Record Earnings, Collective Decline After Hours—What Exactly Happened to Photonics Stocks?

Over the past period, most investors have caught at least a portion of the photonics stocks rally. LITE has nearly doubled since the beginning of the year, COHR has climbed steadily, and stocks across the sector have risen almost across the board. If you held positions, you likely captured a decent wave of profits.

But this week, three earnings reports came out simultaneously, hitting the brakes on that momentum.

Lumentum posted quarterly revenue up 90% year-over-year to a historic high of $808 million, EPS beat expectations, and next-quarter guidance significantly exceeded Wall Street estimates—yet shares dropped about 5% after hours.

Coherent similarly achieved record quarterly revenue of $1.81 billion, with data center business experiencing rapid growth and next-quarter guidance midpoint notably surpassing analyst expectations—but shares fell about 7% after hours.

Fabrinet, which provides contract manufacturing for both companies, simultaneously set company revenue and EPS records—and also declined after the report.

Three companies, three earnings beats, three after-hours declines.

At times like this, investors typically have two reactions: one is "it's fine, short-term pullbacks are normal"; the other is "should I consider trimming positions?" Both reactions are reasonable, but the underlying logic differs. Below, I'll analyze in depth the questions everyone cares about most: why did the stocks fall despite such strong earnings? And is this sector finished climbing, or is there more to come?

Three Earnings Reports—The Numbers First

All three companies reported in the same week. Let's lay out the core numbers first; all subsequent analysis will refer back to these.

Lumentum (NASDAQ: LITE)

Coherent (NYSE: COHR)

Lumentum operates with a smaller base but aggressive growth rates, with profit margin expansion dramatic even by tech sector standards. Coherent operates at much larger scale with more moderate growth, but follows a clear gross margin improvement trajectory approaching management's repeatedly mentioned 40% long-term target. One is accelerating, the other climbing steadily.

Fabrinet (NYSE: FN)

Fabrinet's role differs from the first two. As a contract manufacturer for Lumentum, Coherent, Ciena and other major vendors, it doesn't bet on which company wins—think of it as a thermometer for the entire supply chain. Its numbers contain no fluff; this quarter's revenue and EPS both set company records, with management explicitly stating on the call that data center optical module demand continues to outstrip supply. This echoes Lumentum's mention of supply-demand gaps exceeding 30%, just stated from a different position in the chain.

Yet the market's reaction to all three reports was identical—all declined after hours.

Why?

Beats Yet Declines: Three Trading-Level Explanations

The most common online explanation is that expectations were already priced in—not wrong, but too vague. A more precise answer requires examining three more specific dimensions.

1. Classic Sell-the-News

In the month before earnings, Lumentum shares already experienced a clear rally, essentially pricing in the beat-expectations expectation ahead of time. This represents classic sell-the-news logic: when good news actually lands, early entrants choose to realize gains while new money begins to hesitate.

To determine whether this is merely event-driven profit-taking, typically observe 5–10 trading days post-earnings. If the stock recovers the after-hours decline, institutions may be using the dip to add positions; if it fails to recover, it may reflect deeper risk appetite weakening.

2. Revenue Slightly Below Expectations Triggered Programmatic Selling

Lumentum's quarterly revenue of $808.4 million came in slightly below Wall Street consensus expectations. The absolute dollar difference was minimal, but in algorithm-dominated after-hours trading, the "revenue miss" label itself triggers programmatic reactions, regardless of magnitude.

After Lumentum significantly beat both revenue and EPS expectations last quarter, sell-side analysts adjusted expectations progressively higher. At some point, the magnitude of beats begins diminishing or even turns to a minor miss—this transition itself represents a bearish signal in major banks' models, unrelated to fundamental business quality. Wall Street's scoring method has never been about how much you grew versus last year, but how much you exceeded my last prediction. This concept in sell-side research is called the "expectation gap," worth reexamining for any investor accustomed to focusing only on YoY growth.

