Concerns & Risks -- 6/10

Lumentum's risk profile is dominated by execution/supply, not demand. China exposure is structurally low (<10% and falling as a share) and a stack of near-term catalysts is dated and visible. The offset is a high valuation bar -- the multi-year acceleration is fully in consensus (revenue +90% FY27 already modeled) -- plus customer concentration (~16% one customer) and supply/execution risk. On the primary EV/EBITDA metric LITE trades modestly below the AI-optical peer average. Weight: 15%

EV/EBITDA (FY27)

~31x

Peer avg ~34x

Below peers

P/E (FY27)

~46x

Peer ~35-40x

Above peers

China Exposure

~8-9%

Falling as a share

Low overhang

Consensus Bar

Heroic

Rev +90% FY27 modeled

No room to slip

Valuation -- Primary Metric EV/EBITDA (FY2027)

Metric	FY27 Basis	LITE	Peer Avg
EV/EBITDA	~$2.15B EBITDA (rev $5.67B x ~38% adj-EBITDA margin)	~31x	~34x
EV/Sales (cross-check)	$5.67B revenue	~11.8x	~12-15x
P/E (cross-check)	$18.53 EPS	~46x	~35-40x

Net = at-to-slightly-below peer. On the primary EV/EBITDA metric (~31x vs ~34x AI-optical peer average) LITE trades modestly below peers despite arguably the cleanest supply story; on EV/Sales it is roughly in line; on P/E it screens above peers (mix and share-count dilution from the NVIDIA-funded raise). Peer set: Coherent (~38x), Corning (~34x), Fabrinet (~32x). FY27 EBITDA is estimated by applying LITE's exit-quarter adjusted-EBITDA margin ($293.5M on $808.4M revenue, ~36% and still expanding) to consensus revenue.

China Exposure (low and declining as a share)

Period	China Rev	Total Rev	China %
FY26Q1 (CY25Q3)	$49.3M	$533.8M	9.2%
FY26Q2 (CY25Q4)	$54.6M	$665.5M	8.2%
FY26Q3 (CY26Q1)	$70.2M	$808.4M	8.7%

China is ~8-9% of sales and shrinking as a share -- the U.S. hyperscaler-driven AI/cloud mix (now 90% of revenue in FY26Q3, up from 65%) is diluting China exposure. Reported "Asia-Pacific" volume overstates true China end-demand because most is Thailand contract-manufacturing throughput, not Chinese end-customers. By the rubric this is the <10% China band, not the >10% overhang band.

Key catalysts

#	Catalyst	Detail
1	FY26Q4 Record Guide	Revenue $960M-$1.01B (midpoint $985M), op-margin 35-36% -- already set. Over half of the sequential growth is components.
2	1.6T Transceiver Ramp	Begins in FY26Q4, ~20% of modules using LITE's own CW lasers; structurally higher margin than 800G.
3	OCS Multi-Year PO	Multi-billion-dollar purchase agreement; backlog "well past $400M," majority shipping 2H-CY26, accelerating into 2027.
4	Scale-out / Scale-up CPO	Multi-hundred-million UHP-laser PO deliverable 1H-CY27; scale-up CPO first shipments late CY27 -- the largest single driver, "still in its infancy" and NOT in numbers.
5	Greensboro (5th InP Fab)	>$5B incremental revenue opportunity, online early 2028 (NOT in current numbers). Plus pump/narrow-linewidth scale-across surge (>30% supply gap).

Regulatory / political risk

#	Risk	Severity	Detail
1	Consensus Bar / Execution	HIGH	Street models rev ~+90% FY27 and ~+52% FY28 -- a vertical consensus leaving zero room for a supply slip. Priced for flawless execution.
2	Customer Concentration	MEDIUM	One customer ~16% of revenue; NVIDIA's ~$2B stake plus top-customer status raises a customer/equity-entanglement question.
3	Supply / Tariff / Export	MEDIUM	FY25 10-K discloses export-control exposure (prior Huawei cessation, BIS/DOJ subpoena history) and tariff risk on imported components that could pressure COGS.
4	Manufacturing Concentration	MEDIUM	Geographically concentrated (Japan InP fabs, Thailand back-end) -- a Japan/Thailand disruption is the live political/operational vector.
5	China End-Demand	LOW	China is ~8-9% of sales and falling; too small to be the primary channel for a demand shock.

Bull case

Lumentum is a >30%-share leader in EMLs and high-power CPO lasers inside the secular AI-optical buildout, with demand running >30% above supply, LTAs locking pricing and capacity through CY27, and NVIDIA as a direct equity holder. Margins are inflecting violently (op-margin +2,140 bps YoY in FY26Q3) and the biggest drivers (scale-up CPO, Greensboro, ELS modules) are NOT in estimates -- classic "management says more, street under-models the out-years." On the primary EV/EBITDA metric it trades below peers despite arguably the cleanest supply story.

Bear case

The street already models rev ~+90% FY27 and ~+52% FY28 -- a vertical consensus leaving zero room for a supply slip, and management itself flags OCS as a "tight rope" and pumps/UHP lasers as suddenly, severely constrained. P/E (~46x FY27) sits above peers, transceiver margins still trail peers, customer concentration is high (one customer ~16% of revenue), and any AI-capex air-pocket or export/tariff shock hits a name priced for flawless execution. The risk is not demand -- it is execution against an already-heroic bar.

Score rationale

Score of 6/10. Two of the three rubric inputs are favorable: China is comfortably <10% and falling as a share, and the catalyst slate is unusually dense and dated (1.6T, OCS multi-year PO, scale-out CPO 1H-CY27, scale-up CPO late-CY27, Greensboro 2028).

Why not higher: the multi-year acceleration is fully embedded in consensus (rev +90% FY27 already modeled), leaving a heroic bar with no room to slip (-1). P/E screens above peers (-0.5). Customer concentration (~16% one customer) and supply/execution risk against a name priced for perfection (-0.5). Regulatory/export and Japan/Thailand manufacturing concentration are real but secondary.

What prevents a lower score: on the primary EV/EBITDA metric (~31x) LITE sits modestly below the ~34x peer average, not expensive; China/regulatory overhang is genuinely low; and the catalyst density plus out-year optionality that is explicitly "not in numbers" provides real upside asymmetry.

Net: the dimension is held below an 8 by the heroic consensus bar and execution/supply/concentration risk rather than by a China or regulatory overhang.

Data sourced from Daloopa (company_id 120), company filings, and earnings transcripts.