Polygon: What Is Hadera Really Worth After All the Haircuts?

What This Follow-Up Is Isolating

The main article already made the bigger point: Hadera is the center of gravity in Polygon’s value, and the real 2026 question is no longer whether the asset works but whether management can turn that value into something accessible to shareholders. This follow-up isolates one narrower issue: what is already embedded in Hadera’s disclosed fair value, and where readers can accidentally double count the same haircut.

That matters because the headline number, NIS 320.6 million, can look like a gross appraisal. It is not. This is not a clean starting point before adjustments. The disclosed fair value comes after a direct vitrines deduction, after another 5% global haircut, and only then includes a separately capitalized management-surplus component. So the real argument is not whether the appraiser forgot the problem. The real argument is whether, even after all of those cuts, the appraisal is still too generous.

There is also an accounting layer here, not just an appraisal layer. Hadera is carried on the balance sheet at NIS 75.5 million under the cost model, while the disclosed fair value is NIS 320.6 million. That NIS 245.1 million gap sits at the core of Polygon’s value gap. But before deciding whether that spread is real or deserves another discount, the first step is to unpack the appraisal math itself.

The key point in the chart is simple: Hadera is not disclosed at NIS 342.1 million. Before the reported NIS 320.6 million even appears, NIS 33.25 million of explicit deductions have already been taken. Only after that does the appraisal add back NIS 11.8 million of management surplus. Put differently, the disclosed fair value still sits NIS 21.5 million below the gross income-capitalization output.

Where The Value Is Actually Cut

The first layer is a direct NIS 16.995 million deduction for the vitrines issue. This is not a narrow estimate for replacing broken glass. The appraisal says explicitly that the amount is based on the company’s experts, themselves relying on the court experts’ work, on the highest cost alternative they presented. It also includes direct and indirect costs such as tenant compensation, guarding, supervision, and a 12% contingency allowance.

That matters for another reason. The same appraisal notes a mediation agreement under which the company is entitled to about NIS 5.3 million for direct and indirect damages. Even so, the valuation still keeps a roughly NIS 16.5 million to NIS 17.0 million deduction in place. In other words, the appraisal does not behave as if the existence of compensation means the issue has already gone away. It still treats the repair burden as a live haircut.

The second layer is an additional 5% global haircut, or NIS 16.255 million. This is no longer just engineering cost. It is the market-transaction haircut. The appraiser says explicitly that a potential buyer would have to be told about the defect, and that the uncertainty around its cost and implications would reduce the price even beyond the direct repair bill. That is a critical distinction. The valuation is already assuming not only physical remediation, but also a market discount for the uncertainty itself.

This is exactly where double counting becomes easy. If someone takes NIS 320.6 million and then says the valuation still needs “the vitrines cut plus another 5% for uncertainty,” they are subtracting the same risk twice. The 5% haircut was not applied to NIS 342.1 million. It was applied to NIS 325.1 million after the direct deduction. That is the difference between challenging an assumption and cutting the same issue twice.

There is a third, smaller item that is still useful because it clarifies what should and should not be cut again. In December 2025 the company received a municipal demand related to open public spaces. The company had already booked the liability, about NIS 2.57 million, and the appraiser says that because this is an owner liability it did not need to be deducted from value. The lesson is important: not every liability tied to the property automatically turns into another valuation haircut if it is already sitting elsewhere in the accounting structure.

What Is Already Embedded, And What People Add By Mistake

The first mistake is to think NIS 320.6 million is a “before problems” number. It is not. Economically, it is a number after two haircut layers. The direct vitrines deduction of about NIS 17.0 million is already almost 5% of the gross pre-deduction value. Together with the global haircut, total explicit deductions reach NIS 33.25 million, nearly 9.7% of the gross value of the center.

The second mistake runs in the other direction and forgets that NIS 320.6 million already includes an add-back. The appraiser states that capitalizing the management surplus for roughly 10 years at a 9.5% discount rate produces a present value of about NIS 11.7 million. That is not a side detail. It is exactly the gap between the ownership-rights value after the deductions, NIS 308.846 million, and the fair value disclosed in the note, NIS 320.610 million.

