BRNX — H1 2026 Financial Report Analysis

BrenX (formerly Brenmiller Energy) recorded no revenue in the first half of 2026, compared with $387 thousand a year earlier. Its net loss narrowed 18% to $6.09 million, mostly because of cost cuts in research and a currency swing. The Israeli company builds bGen thermal energy storage systems, which turn electricity into heat, store it in crushed rock and release it later as industrial steam. Its two first Israeli projects (Tempo Beverages and Wolfson Medical Center) are still being built or commissioned, so nothing qualified as revenue in the period. The bigger story is the capital structure. The company kept running by selling convertible preferred shares to a single investor, Alpha Capital, at falling prices. The ordinary share count went from 23,862 at the end of 2025 to 353,223 at June 30, and to about 1.95 million by mid-September, even after three reverse share splits in 2026.

At a glance

• $0 revenue (H1 2025: $387 thousand). The only revenue a year ago came from one European project (Enel). No project reached a revenue milestone this half.
• $5.50 million of cash burned by operations in six months against $5.69 million of cash at June 30. Without new funding, that is roughly six months of runway. The company says this raises "substantial doubt" about its ability to keep operating (a "going concern" warning).
• Loss per share of $125.90, against $796.45 a year earlier. The smaller figure does not mean the business improved by 84%. The weighted share count rose about 12.6-fold (from 9,359 to 117,661, on a post-split basis), and the loss charged to ordinary shareholders actually doubled to $14.8 million (explained below).

Results for the six months to June 30, 2026

There is no operating margin figure because there was no revenue to measure it against. Share counts and per-share figures are as the company restated them after reverse splits of 7-for-1 (January 2026), 5-for-1 (April 2026) and 6-for-1 (August 2026). A reverse split merges shares (here 210 old shares became one), so the share count falls and the price per share rises without changing anything about the business.

What drove the operating loss

The operating loss fell by $605 thousand, and almost all of the improvement came from lower spending:

• Research and development fell 41%, from $2.41 million to $1.43 million. The company describes the cut as "broad-based" and part of "cost optimization and operational restructuring": salaries down $294 thousand, materials down $251 thousand, depreciation and other down $293 thousand. Part of the depreciation-and-other drop is an H1 2025 one-off, a repayment to the Israel Innovation Authority that did not recur.
• Cost of revenues fell 27%, to $1.35 million, even though there was no revenue to set it against. There are two reasons. First, the $426 thousand of subcontractor costs on the Enel project last year did not recur. Second, the write-down of work in progress was smaller ($473 thousand against $636 thousand).
• Overheads rose. General and administrative costs rose 17% to $2.43 million and selling and marketing rose 23% to $765 thousand. Most of the G&A increase was salaries ($513 thousand more), including $340 thousand more share-based pay (staff compensation paid in shares or options rather than cash).

What the headline numbers hide

1. More than half of the smaller net loss came from currency, not operations. The net loss shrank by $1.37 million. Of that, $605 thousand came from the operating line. The other main driver was the financial line, which swung from $837 thousand of expense to $4 thousand of income. Most of that swing is foreign exchange: a $104 thousand gain this half against a $644 thousand loss a year earlier. That gives about $748 thousand from currency movements, which can reverse in any period.

2. The loss that counts for shareholders doubled. The income statement shows a $6.09 million net loss. Loss per share, however, is calculated on $14.81 million charged to ordinary shareholders. The extra comes from an 8% dividend on the preferred shares ($149 thousand) and an $8.58 million "deemed dividend". The preferred shares and warrants (rights to buy shares at a set price) sold to Alpha carry anti-dilution "ratchet" terms. Each time the company raised money at a lower price, the conversion price of all earlier preferred shares was cut to match: it ended June at $10.02 per share, compared with $76.23 on the February tranche. The deemed dividend is the accounting value of that transfer from ordinary shareholders to the preferred holder ($6.74 million) and warrant holders ($1.84 million). It does not cost cash, but it is real value moving away from ordinary shareholders.

