Sequans’ Miss Was Not the Revenue Line, It Was the Credibility Gap Between a $40 Million Plan and a 37.7% Margin Base

The important conclusion from this print is not that Sequans missed revenue by -1.4%; that was too small to be the real event. What matters is that the company is asking investors to underwrite a 2026 inflection from approximately $27,200,000 of 2025 revenue to a plan of approximately $40,000,000 to $45,000,000 while exiting Q4 FY2025 at $7.0 million of street-comparison revenue, 37.7% gross margin, and -$1.19 EPS versus the Street’s -$0.49. What was priced in looked like a modest Q4 revenue reset after Q3 FY2025 revenue of $4.3 million and a Street estimate of $7.1 million; what actually surprised was the depth of the loss, with EPS surprising by -142.9%, and the fact that the company’s bridge to 2026 depends on timing-sensitive shipments, production conversion, and mix normalization rather than a demonstrated margin turn in the reported quarter.

That distinction matters because the headline revenue result was almost exactly what investors were set up to expect. The print showed actual revenue of $7.0 million against the $7.1 million estimate, a -1.4% surprise, and management’s own account was consistent with that framing when Georges Karam said, “In the fourth quarter, it generated $7,000,000 in revenue, which was in line with our prior expectations.” The problem is that a revenue result close to plan did not prevent the EPS shortfall: actual EPS was -$1.19 versus estimate -$0.49. In a cleaner recovery setup, a near-in-line top line would have been enough to defend the quarter; here, it exposed how little operating leverage the model currently has at the present revenue and gross margin base. Q4 FY2025 revenue improved from $4.3 million in Q3 FY2025 to $7.0 million, and the table shows revenue QoQ of +62.2%, but gross margin still moved only to 37.7% from 40.9%, while diluted EPS in the quarterly history was -$6.30. The street-comparison EPS basis and the quarterly history diluted EPS basis are not the same, but both point in the same direction: the revenue rebound did not translate into earnings quality.

The financial trajectory therefore argues against paying for the 2026 target as if it were already de-risked. Sequans’ historical revenue profile has been unstable: $11.0 million in Q4 FY2024, $8.1 million in Q1 FY2025, $8.1 million in Q2 FY2025, $4.3 million in Q3 FY2025, and $7.0 million in Q4 FY2025. Gross margin has also deteriorated from 67.4% in Q4 FY2024 to 64.5% in Q1 FY2025, 64.4% in Q2 FY2025, 40.9% in Q3 FY2025, and 37.7% in Q4 FY2025. The market may be anchoring on the sequential revenue recovery and the 2026 plan, but the quarter says the business is still below the margin level that would make the revenue inflection self-validating. Management gave investors a way to think about a better model, with Georges Karam saying “if you have a gross margin around on the product, 45%, you can do it with $13,000,000,” but that comment itself defines the gap: Q4 FY2025 gross margin was 37.7%, and Q1 FY2026 in the quarterly history also shows 37.7%. The print did not prove the company is at the margin structure management used in that cash-flow discussion.

The 2026 plan is the bull case, and it is real enough to matter, but not yet bankable enough to close the debate. Karam put the company’s internal plan at “approximately $40,000,000 to $45,000,000 of total global revenue supported by improving visibility and a significant order backlog.” That is the sentence bulls will own, because it reframes the company away from a $7.0 million quarter and toward a higher annualized revenue base. The supporting data are also not trivial: the company said it is exiting 2025 with a revenue funnel exceeding $550,000,000 in potential three-year product revenue, including over $300,000,000 from design win projects, and approximately $132,000,000 of potential three-year revenue from production stage projects alone. Management also set a target of having over 50% of current design win projects in production by June. The variant perception is that these numbers should be treated as option value, not as current earnings power, because Q1 2026 guidance still points to revenue around $6,500,000 with the risk that approximately $1,000,000 could shift into Q2 due to manufacturing and shipment timing planned for the end of Q1.

