Himax Hit the Brake Hard: A Q4 Air Pocket, a $100 Million Optics Bet, and a Tape That Hates Optics

Guidance for the fourth quarter collapses 22–26% on Chinese auto-export indigestion — yet management promises double-digit 2027 growth and quietly moves the CPO option closer to reality

HIMX · Earnings Call · 2026-11-05

The cliff that arrived between quarters

Himax's third quarter was fine. Revenue of $246.8 million rose 8.5% sequentially, gross margin of 34.2% landed at the top of the guided range, and diluted ADS profit of $0.107 beat the $0.080–$0.100 outlook on lower operating expenses. Automotive driver IC sales — the Driver IC franchise that is now nearly half the company — grew double digits sequentially for a second straight quarter. Then the fourth quarter. Revenue guided down 22%–26% sequentially, with profit per ADS of $0.045–$0.065. That is not a normal seasonal fade; it is an order book that changed character. Management traced it precisely to Chinese automakers and the panel makers who serve their export programmes. The numbers they cited are stark: China's domestic sales of locally built vehicles fell 22.4% year-over-year in the first nine months of 2026, while exports of locally manufactured vehicles surged 65.8%.

some of our major customers, perhaps sensing early signals for the market conditions to take a turn for the worst, they decided to hit the brake hard, very hard. Actually, that kind of surprised us as well.
Jordan Wu, President and Chief Executive Officer · 2026-11-05

That admission — surprise — is the most honest sentence on the call. Chinese Tier 1 suppliers and their panel partners had spent two quarters front-loading inventory against export expectations that turned out to be too aggressive; the third quarter's strength was, in part, borrowed from the fourth. Management also flagged the compounding pressures: tightening trade barriers in the EU and a October cut to Chinese export tax rebates for new energy vehicles.

Ninety days ago the same voice said the opposite

On the August call, the automotive business was described as durable and additive. “we are well positioned, I think, to see robust sales growth for next year with continuously improved gross margin.” — Jordan Wu, President and Chief Executive Officer · 2026-08-06 Mainstream automotive, then over half of sales, "will continue to grow as well on top of a year of double-digit growth this year." Nothing in that language anticipated a 22%–26% sequential hole. To be fair, the framework has not changed. This quarter the CEO again guides 2027 to double-digit revenue growth, explicitly built on “a flattish global automotive shipment next year” — Jordan Wu, President and Chief Executive Officer · 2026-11-05 rather than a market recovery, with growth across DDIC, TDDI, LTDI and automotive Tcon. Full-year 2026 automotive IC revenue is still expected to grow roughly 10% despite the Q4 collapse, and automotive Tcon is tracking above 40% growth. But note the recurring shape of this cycle. On the November 2025 call, management said the auto market looked to be "showing signs of bottoming out" and budgeted only a mild recovery; what followed was a violent inventory replenishment in Q2 and Q3 2026, then a violent correction. The auto-display cycle has now overshot in both directions inside four quarters. That makes the 2027 confidence a design-win argument, not an order-visibility argument — and management conceded as much by saying near-term visibility is "limited."

What is genuinely new: optics that are not CPO

The freshest content on this call sits away from displays. Three items stand out in the company's own keyword set, all clustered near the top of this quarter's list. First, WLO technology — Himax's wafer-level optics platform — has been selected by “a leading global technology company ... to enable its advanced optical sensing solution to support autonomous driving” — Jordan Wu, President and Chief Executive Officer · 2026-11-05, with shipments starting in coming quarters. This is a new vertical: previously WLO was framed around consumer devices, wearables and AI data centres. Automotive optical sensing is a hard qualification and a long-life revenue stream if it sticks. Second, Qwen Glasses: Alibaba adopted the WiseEye AI processor for its smart glasses, arriving just one quarter after another global heavyweight put WiseEye into an AI glasses product. The technical pitch is ultra-low-power always-on visual sensing — “consuming only a few milliwatts while enabling on-device AI inference” — Jordan Wu, President and Chief Executive Officer · 2026-11-05 — plus a palm-vein module win at WAFERLOCK and, most tangibly, a capsule endoscopy project entering mass production this quarter after China's NMPA Class III approval. Third, mix. Management expects non-driver IC revenue to approach 30% of sales next year from roughly 20% today, led by automotive Tcon and WiseEye — both higher-margin than driver ICs. That mix shift is the reason gross margin can plausibly stay flat to up in 2027 even as “foundries and backend suppliers, including those in China, have announced significant price increases starting from the very beginning of next year” — Jordan Wu, President and Chief Executive Officer · 2026-11-05. Securing 2027 capacity was framed as the company's top priority, which is what a supplier to automotive customers with "far more stringent requirements for supply stability" has to say.

The CPO option, and the tape that disagrees

The CPO business got its most concrete datapoint yet: partner FOCI's board approved NT$3.33 billion — about $100 million — of capex for fab space and equipment, announced the same morning as Himax's call. Engineering ramps are underway, and management reiterated meaningful top- and bottom-line contribution starting in 2027. And yet: “the revenue outlook I just mentioned ... does not include CPO.” — Jordan Wu, President and Chief Executive Officer · 2026-11-05 After four consecutive calls promising CPO as the 2027 inflection — “we believe we can potentially see rather meaningful top line and bottom line contribution starting in 2027” — Jordan Wu, President and Chief Executive Officer · 2025-11-06 (November 2025); “our annualized CPO revenue could reach over $100 million in the so-called early stage of mass production” — Jordan Wu, President and Chief Executive Officer · 2025-08-07 (August 2025) — it is still deliberately excluded from guidance. That is prudent, but it also means the free option has been free for longer than the narrative implies. Here the price tape draws the sharpest contrast in this dossier. The co packaged optics complex is down roughly 43% over the last 30 days across names like VIAV, AVGO, ANET, LITE, CRDO and MRVL, and about 36% over 90 days — with data center interconnects and high bandwidth memory falling in similar measure. Peers who reported this week, Coherent and Viavi, are telling the same optical-signal-chain story into a market that is currently de-rating it. Himax is asking investors to pay for an optics ramp at the exact moment the optics complex is being marked down. The competitive question, meanwhile, has shifted from abstract to routine. Asked about new FAU entrants in August, the CEO argued “we are probably the only one providing optics using wafer-level optics” — Jordan Wu, President and Chief Executive Officer · 2026-08-06. This quarter the same question came back, and the answer was defensive-but-consistent: still exclusively FOCI, still in late engineering validation, and "way too early to talk about anybody's market share." A theme that gets asked twice in two quarters is a theme the market is beginning to price as contested.

What stopped being said

One silence is worth flagging. LCoS microdisplay — a fixture of this company's keyword history across 2024 and 2025, at times a top-three theme around AR glasses — is absent from this quarter's language almost entirely. In August the CEO pre-emptively retired it: LCoS "is not going to be a mass production story for next year." The smart-glasses narrative has been handed over to WiseEye, and the AR display story has quietly gone to the back burner. So the quarter reads as two stories stacked. The visible one is a cyclical air pocket in a business that is nearly half automotive display, caused by Chinese export indigestion the company did not see coming. The quieter one is a genuine attempt to re-rate the equity from display driver house to precision-optics supplier — WLO into autonomous-driving sensing, WiseEye into glasses and endoscopy, FAU into co-packaged optics. The first story is priced quarterly; the second needs 2027 to arrive. Investors are being asked to hold both, at a $1.3 billion market cap, with the optics tape working against them and the company's own near-term visibility admittedly thin.