PaymenT: Is the light-capital model actually working?
PaymenT sold NIS 280 million of credit in 2025, 52.3% of annual transaction volume, yet year-end still carried NIS 191.5 million of net customer credit and NIS 104.2 million of short-term bank debt. The model is clearly lighter in capital than straight balance-sheet lending, but it still relies on a meaningful warehouse phase before distribution.
What This Follow-Up Is Isolating
The main article argued that PaymenT has already built a real growth engine, but that the move to a lighter-capital model was still incomplete. This follow-up isolates the mechanism itself: what actually leaves the balance sheet when PaymenT sells or assigns loans, what still stays on it, and whether 2025 already proves that the company can keep growing without carrying most of the book with it.
The first conclusion: the off-balance-sheet mechanism is real, not cosmetic. The investor presentation shows NIS 280 million of credit sold in 2025, equal to 52.3% of annual transaction volume. The financial statements also say explicitly that loan sale and assignment transactions meet the derecognition rules for financial assets, so the loans are removed from the balance sheet when sold or assigned.
The second conclusion: this is still not a pure "almost no balance sheet" model. At the end of 2025 the company still carried NIS 186.7 million of customer credit held for realization and another NIS 4.8 million of loans at amortized cost that had not yet become sale-ready. Together that is NIS 191.5 million of net customer credit. At the same time, short-term bank credit rose to NIS 104.2 million. In other words, even after selling hundreds of millions of shekels of credit, there is still a meaningful warehouse layer sitting on the balance sheet.
The third conclusion: the right way to read PaymenT is neither as a classic balance-sheet lender nor as a clean marketplace. It is an originate, warehouse, distribute model. So the key question now is not whether the company knows how to sell portfolios. It has already proven that. The question is whether the sale cadence is fast enough so that growth stops settling on the balance sheet and on short-term bank funding.
This chart is the core of the continuation. It does not compare three like-for-like numbers. It compares annual flow with year-end stock. That is exactly why it matters: PaymenT is already pushing a material amount of credit out, but by year-end it still had a large stock of customer credit and a meaningful bank-funding layer attached to it.
What Really Leaves The Balance Sheet
The annual report explicitly describes a mixed model in which most activity already leans on True Sale, alongside some balance-sheet management of the credit book. The company funds operations through equity, bank credit, monthly loan assignments, and portfolio sales. That matters because the off-balance-sheet channel is not a one-off event. It is part of the regular funding structure.
The three portfolio-sale transactions disclosed in note 16 during 2025 totaled about NIS 224 million: roughly NIS 80 million in January, NIS 60 million in March, and NIS 84 million in October. The presentation shows NIS 280 million of credit sold during 2025. That makes it clear that the off-balance-sheet layer is broader than only those three large transactions, and the report indeed describes ongoing assignment activity as well.
What leaves the balance sheet is not just the accounting asset. In the accounting-policy note, the company says these transactions meet the derecognition rules, so sold or assigned loans are removed from the statement of financial position. In note 16, for the PaymenT Series 1 and PaymenT Series 2 transactions, the report also states that the buyer may not return a purchased loan to the company. That is already a strong sign that this is not simply bridge financing dressed up as a sale.
More importantly, the settlement mechanism in the SPC transactions says expected default is already built into the discount formula, while actual default does not affect the company's share. That is the clause that turns the off-balance-sheet sale into something economic, not merely accounting. In those structures, PaymenT is not only removing the asset from the balance sheet. It is also transferring most of the credit risk of the sold portfolio.
That said, the company does not disappear from the picture once the sale is done. It continues to provide servicing and collection work to the SPCs, although the fee is described as not material to the company. So what leaves the balance sheet is the loan portfolio and the main credit exposure, while the operating layer stays with PaymenT. That distinction matters: the company becomes less capital-intensive, not less operationally involved.
| Layer | Does it leave the balance sheet | What stays with PaymenT | Why it matters |
|---|---|---|---|
| Sold loan portfolio | Yes | The company keeps its economic share at the point of sale | This is the core derecognition event |
| Credit risk on disclosed SPC sales | Largely yes | No recourse, and actual default does not change the company's share | This is genuine risk transfer, not only accounting transfer |
| Servicing and collection | No | The company keeps managing the sold portfolios for a non-material fee | Operational involvement does not disappear |
| Sale proceeds | Not always fully and not always immediately | Part of the consideration is received over time | Asset-light does not mean instant full cash release |
This is where a shallow read can go wrong. When investors see "portfolio sale," it is easy to assume PaymenT simply cuts a portfolio and moves on. In reality, the sale is real at the asset and credit-risk level, but the model still leaves the company with servicing, buyer management, and in some cases cash that is released gradually rather than on day one.
What Still Sits On The Balance Sheet
The key number here is NIS 191.5 million of net customer credit at the end of 2025. Of that, NIS 186.7 million is measured at fair value as loans held for realization in the short term, meaning loans the company believes have a significant likelihood of being sold. That is already different from a model that holds the book to maturity, but it is still a book sitting on the balance sheet until the sale actually happens.
