Elspec in 2025: FX cut the profit, but measurement and cash are holding up

Company Overview

Elspec can look like a small niche electrical equipment company. That is only half true. Economically, the group runs on two very different engines: power quality improvement systems, which are more project driven and tied to customer CAPEX, and power quality measurement and network analysis, which is already the real profit engine. If you only read the 2025 net profit line, you miss that distinction and you miss the story.

The annual headline is weak: revenue fell 8% to NIS 92.5 million, operating profit fell 32% to NIS 8.3 million, and net profit was cut 51% to NIS 6.3 million. But the same year also built the counterpoint. In the second half, revenue still slipped slightly to NIS 48.0 million from NIS 49.1 million, yet operating profit improved to NIS 5.2 million from NIS 3.3 million and net profit rose to NIS 3.7 million from NIS 3.6 million. That matters because it suggests 2025 was not a clean operating collapse.

One number set should frame the whole read. About 95% of sales are exports, roughly 60% of revenue is dollar denominated and about 35% is euro denominated. Management estimates that FX moves cut 2025 revenue by roughly NIS 4-5 million and profit before tax by about NIS 5-6 million. FX was not background noise. It was one of the main reasons the consolidated picture looked worse than the underlying business mix.

The active bottleneck is not liquidity. The company ended 2025 with NIS 17.5 million of cash, NIS 6.3 million of short term investments, NIS 89.9 million of equity and only NIS 1.1 million of bank debt. The bottleneck is proof. Elspec is trying to move from defending the installed base to proving the next generation: the new G4, PQBI, medium and high voltage solutions, and the move into BESS and Microgrids. That makes 2026 look like a proof year, not a breakout year.

The quick economic map looks like this:

Segment profit is not shareholder earnings. Corporate overhead, part of selling expense, R&D, financing and tax are not allocated here. But the gap between NIS 34.9 million in measurement and NIS 8.5 million in systems still answers the key question: measurement is what carries the group operationally.

Events And Triggers

What actually hurt 2025

The first trigger: 2025 was much more of an FX year than a demand year. The dollar weakened 12.6% against the shekel during the year and the euro weakened 1.06%. On an average annual rate basis, the dollar weakened 6.7% and the euro 2.7%. Given Elspec's revenue mix, that translated the same underlying foreign currency activity into fewer shekels, while also pressuring margins because inventory bought at higher exchange rates was sold when proceeds translated at lower rates.

The second trigger: the systems segment entered 2025 against a softer end market backdrop. The company explicitly points to weaker wind energy and European auto markets, and those hit the more project based engine. That is why systems revenue fell much more sharply than measurement.

The third trigger: on April 2, 2025, the U.S. tariff program was announced. The direct effect during the reporting period was described as not material, but the fact that the issue is prominent in the filing matters by itself. The U.S. subsidiary imports most of its products from Israel, so even if tariffs did not break 2025, they add uncertainty around pricing and the supply chain.

The fourth trigger: the second half already looked better. Revenue did not recover, but profitability did. Gross profit for the half rose to NIS 25.3 million from NIS 25.0 million despite slightly lower sales. That came from a better mix and tighter cost control. Analytically, this may be the most important point in the whole report: late 2025 looked better than the annual headline.

Management change, but not a capital structure event

Oren Harari became CEO on January 1, 2025, while Yoram Harari stopped serving as CEO on January 12, 2025. On April 1, 2025, Netali Jersey Bachar became CFO, while Ronit Harari moved into a Finance Director role at 60% employment. This was a real transition year in management terms.

Still, the filing does not point to a financial shock from that transition. The NIS 1.04 million payable to the controlling shareholder at year-end reflected adaptation grant and vacation payout obligations recognized in prior years and paid in early 2026. There is a signal here of a family-controlled company with an active controlling layer, but not of a 2025 balance sheet hit.

What the market is likely to measure next

The near term trigger is not just whether new products get launched, but whether they convert into orders. Order backlog rose from NIS 14.0 million at year-end 2025 to NIS 16.8 million by March 22, 2026, yet most of that increase came from systems, up from NIS 8.2 million to NIS 11.8 million. Measurement backlog fell from NIS 5.8 million to NIS 5.0 million over the same period. That is a sharp point: the short term rebound still leans more on the older engine than on the new measurement story the market wants to underwrite.

Efficiency, Profitability And Competition

Which engine is really funding the company

The measurement and network analysis segment was down only 3% in 2025 to NIS 51.7 million, while the power quality systems segment fell 14% to NIS 40.8 million. The gap matters even more in segment profit: NIS 34.9 million in measurement versus NIS 8.5 million in systems. The gap was already large in 2024, but 2025 made it more central because systems weakened just when the group needed another engine to carry profitability.

This is where the distinction between accounting categories matters. Segment profit is not EBITDA, free cash flow or shareholder earnings. But it is a useful answer to a simpler question: which business funds the fixed cost base? In Elspec's case, the answer is clearly the measurement business.

