The CFO admitted he cannot model his own fastest-growing segment and the stock still went up.
Thesis: Content intensity per wafer is real and backed by POR incumbency at leading nodes, but the business mix is drifting toward the structurally lower-margin ICS segment while semi margins compress year over year. The street's 7% advanced node model is too low, but the 20% print may not be sustainable. The asymmetry is moderate and near-term: guidance raise is supported, but Q4 math at the top of the Q3 range implies flat to down sequential.
Verdict: HOLD — Conviction: MEDIUM
Catalyst: Q3 sequential landing above low single digits with semi EBITDA holding mid-30s would confirm the advanced node growth rate is a durable step change rather than pull-forward, and would force upward revisions on the 7% advanced node street model.
Key Risk: Q3 sequential growth below low single digits or semi EBITDA slipping under 33%. Either converts Mix Shift Spin into a guide cut and validates the Q4 pre-blame.
The Tell: Goss volunteered unprompted that advanced packaging and PCB volumes lack reliable third-party data: 'the broader market is still trying to get a handle around third-party data... we're working with some of those third parties to try and figure out how to model that a little bit more accurately... us included.' He is telling you he cannot independently verify the volume assumptions underpinning ICS's 50% platform growth. That is a CFO admitting his fastest-growing segment's trajectory is built on data he cannot yet reconcile.
Friction Level: MODERATE_FRICTION — Both sides agree on the numbers. Bull says 20% advanced node growth versus a 7% model is a durable step change with pricing power backing it. Bear says the same gap is Unsustainable Trend Confidence until a future guide proves otherwise, and the semi segment's margin compression from mix is being reframed as leverage.
2026Q2 · Aug 4, 2026
3 of 5 readings agreed3 HOLD · 2 LONGRead the full call
Nine Beats, Same Three Questions: Semi Margins, Q4 Math, Visibility Gap
Content intensity per wafer is real and backed by POR incumbency at leading nodes, but the business mix is drifting toward the structurally lower-margin ICS segment while semi margins compress year over year. The street's 7% advanced node model is too low, but the 20% print may not be sustainable. The asymmetry is moderate and near-term: guidance raise is supported, but Q4 math at the top of the Q3 range implies flat to down sequential.
Goss volunteered unprompted that advanced packaging and PCB volumes lack reliable third-party data: 'the broader market is still trying to get a handle around third-party data... we're working with some of those third parties to try and figure out how to model that a little bit more accurately... us included.' He is telling you he cannot independently verify the volume assumptions underpinning ICS's 50% platform growth. That is a CFO admitting his fastest-growing segment's trajectory is built on data he cannot yet reconcile.
Ninth consecutive quarter of profitable organic growth. Raised full-year net sales to $5.55-5.65B, EBITDA to $1.675-1.725B, EPS to $4.40-4.60. Adjusted EBITDA up 24% against sales up 22%. Operating leverage confirmed in the print.
Management states targeted pricing actions are absorbing $20M of logistics and energy input costs, half already through, with 'no near-term risk to supply or output.' Portfolio sits in POR at leading nodes with no second source. Customers absorb increases.
Semi gross margin at 49% and EBITDA margin at 34% both down year over year and sequentially per CFO. Explanation is 'product mix' and 'growth investments.' ICS at 44% gross, 29% EBITDA grows 28% organic against semi's 17%. Mix drifts toward the lower-margin half while management presents it as leverage.
Advanced nodes grew over 20% against an Investor Day model of 7%, a 13-point gap. Kemp declines to provide quarterly forward color on the rate, saying only they'll hit 45-50% of portfolio 'early.' The street models 7%. Neither side has proof of durability.
Kemp states $600M of growth capital deployed since 2022 in modular expansions. He directly answers the capacity question: not constrained, footprint in place, 'growth is a bit faster than what we were expecting.' Management is running ahead of the ramp, not behind it.
Goss volunteers that 'the broader market is still trying to get a handle around third-party data' for advanced packaging and PCB volumes, adding 'us included.' These are the two platforms driving 50%+ growth. He cannot quantify his own fastest-growing driver with confidence.
+15
NEUTRAL"Goss raised full-year guidance with specific dollar ranges and quantified $20M input cost mitigation. But he also introduced new hedging: 'a couple of variables we are watching,' 'to the extent those items resolve or improve.' Direct answers mixed with pre-blaming Q4 consumer seasonality and inventory control. Kemp deferred advanced node growth rate quarterly guidance entirely."
Rich financial detail with segment margins, utilization rates, capacity deployment, and cash flow reconciliation. But no GAAP metrics disclosed anywhere, no channel inventory data, no DSO or receivables breakdown, and CFO explicitly states he cannot model advanced packaging and PCB volumes using third-party data.
Q3 sequential landing above low single digits with semi EBITDA holding mid-30s would confirm the advanced node growth rate is a durable step change rather than pull-forward, and would force upward revisions on the 7% advanced node street model.
Q3 sequential growth below low single digits or semi EBITDA slipping under 33%. Either converts Mix Shift Spin into a guide cut and validates the Q4 pre-blame.
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 2, 2026.