LITE's 9% rise in a single day was not due to its own news

NEXT PICK · Market Insights

LITE's 9% rise in a single day was not due to its own news

Monday, August 3, 2026

A company with a market capitalization of $60.68 billion surged 9.24% at $779.89 on Monday, when it didn't announce any of its own announcements. It hit an intraday high of $785.34 and plunged to $671—a 17% swing in one day, wilder than many small-cap stocks.

It wasn't Lumentum's own news that drove it, but rather the overall risk appetite rebound in the optical communications sector triggered by peer Coherent. In other words, those who bought LITE today are buying sentiment across the entire AI optical interconnect sector, not any new fundamentals added by the company that day.

Even more subtle, just last Tuesday, the same batch of stocks collectively plunged due to market doubts about AI capital expenditures, with LITE dropping about 9% in a single day. Within five trading days, the same stock is first rejected and then affirmed by the market with almost the same magnitude. So, is this a confirmation of fundamentals, or a back-and-forth swing of highly volatile chips?

Latest price $779.89 ▲ +9.24%

Rolling P/E ratio 136.58 times ▼ The forward is about 44 times

Single-quarter revenue 808 million USD ▲ +90%

Of the 9% increase in a single day, how much of it was earned by itself?

Looking at today's bullish candlestick, there are actually very few components that can be attributed to Lumentum itself. The trigger lies in peers: news from Coherent has driven the entire optics and photonics sector toward risk appetite, with Applied Optoelectronics, Coherent, and Lumentum almost all rising and falling in a single package.

This high correlation itself sends a message—it shows that the current market does not distinguish between the technology routes, customer structures, and capacity ramp-up paces of these three companies, but treats them all as proxy products for the single variable of "AI data center capital expenditure."

The price of agent varieties is volatility. On July 28, the market doubted the pace of AI spending, with LITE dropping about 9% in a single day; On July 31, sentiment recovered, rising about 5%; On August 3, another stock rose 9.24%. During this period, the company's orders, production capacity, and customers have not changed at all.

So, the honest characterization of this unusual move on the trending list is: emotion-driven, not company event-driven. The real answer will only be revealed after the market closes on August 11.

What they sell is not modules, but the "passport" for AI computing power

Lumentum is headquartered in San Jose, California, and its business is divided into two segments: Cloud & Networking and Industrial Tech. The former is the protagonist of all current stories, supplying photonic chips, components, modules, and subsystems to cloud data center operators and network equipment manufacturers.

The revenue structure for the latest quarter is very clear

Of the $808.4 million in revenue, components contributed $533.3 million, systems contributed $275.1 million, with components accounting for about two-thirds. This means its value centers upstream on the chip side, rather than in the assembly stage of finished modules.

The real spot where it gets stuck is the indium phosphide (InP) laser chip. EML (Electromagnetic Absorption Modulated Laser) and CW-DFB laser diodes are essential light sources in 800G and 1.6T optical transceivers, as well as CPO (Co-packaged Optics) optical engines. Without it, data between GPUs can't leave the cabinet.

Manufacturing these chips involves dozens of precision steps such as epitaxial growth, lithography, etching, cleavage, and coating, and yield control is a skill accumulated over decades. The number of players capable of large-scale delivery worldwide can be counted on one hand: Broadcom, Lumentum, and Sumitomo Electric together hold about 55% of the EML/CW-DFB market share.

Another often overlooked Industrial Tech area is the cycle buffer—short-pulse solid-state lasers, kilowatt-class fiber lasers, and diode lasers, serving semiconductor manufacturing, solar cells, display panels, and electric vehicle battery production lines. When AI cools down, it at least isn't walking on one leg.

So how should we understand this company? Rather than treating it as a "hardware factory selling optical modules," think of it as a toll station for an AI data center: every optical path to 1.6T has to buy a luminescent pass from such suppliers.

433x, 136x, or 44x—which is real?

First, let's correct a widely circulated number. Many materials on the market list LITE with a price-to-earnings ratio of over 400 times, which is a distortion calculated on a GAAP basis when earnings have just turned positive from the loss zone. Based on the current stock price of $779.89 and rolling earnings per share of $5.71, the price-to-earnings ratio is 136.58 times; Public data sources give forward P/E ratios of about 44 times.

The huge gap between 136 and 44 times is the main controversy about this stock—it's not about valuation, but whether you believe profits can more than double again.

Other anchor points are also expensive

The price-to-sales ratio (P/S) is about 25 times, the price-to-book ratio (P/B) is about 20 times, the enterprise value multiple (EV/EBITDA) is as high as 254 times, and the free cash flow yield is only 0.19%. Even adjusting for revenue growth of over 100%, PEG is already at 1.29—at the upper end of the reasonable range, not undervalued.

