Skip to content
primarysourced Photonics sector Lumentum
LITE
~5 min read · 1,207 words ·updated 2026-04-29 · confidence 56%

Comparable transactions and valuation comps

Two questions for the LITE valuation work:

  1. What’s the right peer set for revenue/EBITDA/EBIT multiple comparison?
  2. Is there a precedent for the NVDA-direct $2B Series A Convertible Preferred placement that helps anchor the appropriate “strategic-supplier-with-direct-investor” valuation premium?

Peer set for multiple comps

The AI-photonics value chain peer-set, ranked by relevance to LITE:

PeerRevenue scale (approx most-recent year/run-rate)Pure-play AI-photonics?Comparable to LITE on
Coherent Corp (COHR)~$5.7–6.5B FY2025 run-rateMostly yes (Datacom & Comms = primary AI-photonics)Direct comp on InP EML duopoly + dual NVDA investment
MACOM Technology (MTSI)~$700–900MPartially (RF + photonic ICs)Adjacent component layer; not direct
Marvell (MRVL)~$5.5–6BPartial (DSP + silicon photonics)Different value-chain layer; competitor on transceiver references
Innolight (Shenzhen)$3–4B+ revenueYes (transceiver pure-play)Module-layer comp for Cloud Light segment
Eoptolink (Shenzhen)~$1B+Yes (transceiver pure-play)Module-layer comp
POET Technologies (POET)<$50MYes (passive optical engine for transceivers)Different architecture; LWLG-thesis-adjacent
NeoPhotonics (now part of LITE)Acquiredn/aPredecessor; integrated

The cleanest direct comp is Coherent Corp. Both companies:

  • Hold the merchant InP EML duopoly position
  • Received parallel $2B NVDA Series A Convertible Preferred placements in March 2026
  • Operate at the chip + module layers (vertically integrated)
  • Have Cloud & Networking / Datacom & Comms segments dominating consolidated revenue

The structural differences:

  • COHR has broader revenue base (~3.5x LITE) including industrial-laser, materials, networking
  • COHR grew Datacom +33.6% YoY in Q2 FY2026 vs. LITE +65%+ (LITE growing faster off smaller base)
  • LITE is more pure-play AI-photonics; COHR is the diversified version

Multiple framework

MetricLITE (Q2 FY2026 annualized × 1.3 forward)COHR (current consensus forward)
EV/Revenue (forward)very high (post-NVDA re-rating)meaningfully lower
Forward P/Ehighmid-range
EV/EBITDAhighmid-range

⚠ Specific multiples are highly sensitive to current spot prices and forward consensus estimates. As of April 29, 2026, with LITE at ~$862 and a >15× appreciation from the 12-month low, the multiple framework reflects extreme growth-premium pricing. Both LITE and COHR are trading at AI-photonics-supercycle multiples that exceed historical mean.

The LITE-vs-COHR multiple gap exists because:

  • LITE’s % AI-photonics revenue mix is higher (≥85% Cloud & Networking vs. COHR ~50% Datacom & Comms)
  • LITE’s revenue growth is faster (+65% YoY vs +33% YoY)
  • LITE’s margin expansion is more dramatic (operating-leverage step-function)
  • LITE’s forward EPS trajectory (toward $30 by 2028) implies a multiple-compression as growth moderates

This justifies a higher LITE multiple in absolute terms — but the multiple gap may compress as LITE scales and growth normalizes, and as COHR’s smaller-base AI-photonics segment grows faster than the consolidated.

⚠ There is no obvious one-to-one precedent for NVDA’s direct $2B Series A Convertible Preferred placement into a public InP optical-component supplier, with parallel investments in two duopoly competitors at the same scale and structure.

The closest analogues fall in three categories, each imperfect:

1. Strategic-supplier equity placements

PrecedentStructureLITE-NVDA fit
Microsoft–OpenAI ($10B+)Equity in private AI labDifferent (private investee, not public supplier)
Apple–Imagination Technologies (historical)Equity stake in GPU IP supplierDifferent scale, different tech
Amazon–Rivian ($700M+)Equity in EV manufacturerDifferent industry
Toyota–Subaru cross-shareholdingIndustrial cross-holdingDifferent model
Sony–TSMC JASM JVJV equityDifferent (JV)

None of these involves a customer placing equity in a public component supplier in the structure NVDA chose with LITE/COHR.

2. Strategic-investor stakes in semiconductor suppliers

PrecedentStructureLITE-NVDA fit
Various Japanese keiretsu cross-holdingsIndustrial cross-holdingCultural/historical; less applicable to US public-co context
Apple–GT Advanced TechnologiesPre-payment + capacity dedicationCloser in spirit (capacity-dedication-for-equity)
Apple–Finisar VCSEL pre-payment ($390M, 2017)Capacity pre-payment (not equity)Closer; Apple-Finisar is a partial precedent for the supply-commitment dynamic
Google–Qualcomm (historical strategic)VariousLess direct

The Apple-GT Advanced Technologies (2014) deal is the closest precedent in spirit: Apple pre-paid $578M to GT for sapphire-display capacity dedicated to Apple. The deal failed when GT couldn’t execute. Apple-Finisar VCSEL pre-payment in 2017 was structurally similar — capacity dedication via supplier-direct funding — but at smaller scale. Both are precedent for the operational logic of NVDA-LITE but not the financial-instrument structure (NVDA chose Series A Convertible Preferred equity, not pre-payment).

3. Convertible preferred placements at strategic-investor scale

The Series A Convertible Preferred instrument structure is more common in growth-stage private placements than public-company strategic transactions. The terms NVDA negotiated (parity conversion, no redemption, no preemption, voting on as-converted basis except director elections, no disclosed board seat right) are atypical for a strategic-investor placement of this scale — suggesting NVDA prioritized economic exposure and capacity allocation over governance influence.

Implications for LITE valuation

The lack of a clean comp for the NVDA-direct strategic-investment dynamic means the LITE re-rating is partly uncomped optionality value: how much equity-multiple should accrue to a company that has a $2B / multibillion-purchase-commitment / capacity-allocation arrangement with the dominant customer in the fastest-growing AI infrastructure layer?

The market’s answer (LITE up >15× over the 12-month window, currently ~$862) prices a meaningful premium for this dynamic. Reasonable framings:

  • Structural revenue floor — NVDA’s purchase commitment functions as a take-or-pay contract for the dedicated capacity, providing the equivalent of a multi-year-revenue forward sale
  • Margin protection — NVDA’s bilateral structure (parallel COHR investment) preserves duopoly economics and pricing power
  • Optionality on follow-on (CPO, NVLink CPO, future generations of advanced laser components)

These three premiums together support an above-historical-mean multiple but do not justify any specific number. The valuation work needs scenario-based DCF anchoring (see DCF assumptions).

What would compress the multiple

  1. Hyperscaler AI capex pause — would compress both the revenue trajectory and the duopoly pricing premium
  2. NVDA in-sourcing of optics — NVDA’s parallel investment in COHR is a reduce-this-risk signal but does not eliminate it
  3. CPO timeline acceleration that disrupts pluggable economics — Cloud Light segment most exposed
  4. Third merchant supplier emerges — Sumitomo, HG Genuine reaching Western datacom volume
  5. Macro / risk-off — high-multiple growth names compress fastest in tightening cycles

None of these is a near-term catalyst as of April 2026, but each is a tail-risk trigger that the bear case enumerates.

Sources