The mix improved, but its quality still needs testing

The second half of 2025 already showed a better mix. Gross profit was almost unchanged despite lower revenue, and operating margin improved to 11% from 7% in the comparable half. That says the business model can still produce respectable profit when measurement carries a bigger weight and expense discipline holds.

But that is not the whole picture. In systems, Equalizer revenue fell roughly 21% to NIS 31.9 million, while Activar rose to NIS 8.9 million. That shift helped the mix somewhat, but it did not change the basic fact that the more project based engine weakened materially. So the better second half should not be read as if the systems issue has already been solved.

The moat is real, but less clean than it used to be

One of the more important non-obvious points in the filing is that the historical electrical signal compression patent expired in April 2024. Management argues that the main barrier is now not legal protection but know-how, accumulated applications and the ability to turn large data volumes into useful information for customers. That is a reasonable argument, but it changes the nature of the moat.

Once a patent expires, the moat shifts from legal protection to execution protection. That is not the same thing. It does not mean the advantage disappears. It does mean Elspec has to keep renewing the product, the software layer and the customer value proposition if it wants to defend high measurement margins over time.

The systems side makes that even clearer. The company itself says Equalizer and Activar still rely on a relatively old controller, which may limit the ability to meet new market requirements. So the legacy engine still generates revenue, but it also needs a refresh if Elspec wants to preserve competitiveness and pricing power.

Cash Flow, Debt And Capital Structure

The all-in cash picture actually improved

This is where the framing matters. The more useful lens here is all-in cash flexibility, meaning how much cash was left after actual cash uses during the year. On that basis, 2025 looked much better than the earnings line.

Operating cash flow rose to NIS 13.4 million from NIS 9.6 million in 2024. After NIS 3.3 million of fixed asset investment, NIS 1.8 million of capitalized development, NIS 1.9 million of lease cash payments and NIS 0.3 million of bank debt repayment, cash still increased by NIS 6.2 million to NIS 17.5 million. That is not the picture of a balance sheet under strain.

It is also important to ask where that came from. Working capital helped. Receivables declined by NIS 1.6 million and inventory declined by NIS 3.25 million in cash flow terms. But there is a mild yellow flag here: the company also says it typically holds around nine months of raw material inventory, and in 2025 it increased measurement inventory for certain models because of the war and split storage across multiple sites. So part of the working capital improvement is real, but it does not erase the fact that inventory policy remains fairly heavy.

This is not a leverage story

The balance sheet is a clear strength. Equity rose to NIS 89.9 million, equal to 79% of total assets. The current ratio improved to 4.39 and the quick ratio to 2.30. Total liquidity, cash plus short term investments, stood at NIS 23.8 million. Against that, the group had only NIS 1.066 million of bank debt, taken at the Portuguese subsidiary, plus NIS 5.719 million of lease liabilities by payment schedule.

In its current capital structure, 2026 does not read like a survival year. It reads like a year that needs to justify the strategic story. That is an important distinction. In small caps, the first question is often whether the company will need fresh capital. This filing does not point to that kind of pressure.

Even this strength is not cost free

Still, Elspec is not a low CAPEX software company. Cost of sales included NIS 4.2 million of depreciation in 2025, capitalized development was NIS 1.829 million, and the company expects to spend about NIS 12.7 million on R&D over the next 12 months from the report date. So it would be a mistake to translate strong liquidity into a story of surplus cash with no investment needs.

The NIS 1.5 million dividend approved on March 26, 2026 is a positive signal. It suggests the board is comfortable enough with liquidity even after a weak year. But the size is still modest. This does not look like a company that has run out of productive uses for cash.

Outlook

Finding one: 2026 looks like a proof year for new products, not a harvest year.

Finding two: the rebound in backlog through March 2026 leaned mainly on systems, while the measurement engine that underpins the future narrative did not show the same backlog acceleration.

Finding three: the new G4 already has accounting support, but it still lacks commercial proof.

Finding four: Elspec improved costs in sales and R&D, but part of that flexibility came from moving toward contractors and external specialists.

Finding five: if FX stabilizes, 2025 may later look like a transition year hit mainly by external factors. If FX stays hostile, 2026 can still look weak in reported numbers even if the operating business improves.

The real test is commercialization

The company expects the official launch of the new G4 by the end of 2026. It also discusses a 2026 release path for G4K and PQBI, and further development of the energy management and BESS platform. These are not side projects. They are the core of the forward story.

But created value is not yet accessible value. At year-end 2025, the carrying value of the new G4 stood at NIS 5.873 million. The company ran an impairment review with an external appraiser, which produced a recoverable amount of NIS 20.372 million using a 14 year cash flow model and a 21.1% discount rate. That supports the accounting treatment. It does not prove commercial traction.

That is the key distinction for 2026. The market does not need to ask whether the asset looks valuable on paper. It needs to ask whether the product sells, at what pace, to whom, and at what margin.

What kind of year is 2026

The filing does not use this label, but the picture is fairly clear: 2026 looks like a proof year with elements of a bridge year. There are three parallel tasks:

1. Prove that the next generation measurement products are actually reaching the market.
2. Keep the existing measurement engine highly profitable while widening the product stack.
3. Show that systems can recover from 2025 weakness without relying only on friendlier FX.

