Blender Techno 2025: Profit Is Starting To Return, but the Real Story Is Life After P2P

Company Overview

Blender in 2025 is first and foremost a digital lender, and only then a software story. That matters because the headlines around distribution partnerships, a new SaaS style service bureau, and a proprietary underwriting system can create the impression of a fintech platform already moving into a software monetization phase. In practice, the two engines that still generate almost all of the economics are consumer lending in Israel, mainly at the point of sale, and the legacy P2P credit intermediation platform. Recognized revenue from technology services was only NIS 198 thousand in 2025, and all continuing operations revenue was generated in Israel.

What is working right now? The Israeli lending segment grew quickly. Its credit book rose to NIS 181 million from NIS 117 million at the end of 2024, gross segment revenue increased 59% to NIS 26.1 million, and segment operating profit increased 64% to NIS 4.6 million. At the consolidated level, continuing operations revenue rose 37% to NIS 41.8 million, net revenue rose to NIS 28.1 million, and the second half already showed operating profit of NIS 266 thousand versus an operating loss of NIS 4.3 million in the second half of 2024. That is real progress.

What is still not clean? The P2P engine is no longer moving toward growth, but toward contraction. The managed book fell to NIS 381 million from NIS 435 million, and in January 2026 the company stopped originating new P2P loans altogether. The security fund was depleted by the report publication date, and from January 2026 it began making only partial payments to investors based on whatever cash was available on each payment date. At the same time, cash on hand rose to NIS 32.6 million, but that increase was built mainly on a roughly NIS 20.8 million net equity raise and NIS 19.2 million of positive financing cash flow, while operating cash flow was negative NIS 1.9 million.

This is the main trap in the story. Anyone reading only the "turnaround year" headline could conclude that Blender has already moved from a volatile credit company to a cleaner and profitable fintech platform. That is too early. The company itself explicitly states that it has no data showing that the system's predictive ability around credit default is better than that of other underwriting systems. So even if the technology layer is real, the proven edge today sits more in speed, operations, and distribution integration than in a software engine that has already demonstrated clear standalone commercial superiority.

There is also a practical market filter to keep in mind. On the latest trading date in the market snapshot, turnover in the stock was only NIS 49.5 thousand, and short interest was effectively zero. That means the business story may improve faster than the broader market's ability to act on it comfortably.

Before getting into the details, four points matter more than the headline:

1. The move to operating profit in the second half came from a combination of rapid growth in Israeli lending and cost cuts in a shrinking P2P segment, not from a breakthrough in technology sales.
2. The P2P segment moved into operating profit precisely because activity was shrinking and expenses were lower, not because the engine improved fundamentally. A few weeks after year end, it had effectively become a runoff business.
3. The rise in cash and the move to positive working capital depended heavily on fresh equity and financing, not on a model that is already funding its own growth.
4. The software thesis is still option value. In 2025, accounting revenue from technology services was only NIS 198 thousand, while one technology development MOU generated a non material phase one payment and then stopped before a dedicated vehicle was formed.

The Economic Map

Blender now operates across two worlds, but they are not equal in economic weight:

The headcount tells the same story. At the end of 2025 the group had 60 employees versus 68 a year earlier. Employee count in the Israeli lending segment rose to 15 from 12, while P2P headcount fell to 23 from 28. In practice, Blender is reallocating people and cost toward point of sale lending while the legacy platform contracts.

This chart sharpens what is easy to miss in the headlines. Even after a year full of technology language, almost all reported revenue still comes from two Israeli credit engines. That does not mean the technology has no value. It does mean it still cannot carry the core thesis.

The picture is straightforward: Europe is almost gone, Israeli lending is growing, and P2P is shrinking. So the right way to read Blender today is not as a company with three equal engines, but as an Israeli lending company rebuilding itself around BNPL while the P2P engine fades and the software story still needs commercial proof.

