Freesbe Finance: Credit Growth or a New Risk Center

The main article already argued that Carasso's bottleneck shifted from deliveries to the amount of capital and funding required to sustain growth. This follow-up isolates one layer inside that stretched balance sheet: Freesbe Finance. On the revenue line it can still look relatively small, with NIS 124.0 million of financing revenue in 2025. On the balance sheet it looks very different: a NIS 1.358 billion loan book, 20,068 loans, an average loan size of NIS 68 thousand, and a covenant that requires tangible equity of at least 12% of total assets.

That is exactly why this activity now has to be read as a real lending business, not only as an add-on that helps sell more vehicles. The real question is no longer whether the book is growing. It is how it is growing, what the underwriting metrics actually say, and what happens when the same platform is trying to open a mortgage channel at the same time.

Four points organize the picture:

• The activity is already too large to dismiss as marginal. The loan book rose to NIS 1.358 billion from NIS 1.116 billion in 2024 and NIS 863.6 million in 2023, and the investor presentation already plots the business on a growth curve from NIS 424 million in 2020.
• Growth still looks more disciplined than aggressive. The average loan rose to NIS 68 thousand, but average LTV fell to 58%, duration fell to 1.8 years, and the average interest rate fell to 7.5%.
• The real issue is not current credit losses, but capital. Freesbe Finance's covenant requires tangible equity of at least 12% of assets and restricts dividends or shareholder-loan repayments that would hurt compliance.
• Mortgages open a new runway, but without a real performance history yet. Headcount rose from 37 at year-end 2025 to 48 near the report date, and the new mortgage activity only began to roll out in March or April 2026, depending on which filing is used.

A Lending Business That Can No Longer Be Read As A Side Service

The easy way to underestimate Freesbe Finance is to look only at the revenue line. The company itself says that financing revenue still does not account for 10% or more of consolidated revenue, and it also says it cannot precisely quantify its market share in vehicle financing. That framing works for the income statement. It is incomplete for capital analysis.

The key table is not the revenue table. It is the book table:

The chart matters because it exposes the gap between the revenue line and the real balance-sheet weight of the business. Over five years the book more than tripled. Anyone reading the activity only through NIS 124 million of revenue is missing that a real credit book has already been built here.

What supports that growth is not an open-market lending brand sourcing customers independently. It is internal distribution through the group. Freesbe Finance operates through 75 sales points, depends on cooperation with Carasso's showrooms and used-car lots, and is described in the presentation as serving mainly group customers. That is a real advantage: direct access to the customer, a tight link to the vehicle transaction, and a sales environment built for speed. But it is also a real limitation: the growth rate of the loan book remains tied to the quality and mix of Carasso's own vehicle sales engine.

The customer mix says something important too. Roughly 90% of Freesbe Finance customers are one-time private borrowers taking a single loan, and the company says there is no concentration in any single customer. That reduces concentration risk. It does not eliminate risk, but it means the issue here is not a large-anchor exposure. It is broad underwriting quality at scale.

Growth Still Looks Controlled, But The Book Is Getting Heavier

The first headline here is actually constructive. On the surface, 2025 does not look like a year in which Freesbe Finance bought growth through looser terms. If anything, the opposite appears in the disclosed averages. The average loan rose by 6.2% to NIS 68 thousand, but average LTV fell to 58% from 60%, duration fell to 1.8 years from 1.9, and the average interest rate fell to 7.5% from 7.8%.

So the basic combination is this: a larger book, somewhat larger tickets, but no evidence in the disclosed metrics of higher collateral leverage or longer duration. That does not prove the underwriting model is perfect. But it does suggest that the company expanded volume without showing a clear deterioration in the core terms of the book.

What did change is origination mix. In 2023, loans for new vehicles accounted for 66% of origination volume. In 2024 that rose to 76%, and in 2025 to 80%. In shekel terms, lending for new vehicles reached NIS 614.1 million, versus NIS 152.4 million for used vehicles. This cuts both ways. On the one hand, new vehicles usually come with a cleaner collateral profile. On the other hand, it ties Freesbe Finance even more directly to Carasso's primary sales engine.

The annual movement in the book looked like this:

The book grew by 21.6% in 2025. But it matters how. New lending rose by 15.1% to NIS 766.4 million, while principal repayments also rose by 27.1% to NIS 525.0 million. This is not a static book that simply sits on the balance sheet. It is an active book with ongoing turnover, collection, and fresh production, which is still getting larger over time.

This is where caution enters. The presentation frames the picture as "low credit losses with no significant change over the past two years." At the ratio level, that description is broadly defensible. In the annual report Freesbe Finance records an allowance of about NIS 30.6 million, equal to 2.2% of its customer book. Relative to 2024, that is indeed broadly stable. But in absolute shekels, the allowance still rose from NIS 25.9 million to NIS 30.6 million, an increase of 18.2%.