So what gives the market such a price? The implicit assumption is: in the next three to five years, revenue growth rates will maintain a compound annual growth rate of over 50%, and operating profit margins will expand all the way to over 25%. This is not a "conventional premium for growth stocks"; it is pricing that "must consistently exceed expectations."

The cash flow side has indeed improved. Operating cash flow for the first nine months of fiscal year was $388.4 million, compared to $62.3 million in the same period last year—a more than fivefold increase; Cash and equivalents on hand rose from $520.7 million to $2.6178 billion, indicating ample liquidity.

But on the same table, there are two hidden reefs that cannot be ignored

Capital expenditure rose from $177.1 million to $284.5 million, with capacity expansion moving on cash; Meanwhile, long-term debt due within one year surged from $10.6 million to $3.2386 billion (including reclassification of convertible preferred shares and 2032 notes), sharply increasing short-term debt repayment pressure.

Overall, daily operations and capacity expansion can now be covered by its own cash flow, and the company no longer needs to rely on financing to survive. But "self-sufficiency" and "valuation safety" are two different things—a reasonable value calculated by scenario weighting is actually about 12% lower than the current stock price. Today's 9.24% increase reflects sentiment, not a margin of safety.

From proof of concept to large-scale shipments, the market has already accepted it

In terms of stage assessment, Lumentum has long surpassed commercial validation and entered the break a surge in volume out stage, based on three sets of hard numbers.

First, revenue accelerated for three consecutive quarters: 533.8 million→ 665.5 million→ $808.4 million, with a 90% year-on-year increase in the most recent quarter, and a cumulative total of $2.0077 billion for the first nine months, up 72.4% year-on-year.

Second, profits have turned around with certainty. The full fiscal year posted an operating loss of $180.1 million, but operating profit for the most recent quarter reached $174.5 million, gross margin expanded from 28.8% to 44.2%, GAAP earnings per share were $1.50, and non-GAAP earnings per share were $2.37. Net profit for the first nine months turned positive from -$187.4 million in the same period last year to +$226.6 million.

Third, NVIDIA's CPO has entered mass production and serves the Vera Rubin platform, no longer just a roadmap on a PPT, but rather shipments on the production line.

Does the market recognize this stage? The technical answer is: "I admit it, but I've already gotten tired of it." The stock price is still nearly 30% above the 200-day moving average — the long-term trend remains intact; However, it also fell below the 50-day moving average and briefly stood above the 1,000 yuan mark mid-year before entering a correction. The cumulative return over the past 30 days is -12.8%, with the RSI (14) returning to the neutral zone at 51.1.

Long-term momentum remains on track, but medium-term momentum has weakened—this is exactly what a stock moving from "value discovery" to "valuation digest" should look like.

In the dual-oligopoly structure, its rivals most resemble it are also on the same contract

Lumentum faces few competitors, but none are easy to deal with.

The most direct is Coherent (NYSE: COHR). Both are signatories for NVIDIA's major order, and their product lines also cover InP lasers, optical transceiver modules, and CPO light engines. Coherent posted revenue of $1.805 billion in the most recent quarter, up 21% year-over-year, with the Data Center & Communications segment jumping 41% to $1.361 billion, accounting for 75% of total revenue. At the module level, it has stronger vertical integration, but in terms of EML bare chip capacity, the two are evenly matched, forming a typical duopoly in NVIDIA's supply chain.

Broadcom (AVGO) ranks among the top three in both EML and CW-DFB markets, but its core profits come from the Tomahawk and Jericho series of switch chips. Its optical chips are mostly self-supplied with limited export shares, making its direct impact on Lumentum controllable. Sumitomo Electric's strength lies in the vertical integration of InP substrate materials, allowing it to integrate CPO light engines in a deeper integration with customers compared to Lumentum.

As for new entrants, emerging InP laser suppliers like SIVE are seen by some research circles as potential "laser throat" participants. If foundry integration goes smoothly, some mid- to long-term EML demand may be diverted, but it will be difficult to reach NVIDIA-level high-end product lines in the short term.

Just how deep is the moat? Just look at one number: from design import, sample validation, small-batch trial production to scaling up mass production, optical chips usually takes 12 to 18 months, and deep collaboration with customers' switching chips, DSPs, and packaging solutions is required. Once mass production begins, switching suppliers is like setting back the clock of the entire production line by a year and a half. In years when everyone is competing for capacity, no hyperscale customer is willing to pay this price.

The ceiling is high enough, but market share and penetration are the real variables

In terms of total volume, the AI cluster optical market is expected to grow by 60% in 2026, reaching $26 billion. Annual shipments of optical transceivers are expected to exceed 92 million units, with 800G and 1.6T being the main growth drivers. ERCOT Texas received 474 GW of data center interconnection applications in just one state, about five times its peak demand—a figure that shows both the staggering potential scale and the fact that many of these projects are destined to fail.

