Anna Wong

The AI capex cycle is in its early stages with at least another one to two years of growth, supported by strong demand that outpaces supply, as evidenced by Google's AI product revenue growing 800% and margins increasing, and Intel's CPU orders being full. This is not a bubble because the market is differentiating between successful and overinvesting hyperscalers.

The AI capex cycle is in its early stages with at least another one to two years of growth, supported by strong demand that outpaces supply, as evidenced by Google's AI product revenue growing 800% and margins increasing, and Intel's CPU orders being full. This is not a bubble because the market is differentiating between successful and overinvesting hyperscalers.

States Micron is diverting production from consumer electronics to meet AI data center demand, contributing to a 100-200% increase in memory chip prices. The AI-driven shortage has "no end in sight" and no marginal producers, unlike oil, making this a persistent supply shock. Only half of the estimated price shock has passed through to consumer prices so far. Continued, inelastic demand for AI memory chips should support sustained pricing power and revenue for producers like Micron. An unexpected collapse in AI infrastructure investment demand.

States that if oil stays at ~$116, US production will rise from 13 to 17-18 million barrels per day within ~9 months. Also notes that war expansion (a driver of $150-$200 oil) would necessitate higher defense spending. Higher oil prices incentivize increased domestic production (adding to GDP), and a war scenario would spur defense manufacturing, which requires commodities. The energy sector is positioned for growth in both production volume and, in a geopolitical escalation scenario, supported demand from the linked defense industrial expansion. A rapid de-escalation of conflict leading to a swift collapse in oil prices.