The implication is straightforward. If someone takes NIS 320.6 million and then adds the management surplus on top, they are counting the same component twice. If someone takes NIS 320.6 million and then subtracts another NIS 17.0 million plus another 5%, they are cutting the same risk twice. In both directions, the right question is not “let’s be a bit more conservative” in the abstract. The right question is which specific component you disagree with.

This is also where economic value and accessible value diverge. Even after all the cuts, Hadera still sits at a fair value of NIS 320.6 million versus a carrying value of NIS 75.5 million. Even if you remove the management surplus entirely and stay only with the NIS 308.8 million ownership-rights value, the gap to book value remains huge. So the debate is not whether there is a gap. The debate is how much of that gap a reader is willing to accept, and which additional haircut they can actually justify.

Where The Real Argument Starts

Once double counting is stripped out, three more serious questions remain.

The first is the cap rate. The appraisal uses a weighted 7.37% cap rate, with 7.26% on occupied area and 7.94% on vacant area. That does not read like a dream input. The same appraisal cites the government appraiser’s survey, where the retail yield was 7.0% in 2024, and also lists comparison transactions broadly around 6.7% to 7.4%, alongside a Tel Aviv office transaction at about 6.4%. You can still argue that Hadera deserves a higher risk premium, but it is hard to say the appraisal picked a cap rate detached from market evidence.

The second is how aggressive the representative income is. Here too, the numbers are fairly restrained. Actual average occupancy in 2025 was 98.6%, but the representative occupancy used in the valuation is 97.7% for the complex and 98.5% for the ground floor. Actual average rent was NIS 95 per square meter per month, while the representative rent in the valuation is NIS 1,128 per square meter per year, or roughly NIS 94 per month. This is not a valuation built on a sharp step-up that has not happened yet.

The third is how much weight the management surplus deserves. Here there is more room for a real debate. The appraiser links it to the photovoltaic system and the expected reduction in electricity expense, then capitalizes it for 10 years at 9.5%. It is a relatively small piece, NIS 11.8 million, about 3.7% of the disclosed fair value, but it is still a distinct component. If an investor thinks that surplus is less durable than the appraisal assumes, that is the place to argue directly. It is a much more serious argument than automatically asking for another general haircut.

The sensitivity analysis reinforces the point. A 0.5% increase in the cap rate cuts value by NIS 15.442 million. A 5% decline in rent cuts value by NIS 16.316 million. A 2.5% occupancy decline cuts value by NIS 7.721 million. So the real fight over this valuation is less about a modest occupancy slip and much more about whether the cap rate is demanding enough and whether rent durability and the management surplus are truly solid.

One more point is easy to miss: Hadera alone represents about 92.7% of Polygon’s total disclosed fair value for investment property, NIS 320.6 million out of NIS 345.9 million. So this is not a side argument about one asset in a broad portfolio. It is close to the entire value debate in the company.

Bottom Line

The central message of this continuation is simple: NIS 320.6 million is not a before-haircut number. It is an after-haircut number. It already includes the direct vitrines repair burden, an additional 5% market haircut for uncertainty, and only then a capitalized management-surplus component.

That means the genuinely conservative read is not “let’s subtract again what has already been subtracted.” The genuinely conservative read is to ask whether 7.37% is a tough enough cap rate for this asset, whether NIS 11.8 million of management surplus deserves full capitalization, and whether representative rents can stay intact against local competition. Those are legitimate arguments. Double counting the deductions, or double counting the management surplus on the upside, is no longer an economic debate. It is just arithmetic error.

That takes the reader back to the main article. If Hadera is the core of Polygon’s value, then understanding this appraisal correctly is what determines whether the company’s discount is anchored or misread. Even after all the haircuts, the fair value still stands at more than 4 times carrying value. So the real question is not whether the problem disappeared. The real question is whether, after the problem has already been priced twice, once in repair cost and once in a market haircut, the investor still wants to demand a third one.