That leaves Elspec between two worlds. On one hand, it already has a very profitable measurement engine. On the other, much of the upgrade narrative, new G4, PQBI, BESS, still sits more in investment and development than in mature sales.

What must happen for the read to improve

The first step is commercial. Elspec needs to show actual market release and early shipments for the next measurement generation, not just ongoing development. The second step is operational. The systems backlog that improved by March 22, 2026 has to convert into revenue without another hit to margins. The third step is external. The company needs at least reasonable FX stability so that reported numbers stop wiping out underlying improvement.

That matters because most of the constructive case on Elspec rests on the idea that measurement can become larger, not simply stay profitable. If 2026 ends without commercialization proof, the new G4 and PQBI may remain another year of potential value instead of accessible value.

Risks

FX remains the first risk

The company is not solving currency exposure with a full, systematic hedge. It points to price adjustments where possible, immaterial point hedges and a natural hedge from raw material and subcontractor costs in dollars and euros. That is better than nothing, but it does not neutralize the problem.

The sensitivity analysis shows how material this is. A 10% move in the dollar changes profit before tax by NIS 1.666 million, and a 10% move in the euro changes it by NIS 0.896 million. In 2025 this was not a theoretical sensitivity. It was a real operating issue.

Geopolitics and supply chain friction are not abstract

The company explicitly says the war and the closure of Israeli airspace delayed inbound goods, prevented timely completion of systems and raised shipping costs. After the reporting date, on February 28, 2026, it already describes a renewed emergency environment that it says it cannot yet fully quantify in future operating terms.

This is not just macro framing. A company that exports most of its sales, manufactures mainly in Israel and carries a long inventory policy is highly exposed to these disruptions. The higher measurement inventory held in 2025 was meant to improve resilience, but it also raises the risk of heavier working capital and slower inventory turns if commercialization disappoints.

The new technology still needs proof

The new G4 had not generated sales by the balance sheet date, even though the company has already received the full innovation grant of NIS 919 thousand for it and expects to pay 3% royalties on future sales. Economically, that is a sensitive point: the development asset is already capitalized, the valuation work is already done, but the commercial proof point is not there yet.

Add to that the patent expiry and the company's own comment that Equalizer and Activar rely on a relatively old controller, and the risk becomes clearer. Until the new generation is commercially proven, Elspec is relying partly on deep know-how and partly on a legacy product base that still needs renewal.

The governance signal is not dramatic, but it exists

Elspec remains a company with an active controlling family layer. At year-end 2025 the controlling shareholder had a NIS 1.04 million payable balance, and Ronit Harari, the controlling shareholder's spouse, serves as Finance Director. There is no sign here of an acute governance event, but there is a reminder that this remains a family-shaped small cap rather than a fully institutionalized operating model.

Short Interest

Short interest was only 0.10% of float on March 27, 2026, with SIR at 0.54. The local high during 2026 reached just 0.20% of float and 2.01 days to cover on January 30, 2026, which is still very low in absolute terms. Sector averages are lower, but the broader point is clear: there is no meaningful short crowd pressing a strong bearish view against this story.

In practical terms, the market may be cautious on Elspec, but that caution is not showing up through a heavy short setup. That fits a company where the debate is more about proof, timing and commercialization quality than about balance sheet distress.

---

Conclusions

Elspec ended 2025 with a report that looks weak, but with a business that looks less weak than the headline implies. What supports the thesis today is a strong measurement engine, a solid cash position and a clearly better second half operating profile. What blocks a cleaner thesis is the need to prove that the next product generation can move from development into real sales, while FX remains a meaningful drag.

In the short to medium term, the market is likely to care less about another product story and more about three concrete tests: backlog conversion, actual rollout of the new measurement products, and whether the next few quarters look more like the second half of 2025 than like the full-year summary.

Current thesis: Elspec already owns a better measurement business than the annual headline suggests, but 2026 still needs to prove that the next generation can move from development into commercialization.

What really changed in 2025 is that mix mattered more than topline growth. Measurement stayed highly profitable, systems weakened, FX distorted the consolidated picture, and the balance sheet actually improved.

Counter-thesis: the constructive read may prove too forgiving if 2025 was not mainly an FX year but an early sign of deeper structural weakness in systems, stronger competition after the patent expiry, and slower than expected commercialization of the new products.

What could change market interpretation over the next few quarters is not another development narrative, but measurable evidence: orders, shipments, preserved measurement margins and continued cash discipline.

This matters because if Elspec can connect the current strength of measurement with real commercialization of the next generation, it can come out of 2025 as a higher quality company than the net profit line implies. If it cannot, it remains a profitable but more volatile niche manufacturer with too much exposure to FX and customer CAPEX cycles.

For the thesis to strengthen over the next 2-4 quarters, the company needs to show real commercial rollout of the next generation, convert backlog into revenue and keep the balance sheet strong while still funding R&D. What would weaken the read is another commercialization delay, renewed systems erosion and further FX damage without pricing recovery.