Events And Triggers

The first trigger: Blender has almost fully closed the European chapter. During 2025, Blender Lithuania sold the loan portfolio in Lithuania and Latvia for total consideration of roughly NIS 27.9 million, and that cash was used to repay the European credit lines. By year end, all institutional borrowings in Europe had been repaid and the financing agreements had ended. This is not only a reduction in noise. It removes a layer of foreign funding and an activity that was no longer central to strategy.

Alongside that, on August 14, 2025, the company also signed an agreement to sell its full holding in Blender Lithuania for about EUR 725 thousand, with a possible additional EUR 100 thousand if the buyer receives approval to connect to the CENTROlink payment system within 12 months of central bank approval. By the report date, the central bank approval had already been received, and management expected the remaining conditions to be completed during the first half of 2026. In other words, Europe is no longer a growth leg. It is an exit process still finishing the remaining administrative steps.

The second trigger: in January 2026 the company did what the 2025 numbers were already pointing toward, it stopped originating new P2P loans. The formal explanation was that changes in investor preferences driven by the higher rate environment led to lower deposits, lower origination volume, and erosion of the security fund. That is a relatively polite way to describe a sharp move. Economically, it means a funding model that was supposed to connect investors and borrowers lost competitiveness against other liquid savings alternatives.

The third trigger: a few days after stopping new P2P originations, Blender signed an agreement to establish a service bureau in a SaaS model with a technology partner and a third party. Blender and the partner will each hold 40%, while the third party will hold 20%. The new vehicle is meant to sell technology, operational, and regulatory credit services to credit players in Israel. Strategically this makes sense, because it attempts to turn Blender's internal system into an external product. But the right framing is still optionality, not a current profit engine.

This is exactly where the necessary balance comes in. During 2025, Blender also signed an MOU with an unrelated US financial entity to provide technology development services for an auto dealer financing platform. The company completed its phase one work and received a non material payment, but the partner later said the special purpose vehicle was never formed and the engagement stopped. That is the core point. The technology sales story exists, but so far it still lives more in partnerships, MOUs, and strategic language than in recurring and material revenue.

The fourth trigger: in November 2025 Blender completed a private placement of 4.13 million shares and 2.06 million options for total proceeds of around NIS 21 million, about NIS 20.8 million net. The company itself tied the raise to three uses, strengthening equity, supporting faster BNPL growth, and continuing to develop and sell technology services. This matters because it explains both why equity rose to NIS 54.6 million and why 2025 should not be read as a clean year of organic profitability.

Efficiency, Profitability And Competition

The profitability improvement is real, but it does not come from one place. At the consolidated level, continuing operations gross revenue rose to NIS 41.8 million from NIS 30.5 million, net revenue after funding costs and credit losses rose to NIS 28.1 million from NIS 21.6 million, and operating loss narrowed to NIS 1.5 million from NIS 10.2 million. That is a major improvement, and it came from both growth and cost discipline.

What Actually Drove 2025

In the Israeli lending segment, Blender grew both volume and contribution. It originated NIS 164 million of new credit in 2025 versus NIS 108 million in 2024, and the second half alone already reached NIS 93 million versus NIS 58 million in the second half of 2024. Segment revenue rose 59%, and segment profit rose to NIS 4.6 million. The company's presentation also shows that unit economics did not collapse as the book scaled. Average annualized gross income in the Israeli lending segment was 17.5% versus 17.7% in 2024, interest expense rate fell slightly to 6.7%, actual credit loss rate fell to 1.6% from 2.3%, and net margin improved to 9.2% from 8.7%.

That is a positive point, because it suggests the BNPL growth was not driven only by pushing price higher to customers. Average effective cost to the customer actually fell to 16.9% from 17.7%, and the company's average yield from the customer fell to 16.3% from 17.2%. In other words, Blender grew through volume and distribution, not only through harsher pricing. On the other hand, credit quality still needs attention. The allowance ratio on the Israeli loan book rose to 4.7% from 4.1%, and credit impaired loans rose to NIS 12.7 million from NIS 6.8 million.

This chart holds two truths at the same time. The Israeli book grew strongly, but the allowance cushion did not fall with scale, it increased. So 2025 proves expansion, but it also shows that the risk bill did not disappear.