In the serviceable market (SAM), the combined monthly production capacity of EML and CW-DFB laser diodes in 2026 is expected to be about 50.7 million units, doubling year-on-year, with the industry still facing a structural gap of over 30%. CPO penetration is expected to climb to 35% by 2030, at which point single-machine optical chip usage will increase significantly, further expanding SAM.

If Lumentum were to take a standalone share of 15%-20%, this would translate to about $1.2 billion to $2.4 billion in market revenue this year. Adding CPO ray engine and subsystem revenue, the potential could reach $3 billion to $4 billion. On a three-year horizon, a conservative scenario (AI capital expenditure growth dropping to 20%, CPO penetration rate only 15%) corresponds to about $3.5 billion in 2028 revenue; Neutral scenario is about $5 billion; The optimistic scenario could reach $6 to $7 billion.

The gap between these three numbers is nearly double—the height of the ceiling has never been the issue; the question is which price the market is pricing today.

Every card held by the bulls is backed by hard data

For those bullish on this company, there's no need to tell stories—just lay out the financial reports and announcements:

▲ Bull Case

The only question for the bears is this: are these good news already priced in?

Bearish arguments also don't rely on speculation; almost all come from the other half of the same financial report:

▼ Bear Case

Eight days later, all the debates will be repriced by a financial report

In the coming period, LITE's stock price rhythm will be dominated by one main event, with the rest just background noise:

•	After market close on August 11: Fourth quarter and full-year results, company guidance revenue is $960 million to $1.01 billion, non-GAAP earnings per share $2.85 to $3.05, with short-term success or failure determined by direction
•	Mid-2026: NVIDIA Vera Rubin platform shipments, CPO light engines entering a surge in volume, positive news
•	Second half of 2026: Demand for 1.6T optical transceivers will a surge in volume, with 800G and above products expected to account for over 60%, which is positive
•	2026 to 2030: NVIDIA's $2 billion investment will fulfill its procurement commitments and capacity priorities year by year, providing revenue visibility
•	In the next one to two fiscal years: outsourcing capacity in Southeast Asia, expanding manufacturing scale, and diversifying geopolitical risks, with a neutral bias
•	By 2030: CPO penetration rate will climb from low single digits to 35%, determining the long-term slope of optical chip demand

The market first made a mistake in pessimism, and then in optimism

Rewinding the timeline by two years, this company has fully staged the "market misjudgment twice."

The last fiscal year was the bottom of the cycle

Full-year operating loss was $180.1 million, with net losses of $82.4 million and $60.9 million for two quarters respectively, with a loss per share reaching $1.21 at one point. At that time, pessimism in the optical communications industry was so intense that almost no one believed that demand for AI data centers would be transmitted so quickly to optical chip orders.

The turning point came after revenue began to recover

Quarterly revenue climbed from $533.8 million to $665.5 million, and operating profit rebounded from $6.7 million to $64.3 million. However, the market still doubts the sustainability of the recovery, and stock price gains are limited.

The real main rally came from NVIDIA's $4 billion investment announcement in March this year, confirmation of CPO mass production, and a single-quarter revenue of $808.4 million, far exceeding expectations—12-month returns pushed above 600%, with the stock price standing above the thousand-yuan mark midway.

The retrospective review was very clear

At the low point, the market underestimated the impact of AI infrastructure on optical interconnects, and at the peak, it discounted the most optimistic long-term assumption into the current stock price. Both extremes deviate from fundamental anchors. Today's 9.24% bullish candlestick came after a 12.8% pullback from the high—whether it is the starting point of a new round of rally or another rebound within a wide range of volatility remains to be seen.

⚠️ Risk Notice • Valuations are at historical extreme levels; if growth falls short of expectations, the drawdown could reach 30%-50% • Debt due within one year surged to $3.2386 billion, combined with a 39% share dilution, resulting in both leverage and dilution • Export controls, U.S.-China relations, and tariff escalation have been explicitly listed as major uncertainties by the 10-K • Inventory increased by 35% and accounts receivable by 77%, with downstream slowdowns bringing both impairment and payment collection pressures • Rising and falling in tandem with Coherent and AOI, individual stock volatility is amplified by overall sector sentiment

🟡 Neutral The growth is real, but the price already accounts for the best scripts.

💬 Discussion Fluctuating at high levels, better to chase rallies than wait until August 11. Live buy and sell positions can be posted on the official WeChat account for LITE.

Data source

•	Source: NextPick real-time snapshot + company quarterly earnings report and guidance + SEC filing documents + public data from mainstream financial media and industry research institutions, reconciliation date after the U.S. stock market closes on August 3, 2026.