3. Dilution Risk Remains Unresolved

Lumentum recently completed a Series A convertible preferred stock financing, with NVIDIA as the primary investor at approximately $2 billion scale. This financing dramatically increased the company's cash from approximately $1.16 billion to $3.17 billion—a clear strategic endorsement and liquidity improvement.

However, convertible preferred stock means at some future point, this batch of preferred shares will convert to common stock, directly diluting all existing shareholders' ownership percentage. Some convertible notes have already completed conversion. Until conversion fully settles, the dilution variable will persist in the market's pricing logic.

Wall Street Itself Lacks Consensus

After reviewing earnings, the sector may seem overwhelmingly positive, but current sell-side target price distribution for Lumentum reveals some nuances:

There's broad consensus on the sector's growth direction—disagreement centers on magnitude: how much of the future does current pricing already discount?

The bull case roughly goes: 1.6T transceiver upgrade demand will continue through 2027–2028; large-scale CPO deployment will unlock new incremental capacity; persistent supply shortages give suppliers pricing power. Fundamentals currently support this thesis—the four major hyperscalers (Amazon, Google, Meta, Microsoft) project combined 2026 capex of approximately $610 billion, up ~77% year-over-year, with actual spending consistently exceeding analyst predictions for the past two years.

The bear concern isn't that capex will contract—the real worry is a more long-term issue: this growth rate cannot sustain forever, and at some point must decelerate, with market reactions typically nonlinear. Combined with already stretched valuations, any quarter falling short of perfection could get amplified.

But one detail merits attention: post-earnings, all five major banks raised their target prices, including UBS and Morgan Stanley which maintain Neutral ratings, now at $960 and $900 respectively. Target price range from $900 to $1,300 still exceeds 40%—consensus on growth exists, but no consensus on valuation. Yet the collective upward revision itself represents a clear bullish signal.

The Biggest Engine Hasn't Ignited Yet

Seeing target prices from $900 to $1,300, many immediately think: after such gains, are the banks chasing highs? But if you carefully review the earnings call content, you'll discover one thing: this growth cycle may not even have completed its first half.

Lumentum CEO Michael Hurlston made several statements worth careful review.

First, the supply side. EML laser chips, pump lasers, narrow linewidth lasers—Hurlston's exact words: "effectively sold out for the foreseeable future". Supply-demand gaps expanded from last quarter's 25–30% to exceeding 30%, with the company already allocating customers—not all buyers can purchase what they want. This isn't a normal market-clearing signal but a structural shortage: demand growth already exceeds the company's current physical expansion limits.

More critically, the growth driver. Hurlston's exact words: "Our largest single growth driver, scale-up CPO, is still very much in its infancy". The largest single growth driver, Scale-up CPO, remains in very early stages. Current contributions from OCS and Scale-out CPO, he described as "relatively modest," still quite small. In other words, driving this quarter's growth primarily remains traditional optical transceivers and pump laser demand—the CPO new curve has barely begun climbing.

Connecting these statements, the logic becomes clear: this quarter's +90% revenue growth occurred with the largest growth engine not yet ignited. Moreover, margin improvement similarly exceeded expectations—Non-GAAP gross margin improved 540 basis points quarter-over-quarter to 47.9%, Non-GAAP operating margin improved 700 basis points to 32.2%. This means Lumentum isn't just growing revenue, but simultaneously thickening margins, with scale effects materializing.

So where's the next bottleneck? The answer is capacity. The new Greensboro InP facility isn't expected to contribute full capacity until late 2027 to early 2028. Until then, the company can sell as much as it can produce, not as much as the market wants to buy—a rare seller's market. From now until full capacity release spans nearly two years, with the industry's supply-demand structure essentially locked in during this window.

