Lumentum's $6.9B FY2026 GAAP loss is a non-cash convertible-note charge — Q4 non-GAAP net income was +$326.3M. The laser lead over $COHR, 53% customer concentration, and the 23x sales question.
More Macroplane supply-chain guides · Baskets · Investment theses · Macro trends
Lumentum ($LITE) reported a $6.9B GAAP net loss for fiscal 2026 on $3.01B of revenue, and almost none of it is real. $7.8B of that came from a one-time, non-cash loss on debt extinguishment when the company equitised its convertible notes. Strip it out and the fourth quarter produced $326.3M of non-GAAP net income on record revenue. Every screener shows a company that lost ninety-three dollars a share; the underlying business nearly doubled and the stock is up 946% in five years.
👉 See it in context: $LITE sits in the Photonics / CPO basket on Macroplane, alongside the Silicon Photonics & Optical Interconnects trend.
This is the single most misread number in optical components right now, so it is worth being precise.
Convertible notes are debt that can turn into equity. When the share price runs far above the conversion price — and $LITE went from roughly $80 to $867 — holders convert. Accounting rules make the issuer recognise the difference between the carrying value of the debt and the fair value of the shares handed over as a loss on extinguishment. The share price rising is what creates the loss.
| Fiscal 2026 | Figure |
|---|---|
| Revenue | $3.01B (from $1.65B) |
| Gross profit | $1.26B (41.7% margin) |
| GAAP net loss | −$6.94B (−$92.96/diluted share) |
| Of which: non-cash debt extinguishment | −$7.8B |
| Q4 revenue | $1.01B |
| Q4 non-GAAP net income | +$326.3M (+$3.23/diluted share) |
| Cash and equivalents | $2.04B |
No cash left the building. The debt is gone from the balance sheet, the share count is higher, and the company holds $2.04B in cash. A reader who sees "−$92.96 EPS" and moves on has learned the opposite of what happened.
The dilution is the real cost, and it is not trivial — but it is a different thing from an operating loss, and it is the kind of distinction a quote page will never make for you.
Two segments, and the mix has changed completely since the fiscal 2024 trough.
Cloud & Networking is the AI business: EML and continuous-wave laser chips, optical transceivers, ROADMs and wavelength-selective switches. The 2023 CloudLight acquisition bought its way into hyperscaler transceivers, which is how NVIDIA and Microsoft became customers.
Industrial Tech is fiber and solid-state lasers for manufacturing, plus the VCSEL arrays behind Apple's FaceID.
The five-year arc shows how violent the swing has been:
| Fiscal year | Revenue | Gross profit | Net income |
|---|---|---|---|
| FY2022 | $1.71B | $788.6M | $198.9M |
| FY2023 | $1.77B | $569.0M | −$131.6M |
| FY2024 | $1.36B | $251.5M | −$546.5M |
| FY2025 | $1.65B | $459.9M | $25.9M |
| FY2026 | $3.01B | $1.26B | −$6.94B (see above) |
Gross profit went from $251.5M to $1.26B in two years. That is the number to watch, because it is unaffected by the convertible-note accounting.
Here is where Lumentum's bull case gets specific, and where it differs from Coherent's.
Co-packaged optics does not put a whole transceiver next to the switch chip. It needs an external laser source — a separate, very clean, very efficient light source feeding the optical engine over fibre. That is a different product from a transceiver, with different economics, and it is the piece Lumentum has optimised hardest.
The most detailed public case for Lumentum's lead comes from the semiconductor analyst writing as Irrational Analysis, who argues that nobody else publishes competitive linewidth specifications for high-power (300mW+) O-band DFB lasers. Their August 2026 earnings notes put numbers on it:
On Coherent, the same author is blunt: poor CPO laser performance, yield problems on 6-inch InP wafers, and a position selling isolators and fibre-array units at 30–40% margin rather than the laser itself. They also credit Coherent with excellent VCSELs, "second only to Broadcom."
Weigh this carefully. The author discloses Lumentum as their number-two position and notes the 17% short interest with evident relish. This is a well-argued, technically literate, and openly partisan case, not a neutral appraisal. The verifiable part is that the specifications are public and the margin structure is consistent with what a scarce, hard-to-make component earns. The unverifiable part is whether Coherent's yield problems persist.
Our Coherent teardown makes the other side of the argument: $COHR has EUV, VCSEL and silicon-carbide businesses that would survive an optics downturn, which Lumentum largely does not.
Lumentum's filings disclose four customers above 10% of revenue in fiscal 2024:
| Customer | % of FY2024 net revenue |
|---|---|
| Alphabet ($GOOGL) | 18.9% |
| Apple ($AAPL) | 12.1% |
| Ciena ($CIEN) | 11.4% |
| Nokia ($NOK) | 10.5% |
Roughly 53% of revenue from four customers, and they are not correlated in a helpful way: Alphabet is TPU data centres, Apple is FaceID VCSELs in a mature phone cycle, and Ciena and Nokia are telecom capex. A single hyperscaler pausing, or Apple dual-sourcing, moves the whole P&L.
