Vishay’s beat is not the story; the misprice is a capex-heavy trough that is already showing demand repair

Vishay Intertechnology printed the kind of quarter that can look uninvestable on an EPS screen and investable in the order book, which is exactly where the variant perception sits. The headline miss was EPS of $0.01 versus $0.02, a gap that is economically small because the estimate base was near zero, while revenue of $800.9 million beat the $793.6 million Street number by +0.9%. What was priced in was a company still trapped in the passive and discrete semiconductor downcycle, with gross margin stuck below 20.0% and operating leverage absent. What actually surprised was not the penny of EPS, but the stabilization beneath it: Q4 revenue was +1.3% QoQ and +12.1% YoY, gross margin edged to 19.6%, and the next quarter in the history table shows $839.2 million of revenue, 21.0% gross margin, +4.8% QoQ, and +17.3% YoY. The thesis is that investors should treat Q4 as the confirmation quarter for a demand bottom, not an earnings power quarter, because the income statement is still absorbing elevated SG&A, input cost pressure, tariffs, depreciation, and the front end of a capacity program.

That distinction matters because the beat/miss mechanics were almost designed to distract from the turn. The Street comparison basis says revenue actuals were $800.9 million against $793.6 million, only +0.9% above consensus, and EPS was $0.01 against $0.02, off a near-zero estimate base where the percent surprise is not meaningful. Those facts are not enough to underwrite a re-rating. The surprise was that the quarter did not need heroic pricing or mix to stop the revenue slide: revenue has now moved from $714.7 million in Q4 FY2024 to $715.2 million in Q1 FY2025, $762.2 million in Q2 FY2025, $790.6 million in Q3 FY2025, and $800.9 million in Q4 FY2025. That sequence is not a one-quarter inventory refill; it is four quarters without a sequential revenue decline after the trough. The market may be missing that the EPS line is lagging the volume line, not contradicting it.

The financial trajectory is easiest to interpret by separating cyclical recovery from structural margin drag. Revenue improved from -9.0% YoY in Q4 FY2024 to -4.2% YoY in Q1 FY2025, +2.8% YoY in Q2 FY2025, +7.5% YoY in Q3 FY2025, and +12.1% YoY in Q4 FY2025. Yet gross margin moved only from 19.9% in Q4 FY2024 to 19.0% in Q1 FY2025, 19.5% in Q2 FY2025, 19.5% in Q3 FY2025, and 19.6% in Q4 FY2025. That is the crux: demand has turned before profitability has. A value trap interpretation says the margin ceiling has fallen. The better interpretation, supported by the next quarter data point of 21.0% gross margin on $839.2 million of revenue, is that utilization is beginning to matter again but is not yet large enough to offset the company’s stated tariff impacts, higher input costs, and SG&A step-up.

The company’s own wording supports that demand-first, margin-later read, and the useful quote is the one that combines guidance discipline with sequential proof. Joel Smejkal said, “For the fourth quarter, we generated revenue of $801 million, slightly above the midpoint of our guidance of $790 million and 1.3% higher than the third quarter.” This matters because management did not frame the quarter as a mix miracle or pricing event; it anchored the upside to revenue being above the midpoint and sequentially higher. On the company’s own reported basis in the call, David E. McConnell also pointed to “Gross profit was $157 million, resulting in a gross margin of 19.6%, modestly above both the midpoint of our guidance and the third quarter.” The word “modestly” is doing real work: management is not selling a snapback. It is confirming that the first derivative has turned while the absolute margin level remains depressed versus 32.0% in Q1 FY2023 and 28.9% in Q2 FY2023.

That margin gap is why the EPS miss should not be dismissed, but it should be correctly located. SG&A was $142 million in Q4, compared with $135 million in Q3 and $138 million at the midpoint of guidance, which means the company spent above the level management had pointed investors toward. GAAP operating margin was 1.8% versus 2.4% in Q3, and EBITDA margin was 8.8% versus 9.6% in Q3. The conflict is real: revenue and gross margin improved sequentially, but operating margin and EBITDA margin worsened. The reason the thesis does not collapse on that conflict is that the pressure is identifiable in the data rather than hidden in demand. If Q1 revenue is $839.2 million with gross margin at 21.0%, the confirmation signal is that incremental volume is finally carrying some margin even while the company guides SG&A to $153 million, plus or minus $2 million, and depreciation to approximately $55 million for the first quarter.

The balance sheet and cash flow make the same point in more awkward form: the business is producing cash, but not yet clean cash. Operating cash flow was $149 million in Q4, but that included $62 million from securitization of accounts receivable. Free cash flow was $55 million, compared with negative $24 million in Q3, but the quarter also carried $95 million of CapEx, including $75 million designated for capacity expansion projects. Inventory decreased $759 million, and inventory days outstanding improved to 107 days, which supports the demand repair argument because inventory is not expanding to manufacture the revenue increase. Still, the cash flow quality is not pristine when securitization contributes $62 million to operating cash and when full-year CapEx was $273 million only because delivery of some equipment related to the new 12-inch fab in Germany was delayed to Q1. Investors should not pay full-cycle multiples for Q4 free cash flow; they should pay attention to whether the inventory and backlog signals keep improving while CapEx peaks.