P2P Did Not Improve, It Simply Shrunk

This is one of the more interesting gaps in the annual story. The P2P intermediation segment moved from an operating loss of NIS 1.1 million in 2024 to an operating profit of NIS 3.9 million in 2025. On the surface, that looks like a sharp turnaround in a legacy engine. But the company's own explanation is different, lower payroll expense because activity declined, lower media spending than in the comparable period, and lower partner marketing commissions that move with the size of the book.

So this is not the profit of a reopened engine. It is the profit of an engine that contracted and cleaned up expenses. Gross Non-GAAP revenue in the segment also fell to NIS 53.4 million from NIS 57.1 million, and new originations fell to NIS 128 million from NIS 141 million. That is exactly the difference between a real improvement in economics and a temporary accounting improvement.

Where The Real Competitive Edge Sits

Blender does have real strengths. It already operates across more than 2,000 points of sale through the joint activity with Bank Hapoalim, it has diversified funding sources from banks, an institutional lender, and P2P investors, and it can originate credit digitally and quickly. In addition, the distribution agreements with Amdocs and Sapiens open interesting sales channels into financial institutions.

But a clear limit also needs to be drawn here. The company describes an underwriting system, anti fraud tools, and Big Data capabilities, yet it explicitly states that it has no data showing that the system predicts credit default better than other underwriting systems. So the competitive edge that exists today is an execution, integration, and distribution edge, not a credit model that has already proven itself as clearly superior.

There is another point that readers should not miss, the gap between management language and reported accounting. In the presentation and in the directors' report, the company also uses Non-GAAP gross revenue figures of NIS 80 million in Israel, which include, among other things, interest paid to investors on the P2P platform. In the financial statements themselves, continuing operations gross revenue is NIS 41.8 million. That is not a contradiction, but it is an important filter. Anyone reading Blender has to know when they are looking at a wider economic activity figure and when they are looking at the IFRS top line.

Cash Flow, Debt And Capital Structure

The Right Cash Lens Here Is All In Cash Flexibility

In Blender's case, sloppy cash framing can produce a picture that looks too good. When a company grows through externally funded credit origination, accounting profit on its own says very little about real flexibility. The better lens here is all in cash flexibility, meaning how much cash is actually left after real cash uses, not a normalized version that assumes growth is automatically discretionary.

On that basis, 2025 still does not look like a year in which the business funded itself. Operating cash flow was negative NIS 1.9 million, investing cash flow was negative NIS 2.6 million, and lease principal payments amounted to NIS 682 thousand. Cash rose by NIS 14.7 million, but only because financing cash flow was positive NIS 19.2 million, mainly thanks to the November 2025 equity raise.

This chart explains why a higher cash balance is not the same thing as cash independence here. In 2024 financing drained cash. In 2025 it added cash. So the cash position improved, but it improved through funding and capital decisions, not because the business already generates a consistent surplus of free cash.

Working capital also improved, from a deficit of NIS 13.9 million to a surplus of NIS 6.9 million. Loan related working capital rose to NIS 81.4 million from NIS 42.9 million. That is positive, but it mainly reflects a more comfortable balance sheet structure at year end, not necessarily a final answer to whether the model can keep growing without further capital support.

The Debt Structure Looks Reasonable, but It Still Depends On External Funding

At the end of 2025 the company had NIS 153.0 million of funding liabilities versus NIS 127.3 million a year earlier. Of that total, NIS 106.8 million was short term bank debt, NIS 12.0 million long term bank debt, NIS 30.6 million long term institutional debt, and NIS 3.6 million other lenders. Equity rose to NIS 54.6 million from NIS 37.9 million, mainly because of the raise.

Inside the BNPL engine the picture is even clearer. At the report date, credit facilities available to the segment were about NIS 212 million, of which about NIS 162 million was drawn. The Bank Hapoalim facility amounted to NIS 132 million including the supplemental loan, with NIS 98.5 million drawn at year end. There was also a NIS 30 million institutional facility that was fully drawn and another NIS 50 million bank line, of which NIS 20 million had been used.

