, and in 2025 it did grow aggressively. But this is also where the gap between volume and economics is clearest. Business credit rose 28% to NIS 3.259 billion, the average interest-income rate fell to 8.08% from 8.57%, the spread above prime fell to 2.11% from 2.57%, and segment net profit slipped to NIS 199 million from NIS 203 million. In other words, Isracard sold more credit but did not turn that into higher segment profit.
And this happened in a year when credit cost looked favorable. In the business segment, credit-loss expense fell to only NIS 1 million from NIS 22 million. So this is not yet a story about a blow-up that has already happened. It is a story about growth economics weakening before the book has even gone through a full credit cycle. The fourth quarter, when group credit-loss expense climbed back to NIS 74 million after NIS 39 million in Q2 and NIS 45 million in Q3, is the reminder that the argument is not settled.
Price is falling on both sides of the book
Isracard’s own reporting describes intense competition in both consumer and business credit, expressed through pricing, terms and speed of response. The 2025 numbers fit that description well: the book got larger, but the average price on that credit moved lower.
| Metric | Consumer 2024 | Consumer 2025 | Business 2024 | Business 2025 |
|---|---|---|---|---|
| Interest-bearing credit balance | 7,253 | 8,583 | 2,537 | 3,259 |
| Interest-income rate | 11.44% | 11.11% | 8.57% | 8.08% |
| Spread above prime | 5.44% | 5.14% | 2.57% | 2.11% |
| Problematic-loan ratio | 3.28% | 3.30% | 2.60% | 2.19% |
| Non-accrual ratio | 0.78% | 0.83% | 1.30% | 0.72% |
Consumer credit looks more comfortable at first glance. The book rose 18% to NIS 8.583 billion, interest income from it increased to NIS 869 million from NIS 812 million, and the credit-loss expense ratio fell to 1.83% from 2.19%. But price still moved lower. The average interest-income rate fell to 11.11% from 11.44%, and the spread above prime fell to 5.14% from 5.44%.
In business credit the pricing decline is sharper. The book grew by NIS 722 million in a single year, yet interest income rose only to NIS 356 million from NIS 310 million, while the average income rate compressed by 49 basis points. The spread above prime narrowed by 46 basis points. This is no longer just growth. It is growth sold on softer terms.
And the effect is not only theoretical. In the segment disclosures, the translation into profit has already started to show up. Business-segment revenue rose 7.9% to NIS 892 million, but the cost-to-income ratio excluding credit losses worsened to 67.2% from 65.1%, and segment ROE fell to 29.9% from 33.9%.
In business credit, volume is already large, but the economics are weaker
If one place best captures the gap between quantity and quality, it is the business segment. Isracard presents a growth story there, and the volume data supports that. But at the economic level, 2025 shows that this engine is no longer enjoying the same pricing power.
That chart is the core of the continuation thesis. Even after a year in which business-segment credit-loss expense almost disappeared, net profit did not rise. It slipped by NIS 4 million. That matters because 2025 still did not test this book under meaningful credit-cost pressure. It tested it under relatively benign credit conditions.
Put differently, if the business book grew 28% but did not produce higher net profit even in a year with almost no credit-loss burden, the problem is not only future credit quality. The problem is that price has already weakened while operating and marketing expenses are still moving higher.
The structure of the book reinforces that reading. Out of NIS 3.259 billion of business credit at year-end 2025, about NIS 2.883 billion was credit to merchants. Within the total business-credit book, NIS 2.389 billion was originated in 2025 alone and another NIS 547 million in 2024. That means about 90% of the business book was built in the last two years. It is still a very young book.
The economic implication is straightforward: Isracard does not yet know whether today’s pricing on business-credit growth will fully cover risk once the 2024 and 2025 vintages age. What can already be said is that at the current profit level, the answer is not yet convincing.
Annual quality ratios look cleaner, but the book has not fully proved itself
A reader who looks only at annual quality ratios can reach a conclusion that is too comfortable. In business credit, the problematic-loan ratio fell to 2.19% from 2.60%, the non-accrual ratio fell to 0.72% from 1.30%, and the credit-loss expense ratio fell to 0.06% from 0.70%. In consumer credit, the loss-expense ratio also improved, to 1.83% from 2.19%.
