Delek Rehev 2025: Cash Cleaned Up, but the Import Engine Is Still Losing Altitude

Getting to Know the Company

Delek Rehev is no longer just a Mazda importer with a few side engines. The more accurate way to read it today is as a company with three very different layers: a vehicle core that includes import, distribution, parts, and service; an Eurodrive layer that entered in mid-2025 with operating leasing, zero-km vehicle sales, and two-wheel activity; and a holding layer through a 50.07% stake in Veridis, which pulls environmental services, energy, desalination, cardboard, and investment property into the numbers. Market value in early April 2026 stood at about NIS 1.96 billion. That matters, because the 2025 numbers look very different if the reader thinks this is still a pure-play vehicle importer.

The problem is that it is easy to fall into two wrong readings. A reader who looks only at the bottom line sees a net loss of NIS 21.2 million and may think the year broke down. A reader who looks only at cash flow sees NIS 1.176 billion from operating activity and may think the engine cleaned itself up. Neither reading is good enough. The loss was heavily affected by a roughly NIS 242 million write-down in the Hailo investment inside fair-value securities, but cash flow was also heavily affected by an inventory reduction of about NIS 798 million and another large working-capital release. In other words, both the loss and the cash flow tell the truth, but each tells only part of it.

What is working now? Vehicle inventory dropped sharply, year-end cash rose to NIS 179.6 million, the group's share in profits of associates rose to NIS 80.4 million, and investment property delivered a NIS 259.6 million fair-value gain. What is still not clean? Vehicle deliveries fell to 13,650 from 18,592, gross profit fell to NIS 782.4 million from NIS 1.001 billion, gross margin declined to 13% from 16%, and the fourth quarter ended with operating profit of only NIS 7.8 million even before Hailo hit the bottom line through finance items.

That is the core of the story. 2025 was not a collapse year, but it was not a recovery year either. It was a partial balance-sheet clean-up year in which the holding and real-estate layer softened genuine weakness in the import core. That is why 2026 looks less like a harvest year and more like a proof year: the import business needs to show that 2025 was not the start of a slide, Eurodrive needs to prove it adds economics rather than just volume, and Veridis needs to keep supporting the thesis without creating more legal and regulatory noise.

Four points to hold in mind from the start:

• The layer that generates almost half of revenue no longer generates most of the profit. Vehicle import contributed about 49% of 2025 external revenue, but only about 29% of regular profit attributable to company shareholders.
• The strong cash flow is mainly a working-capital release story. It still does not prove that the ongoing economics of the vehicle core have normalized.
• Eurodrive broadened the activity base, but it also brought an operating-lease fleet, zero-km vehicles, and structurally lower margins into the group. More volume here does not automatically mean better quality.
• Veridis and the real-estate layer provided a real cushion, but they also remind investors that value created inside the group is not always equally simple, clean, or accessible to common shareholders.

The second chart may be the single most important number set in the whole article. The "Other" segment, which includes among other things real estate, leasing, used vehicles, and white-paper activity, generated NIS 258.6 million of regular profit attributable to shareholders in 2025, versus only NIS 178.9 million for vehicle import. That is not a footnote. It means Delek Rehev can no longer be screened mainly through the business a superficial reader assumes they are buying, namely a vehicle importer.

Events and Triggers

The key events of 2025 did not happen on the margin. They changed the way the company has to be read.

The import core contracted

Delek Rehev sold 13,650 vehicles in 2025, versus 18,592 in 2024. Market share based on licensing data fell to 4.6% from 6.8%, while the business-description section presents roughly 5.4% of total industry deliveries and about 5.6% of private-vehicle deliveries. The decline was not uniform across brands: Mazda fell to 7,839 vehicles from 13,665, Ford fell to 1,137 from 2,198, while BMW rose to 3,153 from 2,638 and Dongfeng added 1,391 units after entering the portfolio.

What matters is not just which brands went up and which went down, but what that says about profit quality. Mazda is still the largest brand by deliveries, but it also took the heaviest hit. BMW improved, which helps pricing and positioning, but the numbers show it did not offset the full decline. Dongfeng creates a new option, but it is still early, competition-sensitive, and subject to an antitrust arrangement that limits import agreements to four years unless an extension is approved.