Of course, Hurlston also stated something suspenseful: OCS represents the biggest tightrope the company currently walks. Overwhelming demand is positive, but any execution issues—whether yield, equipment delivery, or customer allocation priority judgment—could affect delivery rhythm. Growth ceiling is determined by capacity, but whether you smoothly reach that ceiling is determined by execution—both factors coexist.

Coherent's story differs from Lumentum but similarly remains incomplete. Coherent's core narrative isn't capacity, but gross margin.

CEO Jim Anderson made two points. First, 1.6T transceivers have higher gross margins than 800G, so the upgrade ramp benefits overall profit margins. Second, 6-inch InP wafers yield over 4x the chips per wafer compared to 3-inch at under half the cost—this is engineering-level improvement, not financial adjustment, and once implemented won't reverse. These two factors combined mean Coherent's margin improvement doesn't depend on market sentiment but is written into production processes.

Numerically, this quarter Coherent's Non-GAAP gross margin reached 39.6%, up 105 basis points year-over-year, just 0.4 percentage points from management's publicly stated 40%+ long-term target. Q4 guidance range is 39.0%–41.0%, meaning next quarter could formally breach the 40% threshold. For a company with quarterly revenue exceeding $1.8 billion, each 1 percentage point gross margin improvement represents over $18 million in additional gross profit per quarter—over $70 million annualized.

In simple comparison, Lumentum's story is capacity—how much capacity unlocks determines how much revenue; Coherent's story is structure—selling the same optical modules, unit prices are rising, costs falling, margins expanding. Different paths, but one commonality: core realization still lies ahead.

Beyond Optics, What Other Options Exist?

After covering these two companies' fundamentals, a natural question arises: if you feel concentrating on one stock carries too much risk or worry about mistiming—beyond LITE and COHR, are there other ways to participate in this track?

Optics isn't a single track. From upstream laser chips to downstream switches, the chain is long, with each stock occupying different positions, carrying different risks and elasticity. LITE is the most direct beneficiary of capacity bottlenecks with highest elasticity but also greatest customer concentration risk; COHR operates at larger scale with more diversification and relatively lower volatility; FN is an optical module packaging contractor, with Ciena and NVIDIA NVLink lasers as major customers—regardless of which designer wins, it remains in the chain, representing a sector-neutral participation method.

How you allocate depends on how much elasticity you want and how much single-stock concentration risk you can accept. This table doesn't provide answers, but helps you clarify what you're betting on.

So, Where Are We Now?

Many stocks in this track have already risen 10x, some even more. At this position, whether you can still buy becomes harder to answer than whether fundamentals are good.

Interestingly, market concerns about this sector have never been fundamentals—no one denies demand is real, with verifiable customer order backlogs extending to 2027–2028. The real divergence is a harder-to-quantify question: how far ahead of reality is this cycle's pricing running ?

Bulls will say CPO is just beginning its zero-to-one journey, with 2026 as the commercialization launch year—this curve hasn't even started climbing yet; CPO market scale is projected to grow from $70 million in 2024 to $8 billion in 2030, compound growth exceeding 120%. Current growth rates occurred with this engine not yet ignited.

Bears will say valuations already sit at historic highs, with trading crowding also reaching year-to-date peaks in April 2026. Any quarter with execution falling short of perfection will get amplified. This LITE post-earnings decline already issued a warning.

Both judgments aren't wrong. The distinction lies in which segment you're buying—are you waiting for the capacity story before CPO truly ramps, betting on sustained gross margin structural improvement, or wagering on the entire sector's beta? These three things carry completely different risk-return characteristics.

Before Broadcom's June 3 earnings, many sector disputes lack answers—CPO deployment rhythm, ASIC order scale—we'll see what Broadcom says then.

This content is for informational reference and learning purposes only and does not constitute any form of investment advice, securities recommendation, or buy/sell guidance. All earnings data comes from listed companies' official investor relations releases and public conference call transcripts. Markets contain uncertainty, past performance doesn't represent future results, investors should combine their own risk tolerance to make independent judgments and assume corresponding responsibility.