Direct Finance: The Mortgage Funding Machine After the February 2026 Securitization

What This Follow-Up Is Isolating

The main article already established the big point: in 2025 Direct Finance's mortgage business was no longer just a growth engine. It was already carrying reported profit. This follow-up isolates the layer underneath that headline. Not whether the mortgage segment earns money, but how it funds growth, who gets paid first, and how much of that economics actually reaches Direct Finance's ordinary shareholders once Menora, securitization layers, and short-term debt repayment sit in the middle.

That is the right question now because February 2026 moved the story from theory to execution. On February 19 the company reported the completion of the mortgage securitization book-building. On February 23 the assignment and bond issuance were completed. On February 24 the proceeds were received by the mortgage subsidiary. This is no longer a future option. It is an operating funding channel.

Three sharp findings stand out from the filings:

• Menora is not just an external lender to the mortgage subsidiary. It also owns 15% of that subsidiary's economics, so the same structure that funds growth also shares away part of the profit.
• The January 1, 2026 Menora amendment is mainly about cheaper funding, more time, and a slightly looser covenant package, not about opening a brand-new pool of liquidity. At year-end 2025, Menora funding already stood at NIS 678.4 million out of an NIS 858.4 million facility, and near the report date it had climbed to NIS 831.2 million.
• The February 2026 securitization is a full true sale, but not a full exit from the economics of the pool. The mortgage company keeps servicing income, holds a subordinated layer, and the proceeds are used to repay short-term bank debt rather than dropping straight to shareholder free cash.

Menora Is Both the Equity and the Debt

At year-end 2025, the mortgage business already looked less like a plain lending book and more like a layered funding structure. As of December 31, 2025, funding outstanding at the mortgage subsidiary totaled about NIS 2.223 billion. The annual report breaks that down into three clearly visible layers: NIS 1.280 billion from banks, NIS 678.4 million from Menora, and NIS 264.0 million in Series A bonds at the mortgage subsidiary.

Against that debt layer sits the equity layer. Under the investment agreement with Menora, Direct Finance funds 85% of each capital call and Menora funds 15%, up to a total commitment of NIS 600 million. By December 31, 2025, the two parties had already invested about NIS 335 million in aggregate. So the mortgage business is indeed growing on top of equity, but it is equity that already embeds minority leakage.

That leads to the first key conclusion. Anyone reading the mortgage segment's profit as if it belongs entirely to Direct Finance is reading the wrong number. The mortgage segment earned NIS 73.4 million of net profit in 2025, but NIS 11.0 million of that sat in non-controlling interests, leaving NIS 62.4 million attributable to Direct Finance's shareholders. Menora helps scale the engine, and at the same time it already sits on the side that receives part of the output.

The covenant package shows why this funding relationship matters so much. Under the Menora agreement, the mortgage subsidiary ended 2025 with a tangible equity ratio of about 14.9%, against a required weighted equity ratio of about 14.17%. By contrast, average LTV stood at about 62.6% against a 75% ceiling, and the bucket of loans with LTV above 75% and up to 90% stood at about 8.8% against a 15% ceiling. In other words, the less comfortable constraint at year-end was not average collateral quality. It was the equity cushion.

The same table in the annual report gives another important read-through: near the report date, Menora funding had already risen to NIS 831.2 million, almost the full approved facility of NIS 858.4 million.

That point matters because it explains what the February 2026 securitization is really doing. It is not landing in an empty balance sheet. It is plugging into a machine that was already running close to one of its key funding limits.

January 1, 2026: Cheaper, Longer, Slightly Looser

The January 1, 2026 immediate report and the post-balance-sheet note in the annual report describe the same move from a cleaner angle: the Menora facility was extended through December 31, 2026, and the interest margins were reduced.

In the fixed indexed track, the spread fell to 1.75% to 2.25%, from 2.5% to 3.5%. In the prime track, the spread fell to 0.25% to 0.75%, from 0.5% to 0.9%. And for the NIS 180 million that was still undrawn on the signing date, the revised agreement defined a variable indexed structure with annual repricing.

The economic meaning is deeper than just "lower funding cost":

• Menora improved the pricing not only for future drawdowns, but also for loans already outstanding under the facility.
• The utilization period was extended by almost a full year, which gave the mortgage company time as well as price relief.
• The covenant wording was changed so that the bucket of end-loans with LTV above 80% and up to 90% is capped at 15% of the portfolio, instead of the prior test above 75% and up to 90%.

That last item looks technical, but it is not minor. It does not remove the 75% average LTV ceiling, and it does not remove the 2.5% annual specific damage cap. What it does is make the high-LTV bucket slightly less restrictive. For a mortgage platform operating with both first- and second-lien loans, that is a real operating flexibility improvement.