Eurodrive changed the map, not just the turnover

In January 2025 the group signed an agreement to acquire 50.01% of Eurodrive, completed the first stage in June, and completed the purchase of the remaining 49.99% in July. This was not a cosmetic deal. The financials absorbed NIS 351.5 million of operating-lease vehicles, intangible balances and goodwill increased, and Eurodrive was consolidated starting in the third quarter. At the report date, Eurodrive operated a fleet of about 3,000 vehicles, with a customer book described as mainly pharmaceutical.

This layer can deepen the group's relationship with the customer across the vehicle life cycle, not just at the point of sale. But it also has a clear downside. Eurodrive adds operating leasing, zero-km vehicles, and used vehicles, meaning activity that is more capital intensive, more exposed to residual values, and less generous on margin. Management says the synergies should bring savings and flexibility, but in the first year the consolidation also lowered the group's margin quality.

The numbers make that point clearly. In the second half of 2025 Eurodrive sold about 1,850 vehicles, and separately the group sold 1,515 zero-km vehicles through it, alongside 790 two-wheel vehicles, ATVs, off-road vehicles, work tractors, and utility vehicles. That is real volume, but not yet full economic proof.

Customer mix shifted, but not necessarily into a simpler story

In vehicle import, the share of sales to private customers rose to about 34% of group revenue, versus about 29% in both 2024 and 2023, while sales to large customers fell to about 15% of group revenue, from 24% and 25% respectively. A superficial reading could label that immediately as better quality. That is too fast.

The report itself notes that wartime periods hurt procurement by rental and leasing companies, and that the zero-km market strengthened while the group entered it through Eurodrive. So the mix shift also reflects genuine weakness in the institutional channel and an organizational change in how part of sales now flows through the group. Not every move from a large customer to a private customer is an upgrade. Sometimes it simply means the sales structure changed.

The CFO transition is not the thesis, but it is still a marker

Ronit Bachar stopped serving as CFO on April 1, 2026, after a tenure that began in March 1996, and continues as an adviser in the finance department. She was replaced by Yair Zecer, who had served as VP of strategy and innovation and, before that, had been a partner and CPA at EY Israel.

This does not replace the economic thesis, but it does deserve a marker. An abrupt departure of a long-serving CFO would have looked worse. Here the company appears to have chosen a relatively orderly handoff, without reporting unusual circumstances. The meaning is more continuity than shock, but the timing still makes 2026 a year in which the new finance leadership has to prove control over the more complex group structure.

Efficiency, Profitability, and Competition

The key 2025 number is not the decline in revenue, but the gap between the modest decline in revenue and the sharp erosion in profitability. Revenue fell about 2.9% to NIS 6.203 billion, but gross profit fell 21.9% to NIS 782.4 million, and net profit moved into negative territory. That indicates the problem was not only how many vehicles were sold, but under what terms and at what profit layer.

Breaking it down by profit drivers, three forces worked together:

• Volume: vehicle deliveries fell sharply, mainly in Mazda and Ford.
• Mix: BMW strengthened, but Mazda weakened, and Dongfeng and Eurodrive add volume that is not necessarily margin-equivalent to what was lost.
• Competition and market terms: the report describes a highly competitive vehicle market, with a wave of Chinese brands, price pressure, and a zero-km channel the group itself entered.

The hit to the vehicle core is obvious in the segment table. Regular profit attributable to shareholders in vehicle import fell to NIS 178.9 million from NIS 406.1 million. That is a drop of more than half. By contrast, parts and service improved their contribution to NIS 53.2 million from NIS 47.7 million. That is the split to remember: service and parts provide a more stable base, but it is too small on its own to carry the whole company.

The quarterly chart sharpens the likely reading error. The third quarter looks very strong, with operating profit of NIS 311.2 million, but it benefited from a NIS 221.5 million gain on investment property. The fourth quarter, by contrast, ended with operating profit of only NIS 7.8 million and a net loss of NIS 271.9 million. In other words, even after separating Hailo, year-end no longer looks like a healthy earnings engine.

Another sign sits in gross margin, which fell to 13% from 16%. The company explains that the decline came mainly from weaker profitability in vehicle sales, first-time consolidation of Eurodrive with relatively lower margins, and weaker profitability in a development real-estate project. In other words, there is no single culprit. Several pressure points worked at the same time.

Against that stands competition. The Israeli vehicle market has seen many new brands, especially from China, and competition is no longer only between the legacy importers. To hold share, an importer now needs product, price, financing, residual value, and availability. Delek Rehev still benefits from strong brands, a Nir Zvi logistics center that can store up to about 3,500 vehicles, and a broad showroom network. But the report also says explicitly that the group currently uses about 100% of its maximum storage capacity. That means there is not much logistical headroom left either.

