Follow-up to Buff: How Much of the Second-Half Improvement Was Operating, and How Much Was Accounting

The main article argued that Buff's second half looked better, but it deliberately left one question open: had the business actually started earning its way through operations, or did a large part of the jump come from accounting. That is the issue worth isolating. In a small company with a tight cash position, the gap between operating improvement and improvement created by revaluation, capitalization, or deferred expense is not a technical footnote. It is the gap between a business moving toward self-funding and a report that looks stronger than the economics underneath it.

The answer is not binary, but it is clear. There was real operating improvement in the second half. Revenue rose, gross profit jumped, and the loss from ordinary operations narrowed from 2.444 million dollars in the first half to just 610 thousand dollars in the second half. But the 330 thousand dollar profit from continuing operations did not come from operations alone. It also relied on 919 thousand dollars from revaluing investor warrants, 238 thousand dollars from discounting the founders compensation liability, and 579 thousand dollars of mobile development costs that were moved from R&D expense into an intangible asset.

In other words, the second half was better than the first, but the move from loss to profit still does not qualify as clean operating profitability. That is the real quality-of-earnings test Buff carries into 2026.

What Actually Improved

The operating side of the story should not get lost. Second-half revenue rose to 4.872 million dollars from 3.625 million dollars in the first half. Cost of revenue rose much less, from 1.975 million dollars to 2.172 million dollars, so gross profit jumped by 1.05 million dollars to 2.7 million dollars. That is the most important change in the report, because it says the campaign channel did not just add volume. It added more dollars above the operating cost line.

There was also genuine improvement below gross profit. Selling and marketing expense fell from 1.862 million dollars in the first half to 1.671 million dollars in the second half, while G&A was essentially flat at 904 thousand dollars in the first half versus 901 thousand dollars in the second. That is why the loss from ordinary operations narrowed by 1.834 million dollars. On a bridge between halves, about 69% of the improvement in continuing operations came from the operating lines before financing.

That matters, because without this operating improvement there would have been nothing for the accounting support to amplify. Buff did not manufacture a complete illusion here. It did improve the business in the second half. But the important question is not whether there was improvement. It is whether that improvement already proves normalized profitability. The numbers still say no.

Where Accounting Did Heavy Lifting

The finance line did much more than it first appears

The line separating a second half that was "better" from one that was "profitable" is finance. In the second half, the company recorded net finance income of 940 thousand dollars, versus only 105 thousand dollars in the first half. The half-year bridge shows that the main change came from three items: 919 thousand dollars from revaluing the liability for investor warrants, 238 thousand dollars from discounting the founders liability, and, on the other side, a 57 thousand dollar increase in FX expense.

That is the core point. The second-half profit from continuing operations did not come from Buff already turning operationally profitable. It came from operating losses narrowing, and then the finance line pushing the number above zero.

The largest effect was the warrants issued to investors in March 2025. As part of the private placement, Buff issued 505,050 warrants with a 19 shekel strike price and a 12-month life. By December 31, 2025, the share price used for measurement was only 10.12 shekels, and the fair value of those warrants had already fallen to a negligible amount, below 1 thousand dollars. The standalone valuation report points in exactly the same direction: 0.003 shekels per warrant, or roughly 1.3 thousand shekels in total. That is why the company recorded 1.213 million dollars of finance income in 2025 from revaluing the warrant liability, with 919 thousand dollars of that showing up in the second-half bridge.

Precision matters here. This is not a gimmick. It is valid accounting for a financial instrument classified as a liability and measured at fair value through profit and loss. But it is also not recurring operating earnings. The business did not sell more product because of it. A liability created in a financing round simply lost almost all of its value when the share price traded far below the strike, and the warrants later expired on March 3, 2026.

The founders liability helped the P&L, but it did not solve the problem

The second layer is the founders compensation liability. In 2025, the company recorded 245 thousand dollars of finance income from discounting that liability, alongside 111 thousand dollars of finance expense from imputed interest on the same balance. In the half-year bridge, 238 thousand dollars of the second-half improvement is attributed to that discounting item alone.