Alongside that sits another NIS 4.8 million of loans at amortized cost. The company explains that these are loans that have not yet become sale-ready in their current state, and that it is working to qualify them for sale. This bucket is still small, but it rose from NIS 1.0 million in 2024 to NIS 4.8 million in 2025. That matters because it signals that the issue is not only sale cadence. Part of the question is also how quickly newly originated loans become eligible for distribution.
This chart sharpens the gap between the "light-capital" story and the actual shape of the balance sheet. Most of the book sitting on the balance sheet is indeed classified as inventory on the way to distribution, but that inventory still grew materially. In other words, the off-balance-sheet machine is working, but not yet at a speed that makes the warehouse stage irrelevant.
This connects directly to the funding side. Short-term bank credit rose from NIS 82.4 million to NIS 104.2 million, and the board report explicitly says the increase was caused by the accumulation of a loan portfolio for a near-term assignment. That tells two stories at once. On one hand, the rise in debt is not automatically a distress signal. On the other hand, it proves that the balance sheet still serves as a meaningful intermediate warehouse for growth.
There is also a smaller but instructive friction layer. Other long-term receivables rose to NIS 2.3 million, and the report ties that increase to amounts due from portfolio sales. Note 16 says part of the consideration from assignments is transferred according to the repayment pace of the assigned portfolio and, in the meantime, is deposited and held in the purchasing entities in a bank deposit bearing roughly 4% interest. That is another reminder that removing an asset from the balance sheet is not the same thing as releasing all of the capital immediately.
Around the report date, the company also flagged another roughly NIS 80 million transaction. Whether one reads that as a sale already close to signing or as a year-end balance already lined up for disposal, the implication is the same: part of the end-2025 balance was credit inventory waiting for another sale window, not necessarily credit meant to remain on balance sheet for long.
Is The Light-Capital Model Actually Working?
The short answer is yes, but not all the way. If "light capital" means the company has built an engine that lets it grow without keeping every new shekel of lending on its own balance sheet until maturity, then 2025 gives a positive answer. The company sold NIS 280 million of credit in 2025, more than half of annual transaction volume. Average bank-line utilization was only 19.9%. Out of total bank facilities of NIS 340 million, NIS 236 million were unused at year-end. Financing expense even fell 10% to NIS 4.7 million thanks to better funding costs with the banks. That is hard to read as liquidity pressure.
The covenant layer also does not look stressed. Senior financial debt as a percentage of loans and receivables stood at 54% against an 80% ceiling. Tangible equity as a share of the balance sheet stood at 35% against a 20% minimum, and tangible equity itself stood at NIS 82 million against a NIS 25 million minimum. The company also says it has no material long-term loan or credit exposure. In other words, the new model has created real funding flexibility.
But if "light capital" means the balance sheet is no longer a meaningful bottleneck, the answer is still no. At the end of 2025 PaymenT still carried NIS 191.5 million of net customer credit against equity of NIS 82.2 million. Even after all of the sale activity, the on-balance-sheet loan book still stood at more than 2.3 times equity. That is not the picture of a pure intermediary platform. It is the picture of a credit company that has built a distribution engine, but still carries a meaningful warehouse layer on the way.
That is exactly why the right label here is warehouse-to-distribute. The balance sheet no longer carries the entire life of the loan, but it still carries the transition stage between origination and sale. In that kind of model, the bottleneck shifts from "is there enough capital" to "what is the rotation speed." If origination keeps growing and the sold share stays high enough to prevent quarter-end inventory and bank usage from compounding with it, the light-capital transition is genuinely progressing. If not, this is a more efficient model, but not yet a fully light-capital one.
What Has To Happen Next
- The sold share needs to stay at least around half of annual transaction volume even as origination continues to grow.
- The on-balance-sheet loan book and short-term bank debt need to start growing more slowly than total business activity.
- The bucket of loans that are not yet sale-ready needs to stay small, otherwise the bottleneck shifts from warehousing to portfolio eligibility itself.
- The roughly NIS 80 million transaction flagged around year-end needs to turn into actual completed deals and a repeatable cadence, not remain a one-off.
Conclusion
Asset-light? Yes, but not in the pure sense the headline may suggest. What truly left the balance sheet was the sold loan portfolio, together with most of its credit risk. What still stayed on it was a meaningful credit inventory until sale, loans that still were not ready for distribution, part of the consideration that arrives gradually, and the servicing layer on sold portfolios.
So 2025 does prove that the model works as a funding engine. It still does not prove that the balance sheet has become marginal. The real test over the next 2 to 4 quarters is very simple: can PaymenT keep growing without another similar step-up in the quarter-end loan book and in short-term bank funding. That is the line between an efficient hybrid model and one that has truly become light-capital.
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.