The arrears read is similar. There is no obvious sharp deterioration, but there is no true standstill either. In the financing segment, gross exposures more than 120 days past due rose to NIS 11.1 million from NIS 9.0 million in 2024. In the dedicated overdue-debt section, the company reports 428 customers in legal collection with NIS 15.5 million of overdue amounts, plus 2 customers not yet transferred to legal handling with NIS 64 thousand overdue. This is still not a severe stress picture. But it is a reminder that a larger book also means larger absolute risk even if ratio metrics remain fairly stable.

The composition of the allowance matters as well. Out of the NIS 30.6 million total, about NIS 16 million is general allowance, about NIS 13.9 million is specific allowance, and about NIS 0.7 million is an excess allowance based on management judgment. That is a book still managed primarily through broad-model risk controls, not through one concentrated credit event. That is why the 2026 test is not just whether the allowance remains "low," but whether it can stay controlled while the book grows further and while the activity expands beyond vehicle-backed credit.

The Advantage Is Distribution, But The Bottleneck Is Capital

This is the core analytical read of Freesbe Finance. Anyone focusing only on arrears may miss the actual bottleneck. Right now the main issue is not current credit losses. It is capital discipline.

The company says this openly in its own critical-success-factors section. The activity requires meaningful equity and the ability to raise external credit at competitive cost. That admission matters. Freesbe Finance is not a business that can run on demand and sales execution alone. It also needs capital and funding access.

The covenant makes that reality much more explicit. As a condition for receiving credit, Freesbe Finance committed to keep tangible equity at no less than 12% of total assets at all times. Not plain accounting equity, but tangible equity, after excluding intangible assets and certain related-party exposures. In addition, it committed to avoid dividends or shareholder-loan repayments that would impair covenant compliance, and it also accepted restrictions around change of control, merger or split transactions, and third-party obligations that could hurt lender rights.

In plain terms, every NIS 100 of assets requires at least NIS 12 of tangible equity. That turns growth from a sales question into a capital question. As of December 31, 2025 Freesbe Finance remained in compliance, and its long-term loans net of current maturities stood at about NIS 246 million. But the structure itself shows the point: the company cannot chase volume without also maintaining balance-sheet support.

That is the difference between a financing add-on and a real lender. In an add-on model, the question is whether the product helps close a vehicle sale. In a real lender, the question is also whether the book can expand without consuming the capital cushion too quickly. Freesbe Finance is already in the second category.

Mortgages Open A New Lane, But Without A Real Track Record Yet

The next move matters because it changes the question, not only the product. In June 2025 the company applied to expand the activity into mortgage loans secured by real estate and into general-purpose loans secured by real estate. On November 16, 2025 the Capital Market Authority said it had no objection. The annual report says the activity began to be implemented on March 15, 2026 after the required preparations. The presentation published alongside the report describes the start of activity as April 2026.

That small date gap matters precisely because it shows how early the activity still is. Whether operational readiness began in March and market launch followed in April, or the opposite framing is preferred, the core point is the same: there is still no performance history to judge.

The presentation frames the move as entry into a new non-bank financing field aimed at mortgage advisers and at private and business customers, especially in more complex transactions, with faster approval processes and more flexible structures. That sounds attractive, but it also clearly changes the activity mix: from vehicle-backed lending built around group distribution, into real-estate-backed lending and a market with different sourcing, competition, and operating dynamics.

The staffing build already shows that this is not a minor pilot. At year-end 2025 Freesbe Finance employed 37 people. Near the report date the company was already reporting 48 employees, and it explicitly links the increase to preparations for the expansion of financing activity. That means cost and operational build-out came before investors were given any real history on mortgage volume, yield, arrears, or capital intensity.

It also matters how management itself defines the strategy. Mortgages do not appear as a side experiment. They appear as a formal strategic objective alongside broader product diversification, expansion into trucks, commercial vehicles and heavy equipment, efficiency measures, deeper cooperation with the group's showrooms and used-car lots, and higher profitability. In other words, this is not framed internally as a small option. It is framed as a planned growth layer.

That is why the right reading for 2026 is not "more upside." It is a first proof year. Investors will need to see not only whether there is demand for the mortgage product, but also how that new activity fits alongside a 12% tangible-equity covenant, a vehicle-credit book that is already at NIS 1.358 billion, and the need to preserve underwriting quality while entering a new lane.

Conclusion

Freesbe Finance is no longer a thin support layer whose main purpose is to help Carasso sell one more car. It is already a real credit business, with a meaningful book, a powerful internal-distribution engine, and a growth pace that deserves a standalone read.

The good news is that 2025 does not currently show a sharp underwriting deterioration. LTV is lower, duration is shorter, the allowance ratio remains reasonable, and customer concentration is low. The less comfortable part is that the activity is already subject to hard capital discipline exactly when it wants to expand into mortgages.

Bottom line: Carasso is gaining another growth engine, but also another risk center. If 2025 was the year in which Freesbe Finance became too large to ignore, 2026 will be the year that shows whether it can keep growing without consuming capital faster than it creates value.