Lumentum: The Contradiction Hidden in $40

Sources: Lumentum FY2026 Q4 results (released Aug 11, 2026); management commentary at the Citi Global TMT Conference, Sep 9, 2026.

01Company overview

Lumentum Holdings Inc. (Nasdaq: LITE) is headquartered in San Jose, California, and is a leading global supplier of optical components and photonic technology serving AI data centres, cloud computing and communications networks. Its fiscal year ends on the last Saturday of June — FY2026 runs from late June 2025 through June 27, 2026.

Components: EML laser chips (100G / 200G), continuous-wave (CW) lasers, pump lasers, narrow-linewidth laser assemblies, VCSELs (3D sensing). Systems: 800G / 1.6T cloud optical modules, OCS optical circuit switches, ROADM / WSS, industrial lasers. Source: company FY2026 Q4 results and earnings call (Aug 11, 2026).

02How the share price has moved this year

As of the Sep 25, 2026 close of $941.65, the stock is up 155.47% year to date, measured against the 2025-12-31 close of $368.59. The intraday low for the year was 317.44 (Jan 16) and the intraday high 1,085.68 (May 12); the highest close came on May 11 at $1,053.09.

Three re-ratings in one year, with a 43% drawdown in between

Flatten the curve and what you see is not a smooth uptrend but three event-driven re-ratings in a single year, with a 43% drawdown sandwiched in between. Much of the story is already in the price — that is the premise for everything that follows.

03The optical bottleneck is moving closer to the compute chip

Bandwidth demand from Lumentum's end applications is growing far faster than compute itself. At the Citi 2026 Global TMT Conference, management set out the bandwidth multiple of each tier of the architecture relative to a telecom DCI baseline:

Underneath it all is a hard physical constraint: copper tops out at roughly 1.5–2 metres at 800G, and a standard rack is about 2 metres tall, meaning copper is essentially confined to within a single rack or to adjacent racks. At 1.6T, passive copper drops below 1 metre, and even with active electrical cables (AEC) it only reaches 2–3 metres. At 3.2T, copper transmission becomes largely unworkable.

The reported numbers bear this out: FY2026 Q4 revenue jumped 109.3% to $1.006bn, the first quarter in the company's history above $1bn. Components revenue was $649.4M, up 102.7% YoY, while Systems was $356.9M, up 122.6% — both triple-digit. More telling still, quarterly revenue had only just passed half of management's $2bn target when non-GAAP gross margin already reached 50.4% (47.4% on a GAAP basis).

That matters: Lumentum's profit growth is not purely a volume story. The CFO attributes the margin improvement to higher factory utilisation, a richer product mix and price increases on selected products; a tight laser supply further supports pricing, and management expects the shortage to persist into 2027 and even 2028, with prices "flat to above the levels already raised."

04OCS is the biggest variable behind the $40 EPS target

Here is the crux: the incremental driver behind management's $40 EPS target is not conventional optical modules but incremental OCS sales.

At the Citi Global TMT Conference on Sep 9, 2026, Vice President of Investor Relations Kathryn Ta said:

She described it as the company's first full-year earnings guidance issued with this level of confidence. The incremental volume comes almost entirely from higher shipments driven by the hardware success of the largest customer — in management's words, "we need to ship them more units." EPS is expected to accrue linearly through FY2028, rising every quarter rather than arriving in a single step.

Why does it matter? OCS lifts Lumentum a rung up the value chain. Ordinary optical components are transmission hardware; an optical circuit switch can dynamically schedule an AI cluster and re-partition compute infrastructure. Management has flagged two new use cases:

• Cluster reconfiguration — dynamic re-sizing of XPU clusters beyond the standard 576-accelerator pod;
• Resiliency routing — using OCS to steer traffic around overloaded or underperforming accelerators.

Management said the $8bn OCS total addressable market (TAM) it gave only six months ago is already stale and conservative, and plans to put out a materially higher figure at OFC in March 2027. As a first-order approximation, management put the accelerator-to-OCS-port ratio at about 1.5 : 1.

The second new growth vector is external light sources (ELS). Lumentum has won its first ELS module order, with deliveries due to begin in the second half of 2027. Management says the gross margin here is above the company average, though still below the laser business.

For Lumentum, CPO and NPO are not an either/or contest. Management expects the 2028 optical scale-up market to split roughly 50% CPO / 50% NPO by unit volume, and says Lumentum expects to cover 100% of CPO external light sources and about half of NPO. CPO inherently needs an external light source, and thermal-management challenges will push a meaningful share of NPO designs toward external lasers — in management's words, "having the laser that close to the heat source is very concerning," a worry that is steering more customers toward external-light-source architectures.