3. The share count is exploding. At June 30, holders of preferred shares, options and in-the-money instruments could claim another 425,047 ordinary shares, more than the 353,223 then outstanding. The ratchet kept working after the period ended. Fundings in July, August and September pushed the conversion price down to $5.505, then $3.12, then $2.616 (September 17). By that date the company expected 1,946,972 ordinary shares outstanding. That is roughly 5.5 times the June 30 count and 82 times the post-split count at the end of 2025.

4. Cash rose only because of financing. Operations used $5.50 million and financing brought in $6.45 million: $5.5 million gross from Alpha plus $1.08 million net from at-the-market share sales. That is why cash rose from $4.91 million at year-end to $5.69 million. Operating cash burn was a little below the net loss because $715 thousand of share-based pay and $271 thousand of depreciation are non-cash. A $528 thousand build-up of project inventory pulled the other way.

5. The first two projects are being built for less than they cost. Inventory rose from $1.58 million to $2.11 million and is mostly work in progress on the Tempo and Wolfson systems. The notes say both units "have been written down to net realizable value." In other words, the company expects to recover less than it is spending on them, so it books the gap as a cost while building. Under the Baran Energy agreement, Baran pays about $2.9 million in milestone payments and takes ownership of the systems on final commissioning. At June 30, $838 thousand of those advances had been received, and another $833 thousand came after the period.

After the period: debt settlement and a new strategy

• European Investment Bank (EIB) loan. At June 30 the company owed $4.81 million on this loan, and a $1.72 million payment due July 28 was waived while the two sides negotiated. On September 15 BrenX announced a settlement. It pays an initial €1.1 million, and the facility is then cancelled and its security released. Management's own illustration, applied to the June balance sheet, shows liabilities falling from $8.9 million to $4.1 million, equity rising from $4.6 million to $8.2 million, and cash falling by about $1.25 million. Management expects an estimated $3.6 million gain. One condition matters: if a change-of-control or a qualifying sale of assets or IP is signed within a year, BrenX owes a further €2.9 million.
• Hungary. On July 2 the company bought A.R.D. Energy Kft. for about $1.1 million. A.R.D. owns two licensed solar facilities that sell power to the Hungarian grid. On August 28 BrenX signed to buy adjacent industrial land, aiming at a site with up to 20 MW of solar, 6 MWh of batteries and 12.5 MWh of thermal storage.
• Rebrand. The company renamed itself BrenX, and its ticker changed to BRNX. It is shifting from selling storage equipment to owning and operating energy assets.

Takeaway: Per-share figures and the narrower net loss make the half look better than it was. Revenue was zero, the business still burns about $0.9 million a month, and it survives by issuing preferred shares whose conversion price has fallen from $76.23 to $2.616 in seven months. Ordinary shareholders bear most of that cost: $8.58 million of deemed dividends in the half and a share count up roughly 82-fold since year-end.

Outlook

Management did not give revenue or earnings guidance. It named four priorities: finish commissioning Tempo, continue construction at Wolfson, develop the Hungary platform, and "pursue financing structures" for current and future projects. It says it expects to "continue to incur losses and negative cash flows from operations in the near term."

Our read: the second half should look better on paper for three reasons. The EIB gain of about $3.6 million is a one-off and not an operating result. The Hungarian solar plant should bring its first small, recurring revenue. And Tempo may reach commissioning, which is when Baran takes ownership. None of that changes the core problem. Monthly cash burn is close to $0.9 million, and the only reliable funding so far has been Alpha tranches on terms that dilute existing holders at an accelerating rate. Things to watch in the full-year 20-F:

1. Whether Tempo and Wolfson produce recognised revenue, and at what margin after the write-downs.
2. Whether the EIB settlement closes without triggering the €2.9 million contingent payment.
3. Whether the going-concern warning is lifted.
4. Whether the conversion price keeps falling.

First BrenX report on this site; there is no earlier analysis to check against. Source: the company's unaudited condensed consolidated financial statements for the six months ended June 30, 2026 (Exhibit 99.2), MD&A (Exhibit 99.3) and press release (Exhibit 99.1), furnished on Form 6-K on September 9, 2026, plus its 6-K filings of September 15 and 17, 2026. All figures in US dollars.