Optical stocks are breaking out on AI demand, with investors shifting focus from demand toward margins, capacity, and valuation sustainability

Overview

This report focuses on the rerating of the North American optical supply chain driven by AI data center demand. Morgan Stanley notes that optical stocks have risen another roughly 20% over the past month, and that post-OFC investor education and incremental capital inflows are still continuing, leaving few near-term negative catalysts that could disprove the bull case. At the same time, stocks have clearly broken out above traditional valuation ranges, and investors' latest questions are centered on GLW's optical margins, InP/EML capacity, reasonable valuation multiples, the OCS competitive landscape, CIEN's multi-rail product competition, and the long-term implications of 'narrow and fast' versus 'wide and slow' optical architectures.

Core views

The core view is that near-term momentum in the optical sector may continue, because AI capex data points have not yet cooled, many bull-case drivers will not be fully reflected until 2027/2028, and there is still uncertainty around the timing and yields of capacity expansions that could ease supply shortages. However, the report also emphasizes that current 20-25x 2028 bull-case EPS multiples are already very generous; especially if pricing competition emerges, Broadcom expands laser capacity, OCS competition intensifies, or GLW/COHR margin delivery falls short of expectations, valuations could retrace toward historical levels in the high teens.

Analysis framework

The report uses an investor Q&A framework to break the recent rerating of the optical sector into five categories of questions: demand, supply, margins, competition, and valuation. It first compares post-OFC investor interest and capital flows, then analyzes GLW fiber price pass-through, LITE/COHR/InP capacity expansion, CIEN multi-rail product competition, OCS solution choices, and the impact of different optical architectures in CPO on the supply chain, and finally reflects higher earnings expectations through model and target price increases.

Methodology notes

• Valuation frameworkScenario valuation and target price adjustment

  Estimate base, bull, and bear cases by applying target PE multiples to FY28e or CY28e earnings power.

  The target price increases for GLW, LITE, CIEN, and COHR are all based on higher earnings power around 2028 and AI-related valuation multiples above historical ranges.

• Industry supply-demand analysisCapacity shortage and expansion path analysis

  Track demand growth, capacity expansion, and yield uncertainty for InP, EML, CW laser, OCS, and CPO.

  The report notes that LITE has disclosed relatively clear EML expansion plans, and that COHR, Sumitomo, and Broadcom also show signs of expansion, but market share, product mix, and 6-inch transition yields still limit precise forecasting.

• Competitive landscape analysisOCS and multi-rail product competition assessment

  Compare the positioning of LITE/COHR, Cisco, Nokia, Eoptolink, and others in OCS, long haul DCI, 800ZR, and multi-rail.

  The report believes LITE's MEMS solution may be more favored by customers in the near term, but more OCS solutions may emerge over the long term; CIEN holds about 63% share in North American long haul DCI, but Nokia and Cisco multi-rail products may limit its pricing power.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• COHR.US / COHERENT CORP

  Beneficiary tied to the AI optical supply chain, OCS/CPO/transceivers, and InP/EML

  Strengths

  Highest investor interest; demand exceeds supply; target price raised to $290; bull case reflects datacomm share gains.

  Weaknesses

  Further upside requires improved execution, and valuation is already above historical averages.

  Comparison

  Relative to LITE/CIEN, the report believes COHR still has more room if execution improves.

  Risks

  AI ecosystem valuation reversal, EML constraints, and weaker pricing power due to capacity expansion and competition.

• CIEN.US / CIENA CORP

  Exposure to long haul DCI and multi-rail optical communications equipment

  Strengths

  About 63% share in North American long haul DCI; multi-rail products may offer a 50-70% price advantage; target price raised to $405.

  Weaknesses

  More of a challenger in pluggables; valuation is already relatively full.

  Comparison

  Its leadership position is stronger than most competitors, but if Nokia and Cisco launch multi-rail products as planned, CIEN's pricing power would be challenged.

  Risks

  Constrained telecom capex, earlier-than-expected price competition in 800ZR/multi-rail, and intensifying product competition from Nokia/Cisco.

• GLW.N / Corning Inc

  Exposure to fiber and improving optical margins

  Strengths

  Long-term beneficiary of fiber deployment, AI data center demand, and higher-value optical innovation; target price raised to $140.

  Weaknesses

  Near-term investor expectations for margins and price pass-through are high, and the report remains cautious on pass-through from Asian fiber price increases into the US market.

  Comparison

  The report favors GLW's long-term demand, but is more cautious ahead of earnings relative to COHR.

