They're guiding 7.3% sequential growth with their back end nearly full and two wafer fabs running below capacity until 2027.
Thesis: Diodes owns idle front-end capacity at Greenock and South Portland that will convert to incremental margin starting 2027-2028 with zero incremental capex. Back-end is near full, meaning the constraint is on the assembly side, not the wafer side, which supports pricing now while the fab utilization ramp creates an embedded margin expansion path. Channel inventory shrinking while POS grows proves this is structural demand, not channel stuffing.
Verdict: LONG — Conviction: MEDIUM
Catalyst: Q2 results confirming $435M midpoint with gross margin at 32.8% would prove the above-seasonal pattern holds. Greenock and South Portland utilization improvement in 2027 is the longer-dated earnings driver management explicitly flagged.
Key Risk: Back-end assembly near full capacity. Every incremental dollar after that requires qualification of new capacity or outsourcing. If demand stays strong through 2026, the back end caps upside before front-end fabs ramp.
The Tell: Emily Yang's response to Stein's AI sizing question: 'Sizing—so, I would say, beyond the AI servers where we have seen a lot of ramp already...' She deflected a direct numerical question with qualitative momentum language across three different end markets. Management can describe AI exposure in detail but will not put a number on it, which means the AI premium in the multiple is unsized risk.
Friction Level: MODERATE_FRICTION — Both sides agree the numbers are clean. The disagreement is whether the supply disruption tailwind and unquantified AI exposure make the current growth rate repeatable, or whether two underloaded fabs create a margin expansion path nobody is modeling.
2026Q1 · May 7, 2026
5 of 5 readings agreed5 LONGRead the full call
Two Fabs Idle, Back End Loaded, Street Still Modeling Peak
Diodes owns idle front-end capacity at Greenock and South Portland that will convert to incremental margin starting 2027-2028 with zero incremental capex. Back-end is near full, meaning the constraint is on the assembly side, not the wafer side, which supports pricing now while the fab utilization ramp creates an embedded margin expansion path. Channel inventory shrinking while POS grows proves this is structural demand, not channel stuffing.
Emily Yang's response to Stein's AI sizing question: 'Sizing—so, I would say, beyond the AI servers where we have seen a lot of ramp already...' She deflected a direct numerical question with qualitative momentum language across three different end markets. Management can describe AI exposure in detail but will not put a number on it, which means the AI premium in the multiple is unsized risk.
Sixth consecutive quarter of double-digit YoY growth. Q1 came in at high end of guidance with above-seasonal sequential growth of 3.5% vs typical -5%. Q2 guide of $435M at +7.3% seq is another above-seasonal raise. Non-GAAP EPS more than doubled YoY.
Gross margin up 70bps sequentially to 31.8%, guiding to 32.8% in Q2. Non-GAAP opex dropped from 31.1% to 25.6% of revenue YoY while revenue expanded 22%. Two underloaded wafer fabs add incremental margin without capex as utilization improves.
Channel inventory declined in both dollars and weeks while POS grew QoQ in a seasonally slow quarter. Automotive +32% YoY, industrial +31% YoY, computing +21% YoY. Management states no restocking behavior in distribution or customer base.
Analyst William Stein directly asked for AI revenue sizing. Emily Yang gave a three-part response across industrial, communications, and compute with zero numbers. AI narrative carries multiple without dollar contribution being disclosed.
Back end 'almost fully loaded at this moment.' Greenock and South Portland utilization improvement not expected until 2027-2028. Q2 guide requires 7.3% sequential growth against near-full assembly capacity.
Zero hedging language in entire call. No macro uncertainty, no headwinds, no qualification of durability despite six consecutive beat-and-raises and above-seasonal guidance twice in a row.
Gross margin improvement attributed to automotive/industrial mix rising 2pts to 44%. Pricing question answered with 'mix change.' Three-year target of 35% blended margin requires 320bps expansion from current 31.8% with no demonstrated pricing power.