NVIDIA has since joined that list from a standing start. In March 2026 it invested $2B in Lumentum alongside $2B in Coherent, with a multi-year agreement carrying a multi-billion-dollar purchase commitment and future capacity rights. That deepens the concentration at the same time as it validates the technology.
Macroplane tracks 76 supply-chain relationships for $LITE. A few that matter and are easy to miss:
The full graph is on the Lumentum company page.
| $LITE | $COHR | |
|---|---|---|
| Market cap | ~$68.9B | ~$71.0B |
| FY2026 revenue | $3.01B | $7.12B |
| Price / sales | ~23x | ~10x |
| FY2026 gross margin | 41.7% | ~38.5% (Q4 GAAP) |
| Non-optics fallback | Industrial lasers, VCSELs | EUV lasers, SiC, VCSELs, industrial |
| NVIDIA stake | $2B | $2B |
Two companies with near-identical market caps, where one has 2.4x the revenue of the other. The market is paying more than twice as much per dollar of Lumentum's sales.
That gap is the thesis, whichever way you take it. If Lumentum's laser lead is real and the external light source is where co-packaged optics concentrates margin, the premium is a forward multiple on a much higher-margin mix. If the lead is narrower than its advocates believe, or if Coherent fixes its yield, $LITE is priced for a scarcity that erodes.
The fiscal 2026 GAAP net loss of $6.9B was driven by a one-time, non-cash loss on debt extinguishment of roughly $7.8B, recognised when Lumentum equitised convertible notes after its share price rose sharply. No cash was paid out. Fourth-quarter non-GAAP net income was $326.3M, and the company ended the year with $2.04B of cash.
LITE is the Nasdaq ticker for Lumentum Holdings Inc., an optical and photonic component maker based in San Jose, California. It was spun out of JDS Uniphase in 2015.
They are different bets. Lumentum is more concentrated in lasers and optical components, with a widely argued lead in high-power narrow-linewidth DFB lasers for co-packaged optics. Coherent is larger and more diversified, with EUV lithography lasers, silicon carbide and VCSEL businesses that do not depend on AI networking. Lumentum trades at roughly 23x sales against Coherent's 10x, so the market already prices the difference.
Fiscal 2024 filings disclose Alphabet at 18.9% of net revenue, Apple at 12.1%, Ciena at 11.4% and Nokia at 10.5% — about 53% from four customers. NVIDIA became a major customer and a $2B shareholder in March 2026.
On a non-GAAP basis, yes: $326.3M of net income in the fourth quarter of fiscal 2026. On a GAAP basis, fiscal 2026 shows a $6.9B loss because of the non-cash convertible-note charge. Gross profit, which the charge does not affect, rose from $251.5M in fiscal 2024 to $1.26B in fiscal 2026.
Co-packaged optics puts the optical engine on the switch package but keeps the laser outside it, feeding light in over fibre, because lasers degrade fast next to a hot ASIC. That external source has to be unusually clean and efficient, which makes it a scarcer, higher-margin product than a transceiver. It is the segment Lumentum has focused on.
This is research and education, not financial advice. This article cites a partisan third-party analyst who discloses a long position in $LITE; treat their competitive claims as argument, not fact. Do your own work.
The fiscal 2026 GAAP net loss of $6.9B was driven by a one-time, non-cash loss on debt extinguishment of roughly $7.8B, recognised when Lumentum equitised convertible notes after its share price rose sharply. No cash was paid out. Fourth-quarter non-GAAP net income was $326.3M, and the company ended the year with $2.04B of cash.
LITE is the Nasdaq ticker for Lumentum Holdings Inc., an optical and photonic component maker based in San Jose, California. It was spun out of JDS Uniphase in 2015.
They are different bets. Lumentum is more concentrated in lasers and optical components, with a widely argued lead in high-power narrow-linewidth DFB lasers for co-packaged optics. Coherent is larger and more diversified, with EUV lithography lasers, silicon carbide and VCSEL businesses that do not depend on AI networking. Lumentum trades at roughly 23x sales against Coherent's 10x, so the market already prices the difference.
Fiscal 2024 filings disclose Alphabet at 18.9% of net revenue, Apple at 12.1%, Ciena at 11.4% and Nokia at 10.5% — about 53% from four customers. NVIDIA became a major customer and a $2B shareholder in March 2026.
On a non-GAAP basis, yes: $326.3M of net income in the fourth quarter of fiscal 2026. On a GAAP basis, fiscal 2026 shows a $6.9B loss because of the non-cash convertible-note charge. Gross profit, which the charge does not affect, rose from $251.5M in fiscal 2024 to $1.26B in fiscal 2026.
Co-packaged optics puts the optical engine on the switch package but keeps the laser outside it, feeding light in over fibre, because lasers degrade fast next to a hot ASIC. That external source has to be unusually clean and efficient, which makes it a scarcer, higher-margin product than a transceiver. It is the segment Lumentum has focused on. This is research and education, not financial advice. This article cites a partisan third-party analyst who discloses a long position in $LITE; treat their competitive claims as argument, not fact. Do your own work.