That Q1 comment is where the 2026 story either begins to become investable or starts to slip again. A company targeting approximately $40,000,000 to $45,000,000 for 2026 cannot afford too many quarters where manufacturing timing moves approximately $1,000,000 across the boundary, especially when Q4 FY2025 street-comparison revenue was $7.0 million and Q1 revenue is expected to be around $6,500,000. The quarterly history shows Q1 FY2026 revenue of $6.1 million and revenue QoQ of -12.6%, which conflicts with the call guide of around $6,500,000 because the table and the call are different reporting contexts and timing bases. The right interpretation is not to average them; it is to recognize that the setup is still dependent on shipment timing and conversion timing, exactly the two variables management flagged. If approximately $1,000,000 shifts into Q2, investors should not call it immaterial simply because the full-year plan is larger. At Sequans’ current revenue scale, approximately $1,000,000 is large enough to change whether Q1 looks like an orderly seasonal step or another deferral-driven quarter.

The balance sheet and Bitcoin overlay complicate the equity story because they introduce non-operating volatility into a company that still needs operating proof. Karam said Bitcoin NAV was about $150,000,000 when Bitcoin was approximately $70,000, and Deborah Choate said year-end 2025 holdings were 2,139 Bitcoin with a market value of $187,100,000. That asset base is not clean surplus capital: 1,617 Bitcoin, valued then at $141,500,000, were pledged as collateral for the remaining $94,500,000 of convertible debt due in July 2028, while 522 Bitcoin, valued at year-end at $45,600,000, were unencumbered. Q4 also carried a noncash impairment charge of $56,900,000 related to mark-to-market value of Bitcoin holdings, compared with an $8,200,000 charge in Q3, and the company sold Bitcoin to fund $101,000,000 of debt redemption and a $9,400,000 ADS buyback after completing Bitcoin purchases totaling $3,400,000 early in the quarter. This matters for semiconductor investors because the balance sheet is not just a liquidity backstop; it is a source of income statement noise, collateral constraints, and capital allocation questions at the same time the core business is trying to prove production conversion.

The expense line gives management some credit, but not enough to rescue the quarter’s earnings interpretation. Choate said R&D and SG&A expenses declined to a combined total of $11,500,000 in Q4, down from $13,600,000 in the third quarter. She also said non-IFRS net loss was $18,500,000 or $1.19 per ADS, compared with a non-IFRS net loss of $11,300,000 or $0.81 per ADS in Q3, and that the realized loss on the sale of Bitcoin of $8,400,000 is included in the non-IFRS net loss, so the company would have been just over $10,000,000 non-IFRS net loss without that element. This is the strongest defense of the EPS miss: stripping the realized Bitcoin sale loss gets the non-IFRS loss closer to the operating scale implied by the cost cuts. But it does not eliminate the central issue, because the company still needs either materially higher revenue, higher product gross margin, more service mix, or some combination of those to make the model work. Karam’s mix comment pointed to “80-85% product and only 15% services” or “20% services,” with service described as “100% margin,” but Q4 reported gross margin of 37.7% shows investors have not yet seen that mix and product-margin framework in the consolidated numbers.

The call delivery reinforced the same message: management sounded more constructive in prepared remarks, but the guidance tone has not returned to the confidence levels seen before the 2025 deterioration. The tone history shows Q4 FY2025 sentiment at 0.18 and guidance_tone at 0.18, with prepared_sentiment of 0.01 and uncertainty of 89.3. The subsequent Q1 FY2026 call improved to sentiment of 0.30, guidance_tone of 0.46, tone_confidence of 0.38, and prepared_sentiment of 0.50, but uncertainty stayed high at 90.1 and ai_optimism fell to -0.30. The call-over-call deltas sharpen the point: sentiment rose +0.13 and guidance_tone rose +0.28, but uncertainty rose +0.7 and qa_sentiment declined -0.04. The most useful change was qa_evasiveness falling by -72.0, from 89.7 to 17.7, which suggests management was more direct in Q1 FY2026 than in Q4 FY2025. Still, directness is not the same as conversion. The tone data support a view that management is explaining the plan more clearly, not that the plan has already been executed.