  Risks

  Slower AI capex, tariffs and macro uncertainty affecting service-provider and consumer-related businesses, and pricing improvement falling short of market expectations.

• LITE.O / Lumentum Holdings Inc

  Exposure to InP/EML, OCS, CPO, and the AI data center optical supply chain

  Strengths

  Currently in a favorable position with tight supply and pricing power; management says sold out conditions can last through 2028; target price raised to $710.

  Weaknesses

  Valuation is relatively full, and the bull case depends on smooth OCS/CPO ramp with no significant supply disruptions.

  Comparison

  For near-term OCS solutions, the report leans toward customers favoring LITE's MEMS based solution.

  Risks

  A shift in AI ecosystem valuation, changes in EML constraints, and easing bottlenecks from capacity expansion by Broadcom and others, weakening pricing conditions.

Key data

• Recent performance of optical stocksUp about 20% over the past monthThe report says optical stocks over L1M up another ~20%.
• Investor preference rankingCOHR, GLW, LITE, CIENThe report says even the least favored names still have relatively high interest, but LITE/CIEN are more fully valued.
• GLW target price$140, raised from $127Equal-weight maintained, based on FY28e Base Case EPS of $4.66 and a roughly 30x multiple.
• LITE target price$710, raised from $595Equal-weight maintained, based on FY28e earnings power of $23.70 and a 30x multiple; bull case is $1,000.
• CIEN target price$405, raised from $280Equal-weight maintained, based on FY28e EPS of $12.60 and a roughly 32x multiple.
• COHR target price$290, raised from $250Equal-weight maintained, based on FY28e earnings power of about $9.60 and a 30x multiple.
• Current optical valuations20-25x 2028e bull-case EPS; some current NTM PE at 49x-79xCharts show that AAOI, CIEN, LITE, COHR, and FN all currently trade at NTM PE well above historical averages.
• CIEN North America long haul DCI shareAbout 63%Third-party data shows CIEN remains the clear leader.
• LITE EML capacity expansionBy the end of CY26, EML unit capacity will increase another 50%+ versus the end of CY25Charts show FY20-FY25 EML shipments have already expanded by >8x.
• InP optical lane demandAbout 85% CAGRLumentum's chart projects AI data center InP optical lane volume demand for EML, CW, and UHP lasers at about 85% CAGR.
• Broadcom capacity signal4-6x capacity increase over the next yearThe report says timing, details, and yields remain uncertain.
• Fiber price signalAsian carrier fiber prices up 75%The report believes this pricing dislocation in Asia may not directly represent US pricing.

Impact & implications

In terms of investment implications, the report acknowledges that AI data center demand and tight capacity can still support near-term valuations for optical stocks, but emphasizes that risk/reward is now more dependent on execution and margin delivery. COHR is seen as having more upside if execution improves; GLW has favorable long-term demand, but elevated margin expectations ahead of earnings create near-term risk; LITE is in a strong position amid pricing power and tight supply, but valuation is already full; CIEN has a long haul DCI advantage, but multi-rail product competition may constrain pricing.

Risks

• Slower AI capex or a reversal in AI ecosystem valuations.
• Optical stock valuations are already far above historical averages; if competition increases or supply normalizes, multiples may compress.
• Margin improvement, price repricing, or fiber price pass-through at GLW and COHR may fall short of expectations.
• Laser capacity expansion by Broadcom and others could bring supply and demand closer to balance.
• The OCS competitive landscape may become more intense than the market currently expects.
• If multi-rail products from Nokia, Cisco, and others progress smoothly, they may limit CIEN's pricing power.
• Tariffs, macro uncertainty, and weakening demand from service providers/consumers.
• There is uncertainty around the timing of InP/EML capacity expansion, product mix, and 6-inch transition yields.

What to watch

• GLW earnings and investor day commentary on optical margins, Project Springboard, and price pass-through.
• Whether LITE continues to confirm sold out conditions, pricing power, and execution of 50%+ CY26 EML capacity growth.
• Execution improvement and share changes for COHR in OCS/CPO/transceivers demand.
• The rollout pace of CIEN multi-rail products, and whether Nokia and Cisco launch competing products as planned.
• The actual timing, scale, and yields of Broadcom's 4-6x capacity expansion over the next year.
• Whether AI capex data points continue to support nonlinear demand growth.
• The deployment pace of OCS at customers such as Google, Oracle, and Meta, and whether new entrants such as Eoptolink show large-scale supply responses.
• The technology path choice between 'narrow and fast' and 'wide and slow' architectures in CPO/AI clusters.