They couldn't ship enough to meet Datacom demand, yet gross margin fell 30bps to 12.1%. Same quarter, OpEx grew 6.2% on 39% revenue growth.
Thesis: Fabrinet is supply-constrained on demand, not demand-constrained on supply. The Datacom sequential decline is pent-up revenue waiting for upstream lasers, memory, and ASICs to normalize. Meanwhile they're building capacity at $130M for $3B revenue with a six-month operating profit payback. At 1.4% OpEx-to-revenue and 39% top-line growth, every incremental dollar flows through at high leverage. The 12% gross margin caps the multiple but not the earnings power.
Verdict: LONG — Conviction: MEDIUM
Catalyst: Building 10 first floor online June, second floor by September. Combined with supply constraint easing and hyperscaler Datacom programs ramping mid-FY2027, the capacity ceiling lifts in H2 calendar 2026.
Key Risk: Component supply constraints (lasers, memory, ASICs) persist through FY2027. New hyperscaler programs share the same constrained supply chain per Grady's own admission, so diversification doesn't solve the bottleneck.
The Tell: Grady volunteered the exact downside scenario unprompted: 'six months worth of operating profit would pay for the entire 2 million square feet' and 'the gross margin headwind would be about 50 basis points.' CEO voluntarily quantifying both upside and worst-case on capex decisions signals genuine conviction about the demand pipeline.
Friction Level: MODERATE_FRICTION — Both sides agree demand is real and supply is constrained. Bull reads capacity expansion + operating leverage as EPS engine. Bear reads 12% gross margin + negative FCF as a structural cap on monetization.
2026Q3 · May 4, 2026
3 of 5 readings agreed3 LONG · 2 HOLDRead the full call
Demand Exceeds Supply at 12% Gross Margin, Capacity Building for Double the Run Rate
Fabrinet is supply-constrained on demand, not demand-constrained on supply. The Datacom sequential decline is pent-up revenue waiting for upstream lasers, memory, and ASICs to normalize. Meanwhile they're building capacity at $130M for $3B revenue with a six-month operating profit payback. At 1.4% OpEx-to-revenue and 39% top-line growth, every incremental dollar flows through at high leverage. The 12% gross margin caps the multiple but not the earnings power.
Grady volunteered the exact downside scenario unprompted: 'six months worth of operating profit would pay for the entire 2 million square feet' and 'the gross margin headwind would be about 50 basis points.' CEO voluntarily quantifying both upside and worst-case on capex decisions signals genuine conviction about the demand pipeline.
Datacom declined 6% sequentially despite management stating it would have been 'a new record by a wide margin' without supply constraints. Lasers, memory, and ASICs are short. When upstream supply normalizes, deferred revenue converts.
Building 10 adds $3B capacity for $130M capex. Management quantified payback at six months of operating profit and downside risk at 50bps gross margin. Total capacity path from $4.8B to $8.5B with land for $11.5B.
Revenue $1.214B exceeded guidance top, non-GAAP EPS $3.72 beat guidance, 39% YoY growth accelerating. Beat occurred while the fastest-growing segment was supply-constrained.
Demand far exceeds supply yet gross margin declined 30bps to 12.1%. Contract manufacturer in Thailand captures fixed returns. FX and ramp costs are the explanations but the structural 12% margin is a ceiling on monetization.
Current capacity $4.8B against annualized run rate near $4.8B is effectively 100% utilization. Presented as momentum but it means no incremental revenue from existing footprint until Building 10 completes January 2027.
Management claims Datacom demand is 'exceptionally strong' and 'far exceeded what we were able to ship' while reported Datacom fell 6% sequentially. The claim is unverifiable against reported numbers.
+78
CONFIDENT"Grady volunteered capacity math unprompted: $130M capex, six-month operating profit payback, 50bps downside. Answered analyst questions with specific numbers, no deflection on Datacom sequential decline or supply constraints."