The interesting point here is two-sided. On the one hand, this was another accounting contribution to the profit line rather than an operating engine. On the other hand, the liability itself did not disappear. At year-end it stood at 1.213 million dollars, and it is part of the reason equity was negative 771 thousand dollars. The company explicitly says that without this liability, equity would have been positive 442 thousand dollars, and that it intends to examine alternatives for converting the liability in order to reduce the equity deficit.

So this is not a case of a one-off P&L benefit that leaves no trace behind. It is also a balance-sheet issue that still weighs on the company. The founders liability helped the second-half profit line, but it remains part of Buff's pressured capital structure.

Mobile Development Moved Through The Balance Sheet

The third accounting support, and probably the most important one for earnings quality, is the shift of mobile development through the balance sheet. Until June 30, 2025, self-developed costs were fully expensed. Starting in the second half of 2025, management concluded that the mobile-app project met the IAS 38 criteria for recognizing development costs as an intangible asset. As a result, 579 thousand dollars were capitalized in 2025.

The key number here is the half-to-half comparison. R&D expense fell from 1.328 million dollars in the first half to 738 thousand dollars in the second half, a decline of 590 thousand dollars. That almost matches the 579 thousand dollars capitalized during 2025. So even if one fully accepts management's conclusion that the accounting criteria were met, the decline in R&D should not be read as if it came entirely from a deeper economic improvement in the business.

This is not an accusation of aggressive accounting. It is simply a distinction between two types of improvement. If a cost that used to run through the income statement now moves through the balance sheet, reported earnings improve, but the business itself has not necessarily started generating cash or profit from that investment. In Buff's case that matters even more because mobile still did not contribute material revenue in 2025. So the capitalization does not yet prove that the new engine is already working. It mostly proves that the company believes the project has advanced far enough to recognize part of the investment as an asset.

This chart is, of course, an analytical exercise, not a reported figure. It does not replace the audited report, and it does not claim the company should have reported this number under IFRS. It does one simple thing: separate operating earnings from liability remeasurement gains and from an expense that was deferred onto the balance sheet.

So How Much Was Operating, and How Much Was Accounting

If the question is how much of the half-to-half improvement in continuing operations was real versus accounting-driven, the answer is measurable. Out of a 2.67 million dollar improvement, about 1.834 million dollars came from the operating lines, roughly 69%. Another 836 thousand dollars came from finance and tax, roughly 31%. That means the second half was not an accounting-only event. But it also means almost a third of the improvement came from outside the operating core.

If the question shifts from "how much improved versus the first half" to "is Buff already profitable," the answer becomes more demanding. Even after all the operating improvement, the company still posted a 610 thousand dollar loss from ordinary operations in the second half. The 330 thousand dollar profit from continuing operations only appeared after 940 thousand dollars of net finance income. And if one simply expenses the capitalized mobile development, without touching the warrant revaluation or the founders-liability discounting, second-half continuing operations move back to a 249 thousand dollar loss. If those two finance items are also stripped out, the second half becomes a roughly 1.406 million dollar continuing-operations loss.

That does not prove the company failed to improve. It proves the profit was still not clean.

Why This Matters For 2026

The reason this matters now is that part of the 2025 accounting support already cannot repeat in the same way. The investor warrants expired on March 3, 2026, so an effect like the 919 thousand dollar second-half gain will not recur. The founders liability is still on the balance sheet, but it cannot carry the profitability story on its own. And mobile capitalization will now face a tougher standard: if costs continue to move into the asset line, readers will need to see commercialization proof alongside it, not just relief in the R&D line.

That is why the correct test for the next reports becomes much simpler. Not whether the company again reports comprehensive profit, and not whether the finance line helps again, but whether the loss from ordinary operations keeps narrowing without warrant revaluations, and whether mobile starts generating revenue that justifies the asset being built on the balance sheet.

Conclusion

Buff's second half was better than its first, and the report gives that claim real support. But earnings quality was mixed. The operating core improved, mainly through campaigns and gross profit, yet it still did not cross into profitability on its own. What pushed the company to positive continuing-operations profit was a combination of unusual finance-line support and mobile-development capitalization.

For a reader who came out of the main article still holding this question, the takeaway is straightforward: the second-half improvement was not an illusion, but it was not proof of clean profitability either. In 2026, Buff will need to show that the numbers can hold up without revaluation support, and that mobile can move from the balance sheet into actual revenue.