From a covenant standpoint Blender was in compliance. Under the Bank Hapoalim agreement, the company had to keep tangible equity at at least 25% of the customer loan balance, line utilization at no more than 75%, and bad debt ratio at no more than 2.5%. Under the institutional agreement, it had to keep, among other things, company tangible equity of at least NIS 30 million, collections to debt service of at least 1.2, Blender Pay equity of at least NIS 7.5 million, and a revenue to interest plus credit loss ratio of at least 1.3. Under the other bank agreement, it had to keep tangible equity at 25% of assets and senior debt at no more than 75% of the customer loan balance. The company says it was compliant with all of them.

The message is two sided. On one hand, there is no visible covenant stress. On the other hand, BNPL growth is clearly built on external facilities. So any positive reading about "balance sheet strength" needs to stay precise: Blender strengthened equity, exited Europe, and expanded its lines, but it is still not at the point where growth can be separated from ongoing access to funding and capital.

Outlook

Before looking at 2026, four conclusions frame the story properly:

1. This is a proof year, not a breakout year. Second half operating profit and the European exit improve the starting point, but they do not yet prove a clean and profitable company without an active P2P engine.
2. P2P will free up headlines, but not necessarily friction. Even after new originations stop, the existing book still has to be serviced, collections and insurance still have to be managed, and the reputational and legal tail still remains.
3. BNPL is already working, but it still has to prove it can carry the full story on its own. Growth is strong, but the allowance ratio rose and the expansion depended on more funding and more equity.
4. The technology layer still lacks accounting proof. Distribution agreements, a service bureau, and MOUs are not enough. What changes the market reading is material revenue showing up in the accounts.

What Kind Of Year Is 2026?

The right definition for 2026 is a proof year. Not because the company has not moved, but because it has already changed direction and now needs to prove that the new direction is durable. Blender ended 2025 with a cleaner Israeli base, no major European drag, stronger equity, and second half operating profitability. But the new model has not yet gone through a full year without an active P2P contribution, and it has not yet shown that technology sales can become a material layer.

Management also signals this without writing it explicitly. On one side it highlights profitability improvement, BNPL growth, and technology partnerships. On the other side, it stopped P2P, raised equity to support growth, and announced a new service bureau right after year end. This is not the language of "we arrived." It is the language of moving from one model into another.

What Must Happen In The Next 2 To 4 Quarters

The first thing that must happen is continued BNPL growth without a deterioration in credit quality. The company disclosed that from January 1, 2026 until close to the publication date it had already originated about NIS 40 million of loans in Israel at an average nominal rate of around 11%. That suggests the pace has not stopped. The next question is whether that pace translates into cleaner operating profit, not only a bigger book.

The second thing is stabilization in the provision burden. At the end of 2025 the allowance ratio stood at 4.7%, and credit impaired loans reached NIS 12.7 million with NIS 5.0 million of allowance. If 2026 shows another sharp rise in the impaired layer alongside further growth, the market will read 2025 much less cleanly.

The third thing is genuine proof of technology monetization. There are now three relevant anchors, the Amdocs and Sapiens distribution agreements, and the service bureau signed in January 2026. Against that, we have already seen one development services MOU stop after a first stage. So the right question is not whether there is interest in the product. It is whether Blender can convert that interest into material and recurring accounting revenue.

The fourth thing is orderly management of the P2P runoff. The company estimates that post collection default rate will stand at about 3.5%, but that estimate depends on future collections and insurance proceeds. If those flows are delayed, if the fund remains depleted for a long time, or if the legal exposure around the activity expands, the P2P legacy will keep sitting on the story even after the activity stops.

What might the market miss at first glance? That the positive second half is real, but it is not identical to clean profitability. Part of the improvement came from cost reduction in a shrinking segment, part came from a structural capital reset, and part came from growth in an engine that still depends on external funding. What could surprise positively? If BNPL continues to show the same kind of growth pace, net margin, and actual credit loss performance without another fast equity raise. What could surprise negatively? If technology remains mostly a story while P2P keeps absorbing management attention and cash.