Cash Flow, Debt, and Capital Structure

The right way to look at 2025 cash flow is through all-in cash flexibility, not through a normalized cash-generation lens. Why? Because the main story here is not only recurring cash power, but how much real room is left after the group's actual uses of cash.

Cash flow from operations jumped to NIS 1.176 billion, versus only NIS 175.8 million in 2024. But this is not a number that can simply be rolled forward. It was built mainly from an inventory reduction of about NIS 798.1 million, a NIS 260.2 million decline in other receivables and taxes receivable, a NIS 146.3 million decline in trade receivables, and another NIS 467.7 million of P&L adjustments. Offsetting that, suppliers declined by NIS 254.4 million, other payables and balances declined by NIS 212.1 million, and the group purchased NIS 86.2 million of operating-lease vehicles.

That is strong cash flow, but it is balance-sheet release cash flow. Once the rest of the year is brought in, the picture is much less dramatic. Investing cash flow consumed NIS 260.4 million, and financing cash flow consumed NIS 881.9 million. Financing uses included a NIS 194.0 million dividend, NIS 2.027 billion of loan repayments, repayment of a NIS 178.0 million put option to non-controlling interests, and lease payments of NIS 89.7 million. On the bottom line, year-end cash rose only to NIS 179.6 million from NIS 145.7 million.

In other words, the cash-flow improvement is real, but the room left at year-end is still not wide.

The balance sheet itself remains tight. Short-term bank and other credit stands at NIS 2.622 billion, while long-term loans stand at NIS 2.596 billion. The company reports a working-capital deficit of NIS 1.087 billion and explains that it stems in part from loans used to finance long-term assets that are presented within current liabilities. Management says the group can continue to raise funding, convert short-term facilities into long-term financing, or sell assets if needed, and the board report also says the group is in compliance with its financial covenants. But this does not feel like an unburdened balance sheet. It feels like a group operating with active leverage and a constant need to manage the funding side carefully.

Another layer to keep in mind is FX exposure. The report states that hedging volume is negligible relative to exposure, while most costs in the vehicle and parts segments are linked to import currencies and a material portion of liabilities is denominated in foreign currency. Add to that the fact that most bank and other loans carry variable interest rates, and the result is a structure in which rates and FX can move the outcome very quickly.

Guidance and Forward View

Four tests will define 2026:

• Further inventory release can no longer be a driver on the same scale as in 2025, so the market will look for a cleaner return to operating profitability.
• Eurodrive will need to prove that integration lifts customer value and profit, not just the number of vehicles moving through the group.
• The new model pipeline needs to stop the decline in the import core, especially in Mazda and in the electric-vehicle lineup.
• Veridis needs to keep contributing profit and cash without the investigation, subsidy denial, and legal proceedings deepening the view that group value is too "dirty."

If the next year needs a label, it is not a breakout year. It is a proof year with one foot in bridge-year territory. On one hand, the company has several visible triggers. In 2026 it expects additional models, including the Mazda CX-6e and Mazda 6e, the Dongfeng Mage, the new BMW iX3, and NIO's Firefly. That is a clear attempt to refresh the offering exactly where the market is shifting toward electric and hybrid.

On the other hand, the report shows that the question is no longer only which models arrive, but under what terms they sell. 2025 showed that vehicle sales can be defended only up to a point through mix changes, zero-km penetration, and broader service activity. For the market to change its interpretation in a positive way, it will need to see the new models support gross profit and operating profit, not just turnover.

The second point is the non-operating value layer. Investment property rose to NIS 777.5 million from NIS 519.7 million, and specifically the Hadera land designated for a power station was recorded at NIS 277.0 million. At the same time, as of report approval Veridis was in negotiations with OPC Israel regarding a sale of rights in the Hadera land for expected consideration of about NIS 450 million. If such a move is completed, it could turn part of the accounting value into more accessible value. But as of the report date there is no certainty regarding completion, terms, or final price. That makes it an option, not a base case.

Hailo will also remain part of the 2026 picture even though it is not part of the operating core. By year-end 2025 the investment had fallen to about NIS 170 million from NIS 412 million a year earlier, and the company recorded a NIS 242 million value decline. After the balance sheet date, Delek Motors extended a $9 million loan to Hailo with an option for another $3 million in June 2026 if no liquidity event occurs. That matters because it means the accounting noise around Hailo may continue, and the company may still have to put fresh cash into that pocket.

