Follow-up to Turgeman: The new funding stack after the first bond issue

What This Follow-up Isolates

The main article made a simple point: even after genuine operating improvement, financing still drives the Turgeman read. This follow-up isolates one question only, what actually changed after Series A was issued on January 4, 2026.

The answer is sharp. The bond added a new layer of funding, but it did not replace the old hierarchy. The bank still sits on the assets, the project accounts, most of the hard collateral, the subsidiary-level covenants at Amirei Zichron, and a large portion of the controlling shareholder's personal guarantees. Bondholders got a new public instrument, but not a better seat in the structure.

That matters because it is easy to look at January 2026 and see a run of positive headlines: a first bond issue, listing approval, a Hadera project financing agreement, and repayment of the land loan. That reading misses the mechanics. A large part of what happened in January was debt refinancing into new pockets, and some of those new pockets are still short-dated, still bank-controlled, and still tied to collateral and personal support.

The Public Layer: New Money, Not a New Engine

On January 4, 2026, Turgeman issued Series A to the public for the first time, with NIS 123 million par outstanding. Principal is scheduled to amortize in three payments on December 31, 2027, 2028, and 2029, and the CPI-linked annual coupon was set at 3.93%. That is an important move because it opens access to the public market. But the way the deal cleared matters as much as the fact that it cleared.

Pricing did not signal comfort. The auction closed exactly at the maximum coupon that had been pre-set, 3.93%. The company received 47 orders for 130,000 units against an issuance cap of 123,000 units. That was enough demand to complete the deal, but not the kind of book that forces pricing lower. The heavy use of early commitments from classified investors, 120,000 units out of the 150,000 units on offer, also shows that the transaction was built carefully.

The second point is use of proceeds. The proceeds were not ring-fenced for a specific liability. The company presented a general use case, funding business activity in line with management decisions. In the supplemental notice, immediate expected net proceeds were estimated at about NIS 120.0 million after fees and expenses, and in the annual report's post-balance-sheet note, issuance costs were estimated at about 2.5% of gross proceeds. In other words, the bond added room, but it did not come with a closed deleveraging plan aimed at one specific wall of debt or one specific release of collateral.

The chart shows why the bond did not change the balance-sheet read on its own. Going into January, the company had NIS 990.958 million of short bank credit, another NIS 80 million of short non-bank credit, and only NIS 124.808 million of long bank debt. Against that stood just NIS 5.408 million of cash. A NIS 123 million bond is meaningful, but it does not on its own erase a starting point that short.

Why The Banks Still Sit Above The Story

The easiest mistake is to read the bond as a shift from bank funding to capital-markets funding. That is not the right read. The trust deed gives bondholders a framework, but not control over the assets.

On the one hand, the bond does have financial covenants. The equity-to-balance-sheet ratio cannot fall below 20%, and equity cannot fall below NIS 325 million. There is also an earlier coupon step-up mechanism that kicks in if the equity-to-balance-sheet ratio drops below 22% or if equity drops below NIS 400 million, with an extra 0.25% for each breach and 0.5% in aggregate. So there are warning steps before outright acceleration.

On the other hand, the bonds are unsecured, unrated, and the company has no obligation to seek a rating in the future. That is a major distinction. In that structure, the public creditor does not sit on a specific asset package. It sits on a general corporate promise.

The negative pledge is also narrower than it sounds. The company did commit not to create a general floating charge over all assets without bondholder approval. But the same trust deed explicitly says that, other than that restriction, the company may still create specific security over assets or over a specific asset, in any rank, without asking bondholders for permission. That point is central. Bondholders did not stop the bank from sitting first on a specific asset. They only got protection against a general floating charge over the entire company.

The less intuitive point is the solo-debt limiter. On paper, it sounds meaningful, total restricted solo debt cannot exceed 20% of the consolidated balance sheet. But the definition narrows the guardrail materially: it applies only to a specific kind of unsecured parent-level borrowing that is meant to be repaid out of residential-project surpluses. Public bonds are explicitly excluded from that definition. So this limiter does not constrain the bond that has already been issued, and it does not create a real ceiling on the public layer itself.

There is also a mechanism that ties bank risk back into the bond. If material debt of the company or of a consolidated subsidiary, above NIS 75 million, is accelerated and the demand is not removed within five business days, that can become an acceleration trigger for the bond as well. In other words, the public layer is not insulated from the banking system. It sits above it, but it can still be dragged by it.