Risks

The P2P Legacy Still Sits On The Company

The first risk is that P2P stopped growing but did not stop carrying obligations. The security fund was depleted by the report publication date, and from January 2026 investors began receiving only partial payments. That creates a double risk, both operational risk around collections and insurance and trust risk with lenders and borrowers. In addition, in December 2025 a class action request was filed against Blender P2P Israel over allegedly excessive fees and fee changes during existing loans, and the company states that it cannot assess the chances of the claim at this stage.

BNPL Growth Depends On Funding, Covenants, And Regulation

The second risk is that the main growth engine, point of sale lending, is still highly sensitive to funding conditions. The company is currently in compliance with its covenants, but growth depends on banks and an institutional lender, not on a deep internal capital base. At the same time, in May 2025 the regulator published a draft circular for credit offered directly alongside a purchase or service transaction, with expanded disclosure, explicit borrower consent, cancellation mechanisms, and due diligence obligations on merchants. As of the report date the draft had not yet been implemented. That is not a certain hit, but it is a reminder that the area where Blender is building its future is also an area where regulation can change unit economics.

The Risk Of A Technology Narrative Running Ahead Of Numbers

The third risk is a gap between narrative and accounting. Distribution agreements, service bureau structures, development work, and technology partnerships are good raw material for a strong story. But as long as actual technology revenue is NIS 198 thousand, and as long as one MOU already stopped before a dedicated operating vehicle was formed, there is a clear risk that the market gives valuation credit to a story that still has not converted into execution.

Legal, Management, And Liquidity Frictions

The fourth risk is a cluster of smaller issues that still matters. In May 2025 another class action request was filed against Blender Pay, Blender P2P Israel, and ERN over credit extended for services that were allegedly not fully delivered, and there too the company says it cannot assess the outcome. The company also notes some dependence on founders and key executives. Above all, the low liquidity in the stock creates a practical actionability constraint even if the business thesis improves.

Conclusions

Blender ended 2025 in a better place than where it started. Israel grew, Europe was almost fully removed, expenses came down, and the second half already crossed into operating profit. But the main bottleneck did not disappear, it changed shape. Instead of asking whether the company can grow, the market now has to ask whether it can stay cleaner and more profitable after P2P moves into runoff and without leaning on a software engine that is still unproven.

Current thesis in one line: Blender exited 2025 as a sharper and more focused Israeli credit company, but the next phase depends on proving that BNPL can carry the story on its own while P2P retreats and technology still looks for real revenue.

What changed versus the 2024 reading is fairly clear. First, the European exit made the story simpler. Second, the Israeli lending segment is now big enough to carry most of the improvement in the accounts. Third, P2P moved from being a legacy growth engine into a source of friction and management attention. Fourth, technology remained an interesting strategic layer, but not yet a financially meaningful one.

The strongest counter thesis is that the skeptical reading may simply be too harsh. One can argue that this is exactly what a good reset looks like, selling non core activity, stopping an engine that no longer fits the environment, strengthening equity, scaling the strongest business line, and building a new SaaS layer in parallel. If that is true, then 2025 is not a transition year but the beginning of a higher quality platform.

What can change the market's interpretation in the short to medium term is not another technology slogan, but three hard indicators, growth pace in Israel, the behavior of provisions and actual credit losses, and proof that technology revenue starts showing up in the numbers. This matters because in Blender's case it is not enough to see growth. Investors also need to see who funds it, who absorbs the risk, and where the value for common shareholders is actually created.

In the next two to four quarters the thesis strengthens if BNPL keeps growing without a deterioration in credit quality, if the P2P runoff closes in an orderly way without a persistent financial and reputational tail, and if the new service bureau or one of the technology partnerships turns into visible revenue. It weakens if provisions keep rising, if the P2P fund remains effectively empty for an extended period, or if the technology layer remains only a promise quarter after quarter.