NEXT PICK · Market Insights
After dropping 9.78%, is LITE still expensive?
Monday, October 12, 2026
Fundamentals are still accelerating, and stock prices have already been overdrawn: a single sell-off-like pullback neither proves a bubble nor overturns the logic.
No negative news, yet nearly $10 billion evaporated in a single day
On October 12, Lumentum ($LITE) closed at $995.41, down 9.78% in a single day, with trading volume expanding to 1.4 times the daily average over the past three months. Its market capitalization shrank by nearly $10 billion in a single day, returning to $90.17 billion.
Strangely, the company neither released its financial report nor was its rating downgraded that day. At the OCP Global Summit, it jointly showcased an AI rack optical interconnect solution with Astera Labs and Corning. What should have been good news was replaced by the market instead of "selling the news"; On October 1, the old news of CFO Wajid Ali selling 24,542 shares at around $1,048.50 per share was resurfaced and added fuel to the fire.
On the same day, Coherent and Applied Optoelectronics also fell 7%, while the S&P 500 remained unchanged. This was a collective profit-taking event belonging only to the AI optical communications sector.
So here's the question
Why can't a company with quarterly revenue doubling year-on-year and production capacity reportedly scheduled for early 2029 not even be able to hold up its stock price with good news?
|
Latest price
$995.41
▼ -9.78%
|
Quarterly revenue
$1.01 billion
▲ Year-on-year +109%
|
Next quarter's guidance
$1.25 billion
▲ Quarter-on-quarter +24%
|
Selling the "source of light," not optical modules
Lumentum is headquartered in San Jose, California, and its business is divided into two areas: the component business focuses on semiconductor laser chips, laser subcomponents, circuit subsystems, and wavelength management; the systems business covers optical modules, optical circuit switches, as well as short-pulse solid-state lasers and kilowatt-level fiber lasers and other industrial laser products.
If you imagine an AI data center as a city that transmits information through light, companies like Zhongji Accelink are making "street lamps," while Lumentum is making the glowing chip inside the lamp. NVIDIA, Google, Meta, Microsoft, Micron, and SanDisk are all on its customer list.
It is difficult to replace because the two core light sources of AI optical modules—EML (Electromagnetic Absorption Modulation Laser) and CW-DFB (Continuous Wave Distributed Feedback Laser)—have highly concentrated global monthly production capacity. According to TrendForce's data from June 2026, Lumentum, Broadcom, and Sumitomo Electric together account for about 55%.
In the VCSEL (Vertical Cavity Surface-Emitting Laser) field, it is also one of the six major global production suppliers, with the other five being Furukawa, Coherent, Broadcom, ams OSRAM, and HiSilicon. Vertical integration from chip to module to subsystem makes it difficult for customers to change once selected.
The research report rated its irreplaceability as a 9 out of 10. The question is, how many times more revenue is this "chokeneck" capability worth?
After the static P/E ratio fails, which ruler should you use?
Let's look at the most glaring numbers first
For fiscal year 2026 (ending June 2026), GAAP loss per share is $92.96, with a net loss of $6.9 billion, so the static price-to-earnings ratio (P/E) cannot be calculated at all, and the market software shows it as 0.
However, this loss was almost entirely due to the $7.8 billion convertible bond conversion accounting in the fourth quarter, not because the business was losing money. GAAP operating profit for the same fiscal year was $524.8 million, compared to an operating loss of $180.1 million in the previous fiscal year; Non-GAAP full-year earnings per share were $8.67.
Using a more business-oriented scale, based on $3.01 billion in annual revenue, the price-to-sales ratio (P/S) is close to 30 times; Based on Wall Street's earnings forecast for the next year, the forward-looking P/E ratio is between 43 and 45 times. The hundreds of times P/E ratios previously calculated in research reports based on old data are outdated, but the conclusion of "expensive" has not been overturned.
The DCF model in the report is even more stern: assuming revenue growth rates of 50%, 35%, or 25%, the calculated per-share value is only about 30% of the current stock price, and even the optimistic scenario is just over half.
However, this model assumes net margin climbs to 17% year by year, while the company's non-GAAP operating margin guidance for the next quarter is already between 39.5% and 40.5%. The model underestimates profit elasticity, while the market may overestimate its sustainability—the truth is likely somewhere in between.