Put differently, 2026 will be judged less on reported net profit and more on three concrete questions: does the import core stabilize, does Eurodrive prove real economics, and does the holding layer produce accessible value rather than more volatility.

Risks

Dependence on manufacturers remains a core risk

The company does not sugarcoat this point. The group's import activity remains materially dependent on the franchises of its main manufacturers. A large share of 2025 import value relied on Mazda, BMW, and Ford, and the agreements themselves include sensitivity to changes in ownership or management. That does not mean there is a current negative event, but it does mean the moat is not fully proprietary. It also depends on commercial continuity that has to be maintained.

Veridis is both a support layer and a friction point

On one hand, Veridis provided meaningful economic support this year. On the other hand, this is the layer producing the clearest outside warning signal in the group. In April 2025 the support committee decided to deny eligibility for subsidies for 2022 and 2023 in Veridis subsidiaries, and the group recorded a NIS 61 million revenue reduction for 2023. At the same time, the auditors include an explicit emphasis paragraph regarding searches at Veridis offices, questioning of officers, and class actions and claims filed against group companies. That is not a technical detail. It affects value quality.

The unlisted investment book has already proved it can swallow a full year

Hailo alone erased NIS 242 million from the year. That is not only accounting volatility. It is also a reminder of the risk embedded in private technology holdings when there is no clear liquidity path, and the company even had to extend a new loan after year-end. Until this story reaches a clear liquidity event, it will continue to sit between operating performance and the bottom line.

Rates, FX, and a tight debt structure

The group is leveraged, its FX exposure is material, and hedging is negligible relative to that exposure. The company does state that it complies with financial covenants, but it does not provide, in the operating discussion, a detailed headroom picture for the key debt buckets. So any claim that Delek Rehev is balance-sheet clean would be wrong. It managed 2025 well. That is not the same thing.

Short Interest View

Short interest in the stock is still not extreme, but the direction is clear. Between November 2025 and the end of March 2026, short float rose from 0.66% to 3.21%, and SIR rose from 1.04 to 5.7. That is already above the sector averages of 0.54% short float and 1.481 SIR.

That is not proof that the short sellers are right. But it is clear evidence that the market is not buying the clean-up story easily. The increase in short interest fits a skeptical read: strong cash flow, but mostly from working capital; weaker import activity; a holding layer explaining more and more of the reported numbers; and a legal and technology risk book that has not gone away.

Conclusions

Delek Rehev ended 2025 in an in-between state. On one hand, it proved the group can release working capital, lean on profits from associates, and create real-estate value. On the other hand, the very core on which the company name rests, importing and selling vehicles, weakened materially. From here the market is likely to focus less on whether 2025 was "good" or "bad," and more on whether 2026 can replace balance-sheet tailwinds with operating proof.

Current thesis: Delek Rehev is no longer a pure vehicle importer but a layered company, and the problem in 2025 is that the layers that helped it look more stable are not the same layers proving that the import core has recovered.

What changed versus the older read? It used to be easier to view Delek Rehev as a company that mainly wins or loses through the brands it imports. By 2025 it is already clear that Veridis, real estate, and other activities are what cushion the hit, while the import engine itself contributes less and less support. The strongest counter-thesis is that 2025 was simply a one-time clean-up year, and that the new model pipeline, together with Eurodrive and continued support from Veridis, will be enough to put the company back on track without another major shock. That is a live possibility, but it still needs proof.

What could change market interpretation in the near term? First, signs that deliveries, mix, and profitability in the vehicle core are stabilizing without another major inventory release. Second, any step that turns real-estate value or Veridis contribution into more accessible value. On the other side, any worsening in the subsidy and investigation front at Veridis, or any need to inject more cash into the technology-investment book, could pull attention right back to the areas the company is trying to move beyond.

Why does this matter? Because Delek Rehev is now a classic case where the question is not just whether the company earned more or less, but which layer produced the profit and how repeatable and accessible that profit really is. For the thesis to strengthen over the next 2 to 4 quarters, the company needs to show that the import core stops eroding, that Eurodrive adds economics rather than just turnover, and that Veridis keeps supporting the story without increasing the legal burden. What would weaken that reading is another year in which improvement comes mainly from revaluations, working capital, or outside holdings while the import activity itself keeps losing altitude.