+75
CONFIDENT"Gary Yu voluntarily published three-year targets ($2B revenue, $700M gross profit, $4+ EPS) and called Q1 a 'great first step.' Brett Whitmire gave precise guidance with above-seasonal sequential growth. No hedging language anywhere in the call. One evasive moment on AI sizing but otherwise direct."
Rich operational data: channel inventory, POS, inventory days, capacity commentary, pricing trends. Specific guidance with tight ranges. Deflection on AI revenue sizing is the main gap.
Q2 results confirming $435M midpoint with gross margin at 32.8% would prove the above-seasonal pattern holds. Greenock and South Portland utilization improvement in 2027 is the longer-dated earnings driver management explicitly flagged.
Back-end assembly near full capacity. Every incremental dollar after that requires qualification of new capacity or outsourcing. If demand stays strong through 2026, the back end caps upside before front-end fabs ramp.
“Now turning to our outlook, for the first quarter of 2026, we expect revenue to be approximately $395 million, plus or minus 3%.”
“Revenue for the first quarter 2026 was $405.5 million, an increase of 22.1% over $332.1 million in the first quarter of 2025 and up 3.5% compared to $391.6 million in the fourth quarter of 2025.”
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.
“For the second quarter, we expect revenue to be $435 million, plus or minus 3%.”
Changed less than 74% of calls
Wording compared with the Feb 10, 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.
Revenue $405.5M, +22% YoY, +3.5% seq against seasonal -5%. Sixth straight beat-and-raise quarter. Gross margin expanding 70-100bps per quarter on mix and utilization. Channel inventory shrinking while POS grows proves demand is real, not channel fill. OCF $64.3M against $15M GAAP NI means cash conversion validates earnings. Two idle wafer fabs at Greenock and South Portland become incremental margin by 2027-2028 with zero incremental capex. Back end loaded means pricing power holds now. Balance sheet clean: $409M cash vs $55M debt. Three-year targets of $2B revenue and $4 EPS are specific, not vague. Automotive +32%, industrial +31%, computing +21% YoY. Every focus area compounding.
AI revenue sizing: zero numbers given when directly asked. Back end almost fully loaded while two wafer fabs idle until 2027-2028. Supply disruption tailwind acknowledged but unquantified, and it has an expiration date. Computing segment declined 3.7% QoQ on memory shortage. Pricing 'stabilized' but attributed to mix change, not demonstrated pricing power. Getting from 31.8% to 35% target margin requires 320bps with no pricing power evidence. Zero hedging language after six straight beats. The 35% margin target implies $112M more gross profit on $600M more revenue, and management hasn't shown how mix gets them there.
AI Label Without Substance
Stein asked directly for AI revenue sizing. Yang produced zero numbers, deflecting across industrial PSUs, networking, and compute. AI narrative carries multiple but dollar contribution remains undisclosed.
Capacity Ceiling
Back end 'almost fully loaded.' Two wafer fabs won't improve utilization until 2027-2028. Q2 guide demands 7.3% sequential growth on near-full assembly capacity.
Unsustainable Trend Confidence
Zero hedging language across entire call. No concession of cyclicality after six consecutive beat-and-raises. Management hasn't been tested by a downturn this cycle.
Mix Shift Spin
Margin improvement attributed to 2pt mix shift to auto/industrial. 35% target margin needs 320bps from current level. No demonstrated pricing power to bridge the gap.
LONG, medium conviction. The asymmetry is in two idle wafer fabs that become margin without capex starting 2027-2028. Street is pricing Diodes as a mid-cycle analog recovery. Channel inventory shrinking while POS grows, OCF at 4x GAAP earnings, and pricing stabilized across all end markets say structural demand. Back-end full supports near-term pricing. Size it for the utilization ramp, not the AI narrative. If AI sizing ever gets disclosed, that reframes the multiple. Cut if channel inventory turns up or Q2 comes in below $435M.