Full transcript with specific capacity numbers, segment-level revenue, exact supply constraint details, hyperscaler program confirmation with contract status. Limited by no forward margin trajectory guidance.
Building 10 first floor online June, second floor by September. Combined with supply constraint easing and hyperscaler Datacom programs ramping mid-FY2027, the capacity ceiling lifts in H2 calendar 2026.
Component supply constraints (lasers, memory, ASICs) persist through FY2027. New hyperscaler programs share the same constrained supply chain per Grady's own admission, so diversification doesn't solve the bottleneck.
“We expect total revenue to be in the range of $1.15 billion and $1.2 billion, representing approximately 35% year-over-year growth at the midpoint.”
“Revenue is above our guidance range at a record $1.214 billion, with year-over-year growth accelerating to an impressive 39%.”
Companies tend to guide low: the median company we cover reports +1.7% against its guidance midpoint, so a small beat is the norm. A transcript can drop part of a guided range, so read both quotes.
“With that backdrop, we expect total revenue to be in the range of $1.25 billion-$1.29 billion, representing year-over-year growth of approximately 40% at the midpoint.”
Changed less than 96% of calls
Wording compared with the Feb 2, 2026 call, ranked among 97 calls reporting on Q1 2026. A measure of how much the story moved, not a signal about the stock.
Figures are extracted from the transcripts and checked in code: every quote appears word for word. Language change is computed from word frequencies, with no model involved. Descriptive only: none of this feeds the verdict.
Beat-and-raise with a self-imposed ceiling. Datacom declined 6% sequentially solely because lasers, memory, and ASICs are short upstream, not because demand weakened. Grady states it would have been 'a new record by a wide margin.' Two hyperscaler Datacom programs are qualified, shipping, and contracted with ramp through mid-FY2027. Three CPO customers shipping revenue. DCI grew 90% YoY to $197M with all major players as customers. OpEx grew 6.2% against 39% revenue growth, operating income up 46%. Building 10 adds $3B capacity at $130M capex with six-month payback, taking capacity from $4.8B to $8.5B. The supply constraint is temporary and upstream. Every constraint resolves in their favor.
Gross margin is 12.1% despite demand far exceeding supply. A company with genuine pricing power expands margins under those conditions. Fabrinet cannot because it is a contract manufacturer with zero pricing power. Operating cash flow was $53M against $135M net income, 61% conversion. FCF was negative $11M. Cash is going into inventory and receivables for ramps they cannot yet monetize. Current capacity $4.8B against $4.8B run rate is 100% utilization with no room until January 2027. New hyperscaler programs share 'broadly similar supply chain' per Grady, so the same component constraints apply. Every forward catalyst sits in front of reported numbers. The business model captures fixed returns regardless of how hot the AI cycle runs.
Zero Pricing Power Despite Constraints
Demand far exceeds supply but gross margin fell 30bps to 12.1%. FX and ramp costs are cited but a provider with pricing power would offset. Customers set the price at every level.
Say/Do Gap
Management claims Datacom would have been 'a new record by a wide margin' without constraints while reported Datacom fell 6% sequentially. The claim cannot be verified against reported numbers.
High Utilization as Ceiling
100% utilization on $4.8B capacity against $4.8B run rate means no incremental revenue from existing footprint. Presented as operational momentum but it is a hard ceiling until Building 10 completes.
Capacity Ceiling
Managed from both directions: own capacity maxed at 100% and upstream components short. Both resolve but timing is outside management's control.
LONG. The asymmetry is in the supply constraint, not despite it. Datacom revenue that didn't ship will ship when upstream lasers and memory normalize. That is deferred revenue, not lost revenue. Operating leverage is the money printer: OpEx at 1.4% of revenue, growing 6.2% against 39% top-line growth. Building 10 at six-month payback is the cheapest capacity expansion in semis. The 12% gross margin caps the multiple but volume growth at this operating leverage drives EPS hard. Size into the constraint, not against it. Medium conviction because cash conversion at 61% of net income and negative FCF are real near-term headwinds.