But there are two details behind those ratios that matter much more than they first appear.
The first is the denominator. Consumer credit rose to NIS 8.583 billion from NIS 7.253 billion. Business credit rose to NIS 3.259 billion from NIS 2.537 billion. When a book grows that quickly, a stable ratio can still conceal a rise in weaker balances in absolute terms.
That is exactly what happened on the consumer side. Total problematic balances in consumer credit rose to NIS 392 million from NIS 334 million, and non-accrual balances rose to NIS 90 million from NIS 69 million. The ratio barely moved, but the amount did. In business credit, problematic balances rose to NIS 82 million from NIS 80 million, while non-accrual business balances fell to NIS 28 million from NIS 41 million. That is an improvement in mix, not a disappearance of the issue.
The second detail is that coverage moved lower while the book expanded. The allowance ratio on consumer credit fell to 2.92% from 3.23%, and on business credit to 2.39% from 3.31%. In business credit, the total allowance fell to NIS 94 million from NIS 105 million even though the book expanded sharply and problematic balances in absolute terms did not truly decline. That is not necessarily a mistake. It is, however, a choice that assumes 2025 represents a cleaner credit environment. The question is whether that assumption will survive as the younger book seasons.
This chart explains why 2025 is still not a full proof year for credit quality. About NIS 5.533 billion of consumer credit, and NIS 2.936 billion of business credit, comes from 2024 and 2025 vintages. That is roughly two-thirds of the consumer book and almost the entire business book. As long as the book is this young, a low credit-loss ratio is encouraging, but it is not a final verdict.
The fourth quarter broke the comfort
What matters most is that the easing in credit quality did not continue in a straight line into year-end. Group credit-loss expense dropped to NIS 39 million in Q2 and NIS 45 million in Q3, but climbed back to NIS 74 million in Q4. That was also above the NIS 50 million recorded in Q4 2024.
That is not enough to declare that the trend has already reversed. It is enough to break the easy reading that 2025 simply improved quarter after quarter. If Q4 turns out to be the beginning of normalization rather than noise, then 2025 ends not with a clean proof of stronger quality, but with an open question.
This is also where the link back to the main article becomes sharper. Business customers already consume NIS 22.034 billion of the group’s risk-weighted assets, versus NIS 6.214 billion for private customers. So pricing erosion in the business engine is not only a margin question. It is also a capital-efficiency question. If the part of the balance sheet that consumes more capital is growing at a lower price, without showing matching profit leverage, the issue does not stay inside a segment note. It spills directly into the capital discussion from the main article.
Bottom line
2025 does not say that Isracard has lost control of credit. It does say that growth is no longer free. Price fell in both books. In consumer credit, that still sits on top of a year with lower loss cost, but also with higher absolute problematic balances. In business credit the signal is more acute: the book expanded, pricing weakened, and profit still failed to rise even after an unusually benign credit-loss year.
That sets up the 2026 test. Isracard now needs to prove three things at once: that yield and pricing spreads can stabilize, that the business segment can turn balance growth into profit growth, and that Q4 2025 was not the opening stage of a higher credit-loss cycle as the 2024 and 2025 vintages mature.
If that proof arrives, 2025 will look like a reasonable build year. If it does not, the conclusion will be that part of Isracard’s growth was simply bought at too low a price.
Disclosure: Deep TASE analyses are general informational, research, and commentary content only. They do not constitute investment advice, investment marketing, a recommendation, or an offer to buy, sell, or hold any security, and are not tailored to any reader's personal circumstances.
The author, site owner, or related parties may hold, buy, sell, or otherwise trade securities or financial instruments related to the companies discussed, before or after publication, without prior notice and without any obligation to update the analysis. Publication of an analysis should not be read as a statement that any position does or does not exist.
The analysis may contain errors, omissions, or information that changes after publication. Readers should review official filings and primary sources before making decisions.