The capacity program is the biggest reason this print can be mispriced in both directions. Bulls can overstate near-term earnings because 2026 spending is going higher, not lower: Smejkal said, “For CapEx, we expect to spend between $400 million and $440 million during 2026.” Bears can overstate the permanence of weak free cash flow because management also framed 2026 as the peak, saying, “But I think we come over the peak of our CapEx spending in 2026, and we start to return to a more normal type of spending where we do not have a project that is $100 million, $200 million, those projects would be behind us.” The commitment and hedge are both important. “Between $400 million and $440 million” is a hard cash call. “I think” and “start to return” are softer language around normalization. The actionable interpretation is not that free cash flow inflects immediately; it is that the investment case shifts from whether Vishay survives the trough to whether investors are willing to underwrite one more year of elevated capital intensity after 8.9% in 2025 and 10.9% in the prior year.

The order book is the strongest defense against calling the capacity spend reckless. Backlog at quarter end was $1.3 billion, or 4.9 months, which is a tangible buffer against the view that the $800.9 million Q4 revenue number is merely a pull-forward. The company’s Q1 revenue expectation on the call was between $800 million and $830 million, while the quarterly history table shows Q1 FY2026 revenue of $839.2 million. Those are different reporting contexts in the provided data, so the right use is to treat the call guide as management’s setup and the history table as subsequent evidence that the setup was not too aggressive. Gross margin was expected to be 19.9%, plus or minus 50 basis points, including tariff impacts and expected continuing higher input costs, while the Q1 FY2026 table shows 21.0%. That matters because it directly challenges the bear case that tariff and input cost pressure would cap the margin recovery below 20.0%.

The tone data also argues that management delivery improved after the Q4 call, but with a caveat that prevents over-reading the transcript. In the tone history, Q4 FY2025 sentiment was 0.12, guidance_tone was 0.19, tone_confidence was 0.65, prepared_sentiment was 0.05, qa_sentiment was 0.11, uncertainty was 30.7, and qa_evasiveness was 30.8. Q1 FY2026 moved to sentiment 0.49, guidance_tone 0.52, prepared_sentiment 0.64, qa_sentiment 0.24, uncertainty 23.0, and qa_evasiveness -7.2. The call-over-call deltas are the point: sentiment +0.38, guidance_tone +0.33, prepared_sentiment +0.59, uncertainty -7.7, and qa_evasiveness -38.0. The one conflicting metric is tone_confidence, which fell -0.32 to 0.32. That means the transcript became more positive and less evasive, but the model’s confidence in that tone classification weakened. I would use the tone shift as corroboration, not primary evidence; the primary evidence is still revenue, backlog, gross margin, and CapEx.

The second-order read-through is narrower than usual because the data pack names no customers of Vishay and no suppliers to Vishay, so there is no defensible customer-specific or supplier-specific claim to make. The concrete implication is therefore for the competitive set in power discrete and adjacent components: Vishay’s latest peer-table revenue YoY of +17.3% compares with DIOD at +22.1%, 6503.T at +14.3%, 6504.T at +13.3%, ROHCY at +9.5%, 6882.T at +12.9%, 6844.T at +6.4%, and 6707.T at -17.7%. On growth, Vishay is behind DIOD but ahead of the other named peers in the table. On gross margin, however, Vishay’s 21.0% trails DIOD at 31.8%, 6503.T at 32.2%, 6504.T at 31.0%, ROHCY at 26.7%, and 6882.T at 28.3%, while exceeding 6844.T at 14.5% and 6707.T at 0.7%. That spread is the competitive issue: Vishay’s revenue recovery is already visible, but its margin structure still screens as mid-pack to lower-pack among named peers.

That peer comparison is also why the stock should not be bought simply because the cycle has turned. The variant perception is more specific: Q4 reduces downside risk to the revenue trough and raises confidence that the 2026 capacity peak is bridgeable, but it does not yet prove that Vishay can recapture anything close to the 32.0% gross margin seen in Q1 FY2023. The company exited Q4 with $515 million of global cash and short-term investments, but also with $219 million outstanding on its US revolver, and Q4 shareholder returns consisted of a $13.6 million quarterly dividend. That mix is acceptable if the next several quarters convert backlog into revenue and gross margin expansion; it becomes harder to defend if CapEx rises to $400 million to $440 million while gross margin remains around 19.6% and SG&A sits near $153 million, plus or minus $2 million. The market is right to demand proof, but wrong if it treats Q4 EPS of $0.01 as the clean measure of normalized earnings power.

What to watch next is therefore precise. The thesis is confirmed if the next quarter holds revenue at or above the company’s $800 million to $830 million guide range, or closer to the $839.2 million shown in the quarterly history, and if gross margin stays above the guided 19.9%, plus or minus 50 basis points, with the history table’s 21.0% as the level that would validate operating leverage. It is also confirmed if backlog remains near the Q4 level of $1.3 billion or 4.9 months while inventory days stay at or below 107 days, because that would show shipments are not being manufactured by channel stuffing. It breaks if SG&A runs materially above $153 million, plus or minus $2 million, if Q1 tax expense is outside the expected $2 million to $4 million range without a mix explanation, or if CapEx for 2026 moves above $400 million to $440 million without a matching improvement in revenue growth and gross margin. The date is the next quarterly report after the 2026-02-04 Q4 FY2025 call; by then, investors should be able to tell whether Q4 was the first investable trough print or just another quarter where revenue recovered faster than earnings.