← All sectors / The AI transformation
016 · Semiconductors & equipment
The silicon supply chain
Curve position
Binding constraint
Advanced packaging and high bandwidth memory capacity.
Why this sector sits at Flight
We place semiconductors and equipment at Flight. The binding constraint is no longer demand but supply: capacity at the leading edge, in memory and in advanced packaging. Capital is pouring in to relieve it. Global chip sales reached a record $120.6 billion in May 2026, up 104.1% year on year and the 15th straight monthly record.
The sector has already repriced. The SOX rose more than 100% to a record high in late June 2026, then fell as much as 29% as investors asked whether valuations had run ahead of earnings. A correction like that in the middle of a buildout fits Flight, not Orbit.
We would move it to Orbit once growth stops accelerating: sales growth settling at a steady rate and equipment spending flattening. Its old label, Takeoff, is not one of our four positions.
Reviewed on a two month cycle. The position moves only when a dated, verifiable change in the binding constraint justifies it.
Heatmap / How this sector's stocks are moving
| Stock | Price | 1 Day | 1 Week | 1 Month | Year to date |
|---|---|---|---|---|---|
| $AMAT | $369.40 | −3.28% | −11.20% | −30.34% | +43.74% |
| $LRCX | $221.43 | −4.03% | −12.76% | −33.68% | +29.36% |
| $KLAC | $149.79 | −4.10% | −5.66% | −24.14% | +23.28% |
| $ASML | $1,496.30 | −3.60% | −10.02% | −18.42% | +39.86% |
| $TSM | $395.75 | −2.24% | −9.96% | −17.95% | +30.23% |
| Sector average | −3.45% | −9.92% | −24.91% | +33.29% |
Prices as of November 5, 2026. The sector's move is the equal-weighted average of these listed companies, a market proxy for context, not recommendations. See it on the full heatmap →
Sentiment / The mood in this sector's stocks
Since 2018, Semiconductors & equipment has been in Capitulation and steady 8 separate times. After those, its stocks beat their usual three months 4 times (50%), by a median of −0.5 points.
Confidence: low, a split like this is common by chance, so no edge is called.
Semiconductors are the substrate of the entire AI economy, and the capital cycle behind them is historic: leading edge fabs now cost tens of billions each, and every major economy is subsidizing construction inside its borders. The chain that designs, builds, and equips those fabs is the deepest moat system in technology.
Historical context: chipmaking consolidated for decades until each critical layer: lithography, design software, advanced logic manufacturing. Narrowed to a handful of firms or fewer. AI demand is now stress testing that concentrated structure with the largest order wave it has ever seen.
The structural driver is capability plus geography: aI models demand leading edge logic and memory in exploding volume, while governments treat domestic capacity as national security, layering subsidy funded construction on top of commercial demand in a way no prior cycle had.
The technology frontier keeps adding investable layers: advanced packaging that stitches chiplets together, high bandwidth memory stacked beside accelerators, backside power delivery, and the metrology and inspection tools each new node requires more of. Complexity is the equipment industry's revenue model.
Adoption economics are visible in capex plans: foundry, memory, and logic leaders publish spending that flows directly to equipment and materials suppliers with multi quarter lag, one of the more forecastable revenue chains in technology for those who track it.
The beneficiaries span design software vendors, equipment makers from deposition to test, materials suppliers (photoresists, gases, substrates), and the packaging and assembly tier, plus the specialty foundries serving analog, power, and radio frequency chips every AI system also contains.
The value chain rewards different risk appetites at different layers: equipment oligopolies compound steadily, materials suppliers ride volume, and small cap specialists: a niche metrology tool, a substrate innovator. Carry the torque when their layer becomes the bottleneck.
The overlooked layer is exactly there: sub scale equipment and materials names, test and burn in specialists, and the power semiconductor makers riding data center electrical content. Many trade on trailing cyclicality while their AI driven backlog builds.
Competitive dynamics are geopolitical as much as commercial: export controls partition markets, subsidy races relocate capacity, and every restriction reshuffles which suppliers can sell where. Policy literacy is now part of semiconductor analysis.
Risks: this is a deeply cyclical industry, inventory corrections are violent, capex plans get cut fast, and multiple compression arrives before the downturn is visible in earnings. Export control changes can strand revenue overnight, and technology transitions can leapfrog incumbent toolmakers.
What to watch: equipment book to bill and backlog language, memory pricing, foundry utilization, subsidy disbursements converting to orders, and export control revisions. The research follows the chain from capex announcement to component order, and the small caps positioned at the tightest links.
Coverage / Daily Disruptor issues in this sector