Cash flow: From burning cash to self-sufficiency, but it can't support this market value
The report estimates annual free cash flow at $751.4 million, a 495% increase from $126.3 million the previous year, showing the company is now able to expand production on its own. As of the end of June 2026, cash on hand stands at $2.7 billion, an increase of $1.9 billion over the year.
The debt structure has undergone a "major shift"
Long-term debt maturing within one year rose from $10.6 million to $1.5969 billion, while long-term debt fell from $2.5626 billion to $40.5 million. This means the remaining convertible bonds are nearing maturity or conversion, and total debt is actually declining, at the cost of diluting equity from 69.8 million shares to 88.6 million shares.
However, comparing $750 million in free cash flow to a $90 billion market cap, the free cash flow yield is less than 1%. Today's positive news has clearly been mostly priced into the stock price. This decline is less bearish and more a market question: Where is the next surprise?
From concept to a surge in volume, the breakthrough period is far from over
The report marks Lumentum as in the "breakthrough period," with a fairly complete logical chain: Q3 revenue was $808.4 million, up 90.1% year-over-year; Q4 revenue surpassed $1.01 billion, up 109% year-over-year, with non-GAAP gross margin surpassing 50% for the first time.
Management has set revenue guidance of $1.225 billion to $1.275 billion for the next quarter, with CEO Michael Hurlston stating it was "to reach the target model more than a quarter ahead of schedule." TrendForce forecasts a 57% growth in the AI optical transceiver market by 2026, providing cross-validation from upstream in the industry chain.
From a technical perspective, even though the stock price plunged, it remained above the 50-day and 200-day moving averages, and the medium- to long-term trend remained intact. The market still agrees this is an accelerating cycle, but is no longer willing to pay any price for it.
In a three-way tribal oligarchy, neither side can shake off the other
Coherent ($COHR) is Lumentum's most similar competitor, also one of the six major VCSEL suppliers, and also has vertical integration capabilities from wafer to module. The two sides are head-to-head on orders from NVIDIA, Google, and Meta. In the new track of VCSEL and near-packaged optics, both companies are still at the starting point.
Broadcom ($AVGO) is the core competitor in EML and CW-DFB fields. Its greatest advantage lies in its ability to coordinate lasers and switching chips in collaborative design, making it the most comprehensive player in the AI optical interconnect market; Sumitomo Electric, leveraging its Asian manufacturing foundation and cost advantages, influences the supply-demand balance in the EML market.
Looking further downstream, Zhongji Xuchuang's revenue growth is close to 200%. Huawei has a complete product line in optical communications, and Lumentum's annual report clearly lists it as a strategic competitor. Although Lumentum is upstream and does not compete with module manufacturers, the speed at which Chinese manufacturing is catching up should not be underestimated.
A moat is a dam, not a wall
The report rates the moat as strong, with two pillars: technical barriers and switching costs. Optical chips are deeply coupled with switching chips and server architectures; switching suppliers means 12 to 18 months of recertification, design, and testing.
The other three layers of reinforcement come from economies of scale, supply-side "network effects," and export control compliance thresholds: the tighter the capacity, the more customers have to sign long orders to lock in, and the more concentrated supply is. But the height of the dam depends on the water level; once demand retreats, the bargaining power caused by tight capacity also diminishes.
Rather than saying Lumentum has a permanent moat, it's more accurate to say it stands at the top of a supply cycle — how long can this cycle last?
The ceiling is still rising, and the estimates in research reports have already been surpassed by reality
TrendForce forecasts the global AI optical transceiver market to reach $26 billion by 2026, with a year-on-year growth rate exceeding 57%; VCSEL demand is expected to grow at a compound annual rate of about 38% from 2025 to 2031, reaching 650 million units by 2031; EML plus CW-DFB monthly capacity will total 50.7 million units by 2026, with optical transceiver shipments exceeding 92 million units.
The framework of the research report is:
The potential market volume (TAM) exceeds $35 billion, the serviceable market (SAM) is $10 to $13 billion, and the market of value (SOM) corresponds to annual revenue of $2.5 to $3.5 billion; The next fiscal year is estimated to be a conservative $2.1 billion, a baseline $2.8 billion, and an optimistic $3.6 billion.
Interestingly, Lumentum's fiscal 2026 revenue has reached $3.01 billion, with guidance for the next quarter at $1.25 billion, annualizing close to $5 billion. The optimistic scenario in research reports is turning into a realistic bottom line — this precisely shows that demand growth far outpaces all other models.
