Cut, hedge, or hold the AI-infrastructure bet
- CEOs
- Decision deck
- Generic
An investment-committee decision on whether to cut, hedge, or hold a fund's AI-infrastructure exposure. Big Tech is guiding roughly $725B of 2026 capex, up 77 percent, against about $75B of revenue earned today; demand is real and contracted but the financing is fragile and GPUs are booked over six years while they age out in two. The deck holds the demand-backed core, hedges the tail, tilts off the leveraged neoclouds, and pre-commits the tripwires that flip hold to cut.
The deck.
Citations on every slide, sourced back to the row, sheet, or cell behind each claim.
1/–
Click either side of the slide, or use the arrow keys.
The brief.
What we told Huegoo before generation.
Audience. The investment committee and CIO of a diversified institutional fund with heavy AI-infrastructure exposure across hyperscalers, chipmakers, neoclouds, power, and data-center REITs.
Occasion. A decision on whether to cut, hedge, or hold the AI-infrastructure sleeve now.
Time. 26 minutes.
Brief. The room came in skeptical of the AI build-out. The question is not whether AI is real; it is whether the structure financing it survives the gap between the spending and the revenue. Decide whether to cut the sleeve, hedge it, or hold it, and pre-commit the tripwires that flip hold to cut. Establish the demand as contracted, not hoped-for, then weigh the fragile financing and the depreciation gap, and close on a staged stance the committee can vote.