Altshuler Shaham Finance Construction Credit: Growth Is Moving Into Harder-to-Measure Collateral
Construction credit jumped to NIS 82.4 million, but the deterioration is not yet visible in provisions. The more sensitive signal sits in the collateral: second-ranking liens, project-surplus collateral, and a roughly NIS 52 million guarantee tied to the acquisition of a construction-lending portfolio.
The broader first-quarter analysis showed that Altshuler Shaham Finance has enough bank appetite to keep expanding its credit activity, but still has to prove that growth is not coming at the expense of credit quality. This continuation isolates the place where that question becomes sharper: construction credit. The accounting picture still looks calm, because all construction credit is classified in Stage 1 and the expected-loss provision is even lower than it was at the end of 2025. Yet in a single quarter, construction credit grew by 69.2%, and the collateral moved from a simpler structure of project surplus only to a mix of project surplus and second-ranking liens. A roughly NIS 52 million bank guarantee tied to the acquisition of a construction-lending portfolio was added as well, so the exposure is no longer only direct loans to developers. The current conclusion is not that credit has already deteriorated. It is that growth has moved into a layer where collateral quality is harder to measure from one quarterly report. The next proof point is whether construction growth can continue without migration to Stage 3, without a larger share above 70% LTV, and without any sign that the new guarantee is turning from contingent exposure into an actual loss.
Growth Is Showing Up In Collateral, Not In Provisions
Construction credit is still small relative to the total credit portfolio, but it was the fastest-growing part of the book in the quarter. Net construction credit rose from NIS 48.7 million at the end of 2025 to NIS 82.4 million at the end of March 2026. That is a 69.2% increase in three months. The income statement and provision table do not yet show a problem: gross construction credit was NIS 82.9 million, all of it in Stage 1, and the expected-credit-loss provision was only NIS 473 thousand. The expected-loss rate declined from 0.80% at the end of 2025 to 0.57% at the end of March.
That gap between growth and provisioning is exactly why the analysis has to move to collateral. At the end of 2025, all construction credit secured by real estate sat in the 30% to 50% LTV bucket. By the end of March 2026, NIS 16.9 million already sat above 70% LTV, about 21% of construction credit secured by real estate. The amount is not large in absolute terms, but it changes the quality of the growth: this is not just more credit, but credit with part of the balance now sitting in a more sensitive financing layer.
| Construction Credit Metric | 31 Dec 2025 | 31 Mar 2026 | What Changed |
|---|---|---|---|
| Net construction credit | NIS 48.7 million | NIS 82.4 million | 69.2% quarterly growth |
| Construction credit secured by real estate | NIS 49.1 million | NIS 80.9 million | Almost the entire book relies on real-estate collateral |
| Construction credit above 70% LTV | 0 | NIS 16.9 million | A risk layer that did not exist at year-end |
| Expected-loss rate | 0.80% | 0.57% | The provision does not yet reflect stress |
| Collateral type | Project surplus only | Project surplus and second-ranking liens | The collateral became harder to read |
Two Collateral Types Require More Caution
For a non-bank lender, collateral is not a technical detail. It is a core part of risk pricing. Ordinary business credit can look similar in a table even when recoverability is very different between a first-ranking lien on a clear asset, a second-ranking lien on land, and a right to project surplus. In Altshuler's construction credit, that distinction is no longer theoretical.
The construction customer table shows two relatively large customers backed by project surplus: NIS 50 million of utilized credit against collateral value of NIS 147.3 million and 34% LTV, and another NIS 15 million against collateral value of NIS 32 million and 47% LTV. Those ratios look lower, but project surplus is not cash in the bank. It depends on sales pace, collections, remaining costs, and the ability to bring the project to an economically successful finish.
The more sensitive layer is two customers backed by second-ranking liens on land. One shows NIS 10 million of utilized credit and 92% LTV, and the other shows NIS 7 million and 89% LTV. Total collateral value in second-ranking liens is NIS 155.1 million, but lien ranking matters as much as the reported value. When the collateral is second-ranking, creditor protection depends on what sits ahead of it in the lien structure, not only on the gross asset value.
The company also states that construction-credit collateral at the end of March partly consists of second-ranking liens and otherwise of project surplus, and that project-surplus value is forward-looking information that is not fully under the company's control and is uncertain. That sentence turns construction credit into a separate issue from the general discussion of credit quality. As long as there are no arrears or Stage 3 balances, the provision can stay low. But if a project stalls, future surplus and a second-ranking lien may behave very differently from simpler collateral.
The NIS 52 Million Guarantee Adds A Different Risk Path
The guarantee provided by Altshuler Construction on 31 March 2026 is not another loan in the table. It is a bank guarantee of roughly NIS 52 million, provided as part of the financing for a third party's acquisition of a construction-lending portfolio. Altshuler's guarantee equals 10% of the portfolio purchase amount, and it was added alongside a subordinated guarantee from the seller equal to 2.5% of the portfolio. The guarantee may be called if the cash-receipt stream from the acquired construction-lending portfolio is impaired, and only after the seller's subordinated guarantee is exhausted.
The economic meaning is that the company now has exposure to the cash-receipt stream of an acquired construction-lending portfolio, not only to projects it directly finances in its own credit portfolio. The disclosure does not break down the acquired portfolio by project, LTV, collection pace, or borrower diversification. The guarantee therefore does not prove a problem, but it adds exposure that is hard for readers to measure from the quarterly filing. If construction activity remains small and diversified, this could be a complementary product with future profitability. If guarantees grow without better disclosure on the portfolios behind them, the risk quality will be less transparent than the reported growth.
What Will Decide The Quality Of Growth
Altshuler's construction credit does not currently look like a credit-loss problem. It looks like a change in collateral mix, and that distinction matters. The low provision and the absence of Stage 3 balances support a calmer read, but rapid growth, the move into above-70% LTV, second-ranking liens, and the NIS 52 million guarantee set a higher proof threshold for the next few quarters. For the market to treat this growth as a quality expansion, the company will need to show that portfolio size is not running ahead of its ability to measure and realize collateral. The next increase in construction credit will be positive only if it comes with stable arrears, stable provisions, controlled high-LTV exposure, and clearer disclosure around guarantees and construction-lending portfolios entering the platform.
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.