They lifted full-year growth guidance from low-to-mid 20% to low-to-mid 30% while stocking 145 days of inventory at 2.5x turns, and told customers aren't pre-building.
Thesis: AE sits at the etch/dep intensity and rack power content bottleneck of the leading-edge cycle, and pricing power is proven: gross margin ex-tariff 40.7% beat guidance while input costs rose and mature-line price increases passed through cleanly. The market is pricing hyperscaler concentration as a fragility when the real asymmetry is that semi H2 is guided +50% y/y on leading-edge logic/NAND/DRAM intensity while the street still models AE at WFE growth. Earnings quality is the only legitimate discount: $2.74 includes $0.04 tariff refund and a 14% tax rate, so normalize to ~$2.58, and FCF of $36M in a record quarter trails $112M adjusted net income because of the deliberate inventory build.
Verdict: LONG — Conviction: MEDIUM
Catalyst: Q3 print at $640M +/-20M with gross margin 41-41.5% ex any tariff benefit, plus Thailand first production revenue in Q4, proving margin expansion runs on mix and capacity is converting before 2027.
Key Risk: Hyperscaler capex guide breaks while 145-day inventory and $5B of capacity are already committed, and second-wave revenue does not arrive until 2028-2029 on management's own timeline.
The Tell: Asked by Mehdi Hosseini why semi revenue decelerates Q3 to Q4 when customers face an H1 2027 shipment step-up and should be pre-building, Oldham answered: 'it depends kind of how you attribute Q3 and Q4... we're shipping in line to demand.' He deflected the timing question instead of quantifying the split.
Friction Level: MODERATE_FRICTION — Both sides agree the quarter is real with no accounting games. Disagreement is whether the 10-day inventory build plus $5B capacity target against a $2.6B run rate is a Backlog Fortress ahead of a real second-wave ramp, or committed fixed cost riding one hyperscaler capex cycle until 2028-2029.
2026Q2 · Aug 3, 2026
4 of 5 readings agreed4 LONG · 1 HOLDRead the full call
Guidance Raised 10 Points, Inventory Raised 10 Days
AE sits at the etch/dep intensity and rack power content bottleneck of the leading-edge cycle, and pricing power is proven: gross margin ex-tariff 40.7% beat guidance while input costs rose and mature-line price increases passed through cleanly. The market is pricing hyperscaler concentration as a fragility when the real asymmetry is that semi H2 is guided +50% y/y on leading-edge logic/NAND/DRAM intensity while the street still models AE at WFE growth. Earnings quality is the only legitimate discount: $2.74 includes $0.04 tariff refund and a 14% tax rate, so normalize to ~$2.58, and FCF of $36M in a record quarter trails $112M adjusted net income because of the deliberate inventory build.
Asked by Mehdi Hosseini why semi revenue decelerates Q3 to Q4 when customers face an H1 2027 shipment step-up and should be pre-building, Oldham answered: 'it depends kind of how you attribute Q3 and Q4... we're shipping in line to demand.' He deflected the timing question instead of quantifying the split.
Revenue $574M beat the high end of guidance, EPS $2.74 vs $1.50 prior year, and full-year growth was reset from low-to-mid 20% to low-to-mid 30% in one print with data center raised from mid-30% to at least 50%.
Management raised prices on mature lines to offset input costs while gross margin still expanded 380 bps y/y to 41.9% (40.7% ex-tariff), with line of sight to over 43% and Q4 in the 42% range. Price-cost neutral goal holding while mix drives margin up.
Capex raised to $180-195M for 2026 with $50M deployed in Q2, Thailand accelerated to first revenue in Q4, and full network buildout targeting $5B of revenue-generating capacity against a $2.6B current run rate.
AE inventory up 10 days to 145 days at 2.5x turns while CFO says customers are not building stock. Oldham: 'we're shipping in line to their demand... delivering what they're asking.' AE builds, customer builds zero.
Data center is $192M/quarter, roughly a third of revenue, concentrated on one hyperscaler per Citi's analyst, undisputed by management, with second-wave revenue equal to that customer not expected until 2028-2029.
120 bps of the 41.9% gross margin came from one-time IEEPA tariff refunds, EPS carries $0.04 tariff plus a 14% tax rate against a 16-17% target and $5M of convert interest income. Strip it and the quarter is ~$2.58.
Competitors are shipping into ASIC hyperscaler 800V rack programs in 2H 2026; AE expects initial production revenue late 2027 and meaningful ramp in 2028. The market moves while AE qualifies.
+62
CONFIDENT"Management volunteered specifics unprompted: $640M Q3 guide, $2.6B capacity math, 40bps per $50M margin rule, $180-195M capex, quantified distributor metrics (bookings +80%, resales +40%). Oldham's margin walk split into volume, mix, efficiency as three named legs."
Full transcript with specific numbers on every segment, guided Q3/Q4 to records, quantified capacity and capex, clean CFO reconciliation. Gaps only on Q4 semi deceleration attribution and second-wave timing.
Q3 print at $640M +/-20M with gross margin 41-41.5% ex any tariff benefit, plus Thailand first production revenue in Q4, proving margin expansion runs on mix and capacity is converting before 2027.
Hyperscaler capex guide breaks while 145-day inventory and $5B of capacity are already committed, and second-wave revenue does not arrive until 2028-2029 on management's own timeline.
The verdict is public. The case behind it is Pro: the forensic reading of guidance and language, the bull and bear cases, the red flags, and the play. It opens to everyone on Nov 1, 2026.