With Qualcomm and Corning, Lumentum also completed a board-level demonstration of high-density 1060nm VCSEL die-to-die (D2D) interconnect at ECOC 2026 in Málaga, Spain: 32 Gb/s NRZ optical lanes, about 10 Tb/s of aggregate transmit/receive capacity per optical engine, currently around 1 Tb/s/mm of edge bandwidth density against a target of about 4 Tb/s/mm. The near-term contribution is limited, but it opens up longer-term addressable-market optionality.

05The moat is a long production-qualification cycle

Competition should not be underestimated. On the critical high-power laser track, management's view is:

Coherent is also pursuing a liquid-crystal route to OCS. One layer down, in modules and DSP, Marvell and Credo are pushing into optical DSP, silicon photonics and 1.6T interconnect. All of this corroborates the strength of the optical cycle, but it also shows that Lumentum does not have a platform barrier that no one else can replicate.

Lumentum's real edge is a time-based barrier on the manufacturing side. High-performance lasers require lengthy qualification, specialised processes and yield refinement. Customers will not switch to a second source simply because a rival quotes a lower price. That is why the company keeps expanding in San Jose, the United Kingdom and Greensboro.

The risk: today's shortage can turn into tomorrow's overcapacity. FY2026 capex was $451.3M, of which $167M was spent in Q4 on AI/cloud-related capacity. Expanding while demand exceeds supply is fine; the real test comes once competitors' capacity lands and free cash flow has to carry the load — FY2026 operating cash flow was $751.4M and free cash flow $300.1M, with capex already absorbing close to 60% of operating cash flow.

That is why Lumentum's multi-year contracts matter. Management noted that most of the long-term agreements it has struck with major network equipment makers run for three years and already contain pricing terms; pump laser share is around 70%–80% (management's own "rough estimate"), with shipments expected to grow fourfold over the next five quarters. Long-term contracts can hedge expansion risk, but they cannot hold off the eventual normalisation of pricing.

06The $40 EPS target means the valuation has to be re-run

LITE trades at $941.65 (Sep 25, 2026 close; the Sep 24 close was $929.04). The valuation base needs to be re-run, because the consensus figures used previously carried a fiscal-year labelling error. Valuation by scenario, with P/E computed on $941.65 in every case:

So does the "27x P/E × $40 = a $1,080 price target, 16% upside" argument still hold? The arithmetic works, and the market's frame of reference is more complete than a single point estimate.

Putting these lines together: at the $941.65 spot price, management's $40 FY2028 target implies 23.5x P/E, while the $1,148.76 street average target implies roughly +22% upside. Consensus is broadly aligned on direction; the real debate is not whether $40 is achievable but what multiple the market will pay for that level of earnings.

So after this run, LITE is not cheap. Consensus estimates have yet to catch up with management's raised guidance, while the share price has already moved. As long as Lumentum can genuinely deliver $40 of EPS and hold gross margin above 50%, the earnings-growth story can continue.

07When supply scarcity becomes the risk

Is today's high profitability the product of a strong business model, or a windfall from a short-term supply-demand squeeze? Laser shortages are currently holding up prices and margins, but Lumentum and rivals like Coherent are all expanding aggressively, and a shortage can flip into overcapacity at any time. Separately, OCS is tightly bound to a single anchor customer, making the $40 EPS target heavily dependent on one client — a pronounced concentration risk.

Cloud optical module shipments are still predominantly 800G; as 1.6T scales and later generations follow, price normalisation and commoditisation pressure are all but inevitable over the longer run — something to track as a medium-term risk rather than present as an accomplished fact.

Finally, the company is front-loading capex: FY2026 capex of $451.3M against FY2026 free cash flow of $300.1M, with the Greensboro fab not producing until around mid-2028, leaving a near-two-year capacity gap. If scale-up or CPO adoption slows, gross margin, capacity utilisation and free cash flow would come under pressure together, and the downside for the share price should not be underestimated — this July's 43% month-scale drawdown is a ready-made example.

08Conclusion

Lumentum's new technologies are growing faster than its legacy optical component business. AI is embedding optical products deeper into the data centre, and OCS, high-power lasers and external light sources are opening entirely new revenue curves.

At today's price, the stock is a simultaneous bet on three things:

1. the OCS shipment cadence at its largest customer;
2. the sustainability of gross margin above 50%;
3. the multiple the market is willing to pay for that level of earnings.

If any one of the three goes off track, the leverage works in reverse. That is the real risk contradiction inside the "$40 EPS" number.