It is also important to be precise about what this amendment does not do. It does not solve the funding dependency problem. It does not change the fact that Menora is both a minority partner and a major lender. And it does not erase the fact that the line was already close to fully used near the report date. So the correct read is lower friction for an existing machine, not the creation of a new one.

February 2026: True Sale, But Not a Full Economic Exit

The January 25, 2026 immediate report was still a statement of intent: the mortgage subsidiary was examining a sale of a mortgage pool to a wholly owned issuing vehicle, alongside a private placement of two bond series and additional subordinated layers. By February 19 the deal had reached the pricing stage. The company reported the completion of the book-building for a mortgage securitization of a roughly NIS 400 million pool, with NIS 303.905 million par value of Series A bonds and NIS 64.465 million par value of Series B bonds. Total consideration for the senior layer was set at NIS 401.1 million, and the company said completion was expected to generate NIS 13.5 million to NIS 17.5 million of net profit attributable to shareholders.

On February 23 the assignment and issuance were completed, and on February 24 all conditions for the transfer of proceeds were met and the cash was received by the mortgage subsidiary. So the four stages were completed within one month: evaluation, pricing, closing, and receipt of proceeds.

The filings also show that this was not a first one-off structure, but a repeat of a framework already proven in July 2025. Comparing the two deals is the easiest way to see what really leaves the balance sheet and what remains inside the system:

This is the core insight of the continuation. The true sale is real. Both the July 2025 and February 2026 structures explicitly state a full, absolute, non-recourse assignment by way of sale. But that does not mean the mortgage company and the parent walk away from the economics of the pool. Quite the opposite:

• The mortgage company continues to provide servicing, collection, and trustee-like functions for fees.
• Part of the subordinated layer is still issued directly to the mortgage company.
• The proceeds are used to repay short-term bank debt, not to create distributable free cash.

So the securitization is first and foremost a funding-recycling machine. It moves a loan pool out, repays short-term debt, reopens borrowing capacity, and leaves the company with economics in servicing and lower-ranking layers. That is powerful. But it is not the same thing as a clean asset sale in the simple shareholder sense.

What Actually Reaches Shareholders

The right way to read the mortgage engine is through the shareholder layer, not through the segment profit alone.

In 2025 the mortgage segment produced NIS 73.4 million of net profit. That is a real number, and it is supported by a better financing spread of 4.08%, a larger average portfolio, and the start of servicing income. But to understand what belongs to Direct Finance shareholders, two filters matter.

The first filter is Menora as minority holder. NIS 11.0 million of profit went to non-controlling interests. That is not accounting noise. It is the direct result of the mortgage subsidiary's capital structure.

The second filter is the funding structure itself. In both Note 29 and Note 31, the company states explicitly that proceeds from assignments and securitizations were used to repay short-term bank debt. So even when a securitization creates reported profit, it simultaneously funds the next cycle and relieves an existing financing burden. The cash does not simply "stay" at shareholder level.

That is also why the cash flow statement needs careful handling here. In the liquidity section, the company itself explains that long-term debt raises are classified as financing cash flow, while same-day repayment of short-term debt is classified as operating cash flow. That accounting split makes operating cash flow look more negative than the economics of the machine. After neutralizing equity raises and short-term debt repayment tied to long-term funding, the company presents NIS 154.4 million of cash generated from operating activity in 2025, versus reported operating cash outflow of NIS 212.4 million.

That is why the right shareholder question is not "do mortgages earn accounting profit?" They do. The question is whether the full system preserves enough spread after funding cost, minority leakage, and repeated debt recycling to convert mortgage growth into value that ordinary shareholders can actually keep.

Conclusion

After the February 2026 securitization, Direct Finance's mortgage platform looks less like a mortgage book growing on balance sheet and more like a multi-layer funding machine. The parent and Menora supply equity. Menora and the banks fund the growth. Securitization moves mortgage pools out in order to repay short-term debt and reopen room for the next cycle. That is an efficient architecture, but it also means the key shareholder question is not just how much profit the segment reports, but how much of that economics survives the layers in between.

What changed after year-end is twofold. First, Menora reduced pricing, extended the line, and loosened one of the more sensitive covenant tests. Second, February 2026 proved that mortgage securitization is not just a strategic idea but an executable pipe that runs from pricing to assignment to cash receipt.

But the thesis still needs proof. The Menora facility was already close to full near the report date. The mortgage subsidiary's equity cushion was positive, but not especially wide. And securitization proceeds are not the same as free cash to shareholders, because they first repay short-term debt and feed the machine again. So over the next two to four quarters, three checkpoints matter most:

• Whether the securitization and assignment pipeline stays open after the February deal rather than remaining a one-off.
• Whether the financing spread holds or improves after the cheaper Menora pricing, without the company paying for that through a more aggressive LTV mix.
• Whether profit attributable to shareholders continues to grow in step with mortgage profitability, or whether minority interests and funding layers absorb an increasing share of the economics.