Six cards in the hands of the bulls
Bulls on Lumentum are betting on the triple overlap of AI optical internet demand, upstream positioning, and profit inflection points:
| ▲ Bull Case |
| ① | The AI optical transceiver market is expected to reach $26 billion by 2026, growing 57%, with Lumentum being one of the top three EML and CW-DFB players |
| ② | Fourth quarter revenue was $1.01 billion, up 109% year-over-year, with non-GAAP gross margin surpassing 50% for the first time |
| ③ | Next quarter's median revenue guidance is $1.25 billion, with non-GAAP operating margin guidance close to 40%, meeting targets more than a quarter ahead of schedule |
| ④ | With $2.7 billion in cash on hand, free cash flow has turned positive, and expansion does not require external funding |
| ⑤ | Management stated that capacity is scheduled for early 2029, and the optical interconnect solution in collaboration with Astera and Corning is expected to enter mass production between 2027 and 2028 |
| ⑥ | The average seller target price remains above the current stock price, and most institutions maintain a buy rating |
The bears were staring at the six cracks
The logic of bears is equally deadly, and today's decline serves as a footnote to them:
| ▼ Bear Case |
| ① | Static P/E ratio is a GAAP loss effect, with a price-to-sales ratio close to 30 times and a free cash flow yield below 1%. |
| ② | The benchmark DCF valuation in the report is only about 30% of the current stock price, making the valuation heavily dependent on growth narratives |
| ③ | GAAP net loss for fiscal year 2026 is $6.9 billion, with equity diluted from 69.8 million shares to 88.6 million shares in one year |
| ④ | Shortages in EML and other optical chip capacity are hard constraints on expansion, and even with long-term contracts, it's hard to fully lock in upstream supply |
| ⑤ | China accounts for 56% of global optical module foundry capacity; if the U.S. restricts imports, the supply chain will be forced to restructure |
| ⑥ | CFOs sold 24,542 shares near $1,048.50, and insider reductions weakened market confidence |
In the coming months, who will decide the direction of the next step?
In the short term, stock price rhythm will be dominated by financial reports, while in the medium to long term, technical and policy will be the focus:
| • | On November 5, after the market closed, the first quarter financial report was released: the revenue guidance of $1.225 billion to $1.275 billion was validated, with a neutral direction, and the exceeding margin determined the response |
| • | PCIe Optical Interconnect Showcase at OCP Summit: Today has been digested by "selling news," with a negative short-term impact |
| • | VCSEL Near-Packaged Optical Solutions Mass Production in 2027-2028: Opening a New Market for Intra-Rack Optical Interconnect, Long-Term Positive |
| • | By the end of 2026, the AI optical transceiver market will reach $26 billion, with monthly EML production capacity reaching 50.7 million units: the industry outlook is positive |
| • | US restrictions on Chinese optical module imports: uncertain timeline, supply chain restructuring, and negative side |
| • | Nokiado PON line cards to launch in 2027, Gigalight to launch 3.2T hybrid semi-DSP modules: new product cycle, moderate impact |
From building the bottom to Davis's double play, and then switching hands high up
Fiscal year 2024 is a low point: net loss of $546.5 million, the optical communications industry is in a downturn, and AI demand has yet to truly explode. Fiscal year 2025 revenue is $1.65 billion, GAAP net profit turned positive to $25.9 million, yet there is still an operating loss of $180.1 million—the turning point is just beginning.
The real revaluation occurred in fiscal year 2026: revenue grew 83.2% year-over-year to $3.01 billion, operating profit was $524.8 million, and third-quarter net profit was $144.2 million, compared to a loss of $44.1 million in the same period last year. Over the past 52 weeks, the stock price has traded between $152.54 and $1,143, still up 170% year-to-date.
The bulls' demand logic has basically been realized, and the bears' earnings concerns have been temporarily disproven by data, but valuation issues have not disappeared. Today's sharp drop occurred not far from the high, more like a chip turnover at the peak than a trend reversal—so will the next earnings report be a survival, or a retreat?
|
⚠️ Risk Notice
|
|
🟡 Neutral Fundamentals are accelerating, but most of the positive news has already been priced into the stock price; the logic behind the pullback is unchanged and valuation is unclear. |
|
💬 Discussion A good company doesn't equal a cheap stock; chasing highs isn't worth it. Real-time buy and sell levels, post LITE on the official account to check. |
Data source
| • | Source: NextPick real-time snapshot + company earnings and SEC announcements + TrendForce industry data + mainstream financial media. |