January 2026 Did Not Remove The Land Loan, It Refinanced It

The January 19, 2026 financing agreement for Amiri Park Hadera makes the point even more clearly. On first read, it looks like very good news: a Sale Law guarantee facility of up to NIS 287 million, a cash credit facility of up to NIS 72 million, and an additional credit facility of NIS 119 million. But those three buckets need to be read correctly.

They are not cleanly additive. The NIS 72 million cash line overlaps the Sale Law guarantee envelope, so it is not a separate source of fresh money on top of the NIS 287 million. Beyond that, out of the NIS 119 million additional facility, NIS 89 million was already drawn at signing in order to repay part of the land loan, and only NIS 30 million remained for investments and improvements on the remaining land. In other words, a large part of what looked like new money was actually replacement money for old debt.

This is exactly what happened. The new agreement did not eliminate the NIS 144 million land loan through free cash. It replaced that loan with NIS 55 million from the project cash line and another NIS 89 million from the additional facility. That is an important change because it brings the project into a more formal project-finance framework. But it is not genuine deleveraging. It is refinancing.

More than that, one of the new pockets is still short. The NIS 119 million additional facility is in force only until January 7, 2027, and each loan drawn under it can run for only up to one month. Interest is prime plus 0.7%, paid monthly. So even after the January move, there is still a short bridge layer inside the structure. It may be more orderly than the original land loan, but it is still not long, clean money.

That leads to the other side of the agreement, the part that usually does not make the headline. The bank set clear conditions precedent through August 31, 2026, including at least NIS 38.2 million of equity invested in the project, registration of all security, a signed construction agreement, a building permit, and a pre-sale target of NIS 72 million. In addition, the borrower undertook not to take future external capital, and the bank capped contracts with non-linear payment schedules at 25% of project units. The additional facility also carries a maximum LTV test of 80%.

What does that mean economically? The bank is not just financing Hadera. It is also shaping how Hadera can be financed. It puts a ceiling on the use of non-linear contracts, keeps control over the collateral package, and does not leave room for another outside capital layer above it at the borrower level. So Hadera did not become a capital-markets project. It remains, first and foremost, a bank-led project.

Personal Guarantees Are Still Part Of The Architecture

The easiest thing to miss after the bond issue is the continuing role of Oded Turgeman's personal guarantees. On November 27, 2025, the company did approve a plan to work toward removing those guarantees after the bond issuance, but in the same breath it said there is no certainty that the guarantees will actually be cancelled, and that if removal of a specific guarantee requires a change in financing terms on the bank side, the move will be subject to the required approvals.

The numbers are not small. In the related-parties note, guaranteed amounts stood at about NIS 1.3 billion at the balance-sheet date and about NIS 1.33 billion at the signing date of the annual report, with about NIS 927 million of that linked to the financing of Mol HaHof Village. That is material because it shows that Turgeman's funding system still relies not only on assets and covenants, but also on the controller's personal signature.

This is not just a legal detail. It is a question about how institutional the capital structure really is. A real-estate company that has entered the public market but still leans on personal guarantees at that scale has not fully completed the move from an entrepreneurial structure to an institutional one. That may happen later. As of now, it has not happened.

The Mol HaHof financing makes that even clearer. After the February and December 2025 amendments, total credit around the asset stands at about NIS 906 million. The office-tower loan stands at NIS 102 million through the end of 2026, the mall loan stands at NIS 622 million and is mostly repaid at the end of 2026, and there is an additional NIS 100 million facility through the end of 2028 that was still unused as of December 31, 2025. Alongside that sits an unlimited personal guarantee by the controlling shareholder for the total cash credit, plus first-rank mortgages and assignments. That is the core of the story: the bank is not just a lender, it controls the hardest collateral layer.

Bottom Line

Series A was the right and necessary step. Without it, Turgeman would have remained almost entirely inside a banking and non-bank funding system that was too short. But it was a first step, not a finish line.

What actually happened was a broader funding stack, not a cleaner one. A public layer of NIS 123 million was added, but it was added without collateral and without a rating. At the same time, the Hadera land loan was not eliminated, it was recycled into new bank facilities, some of them short, hard, and collateralized. And above all of that, the controlling shareholder's personal guarantees are still alive inside the system.

So the real question for 2026 is not whether Turgeman managed to issue a bond. That question has already been answered positively. The real question is whether the company can turn this first issue into a true maturity extension: less short debt, less reliance on the controller's personal signature, and more funding that rests on asset and cash-flow quality rather than on continuing bridges.