The supply-chain read-through is unusually narrow because the data pack names no customers of SQNS and no suppliers to SQNS, so the second-order implication is not a customer-specific order revision but a timing signal for the LTE, CAT 1 Bis, and early 5G ecosystem around Sequans’ programs. The named quantitative read-through is the company’s own project base: over $300,000,000 from design win projects, approximately $132,000,000 of potential three-year revenue from production stage projects alone, and a target of having over 50% of current design win projects in production by June. For unnamed customers, the print says production ramps exist but are still lumpy enough that approximately $1,000,000 of Q1 revenue could move into Q2 because of manufacturing and shipment timing. For unnamed suppliers, the same language implies that near-term component pull and shipment activity should be judged by end-of-quarter conversion rather than by the size of the three-year funnel. The absence of named counterparties also matters for portfolio construction: there is no clean read-through to a specific customer or supplier stock from this data pack, only a company-specific execution bar.

Against peers, Sequans is not trading in the same operating regime as the larger fabless and platform companies in the table, which is precisely why revenue conversion matters more than end-market narrative. NVDA reported revenue of $81,615.0 million and gross margin of 74.9% with revenue YoY of +85.2%; META reported gross margin of 81.9% with revenue YoY of +33.1%; MSFT reported gross margin of 67.6% with revenue YoY of +18.3%; GOOGL reported gross margin of 62.4% with revenue YoY of +21.8%. Sequans’ Q4 FY2025 revenue was $7.0 million with gross margin of 37.7% and revenue YoY of -37.0%. The comparative point is not that Sequans should look like those companies; it obviously will not at this scale. The point is that the market cannot use sector multiple expansion or broad fabless sentiment as a substitute for proof that Sequans can move from production-stage potential revenue into reported revenue while lifting gross margin from 37.7% toward the product-margin framework management described.

The actionable stance after this print is to treat Sequans as a call option on 2026 conversion with a stricter proof requirement than the stock likely wants. The revenue miss was only -1.4%, so shorting the quarter on top-line disappointment alone misses the point. The EPS surprise of -142.9%, the 37.7% gross margin, the non-IFRS net loss of $18,500,000, and the reliance on timing-sensitive Q1 shipments make the print less clean than the $40,000,000 to $45,000,000 2026 plan suggests. At the same time, the funnel exceeding $550,000,000, over $300,000,000 from design wins, approximately $132,000,000 from production-stage projects, and the target of over 50% of current design win projects in production by June are too large to dismiss. The right variant perception is that the market should not pay full credit for the plan yet, but should be ready to change its mind quickly if Q1 and Q2 show that shipment timing is a deferral rather than demand leakage.

What to watch next is concrete. First, Q1 2026 revenue needs to land around the company’s $6,500,000 guide, and if approximately $1,000,000 shifts into Q2, management must identify that shift explicitly and recover it in Q2 rather than resetting the annual plan. Second, gross margin needs to move off the 37.7% level shown in Q4 FY2025 and Q1 FY2026 if investors are to believe the 45% product-margin discussion can become reported economics. Third, by June, the company needs to show progress toward having over 50% of current design win projects in production, because that is the dated milestone connecting the more than $300,000,000 design-win base to the approximately $40,000,000 to $45,000,000 2026 revenue plan. Fourth, watch whether non-IFRS net loss moves closer to the “just over $10,000,000” level Choate cited excluding the $8,400,000 realized Bitcoin sale loss, and whether R&D and SG&A stay near the Q4 combined total of $11,500,000 rather than reverting toward $13,600,000. If those numbers line up next quarter, the thesis breaks bullish; if revenue slips without Q2 recovery, gross margin stays at 37.7%, or the June production target softens, the market is still over-crediting backlog and underpricing execution risk.