The correlation between Bitcoin and a constructed basket of the largest AI-exposed equities — Nvidia, Microsoft, Meta, AMD, Broadcom, and a handful of pure-play AI-infrastructure firms including Vertiv and Eaton — has climbed to a multi-year high through the second half of 2025. The 30-day rolling correlation reading currently sits at 0.61, the highest level recorded against any constructed equity basket since BTC began trading on regulated U.S. exchanges. Against the broader S&P 500, BTC's correlation has climbed in parallel but more modestly, sitting near 0.5.
Both asset classes have responded to the same set of macro inputs. Federal Reserve policy expectations dominate the daily tape, with futures-implied terminal-rate readings producing immediate cross-asset reactions. Real-yield levels matter even more directly, with the long-duration discount-rate sensitivity of mega-cap AI stocks producing factor exposures that BTC has, surprisingly cleanly, begun to share. And recurring fears that the AI capex cycle is producing valuations that will not survive a recession have generated waves of risk-off behavior that hit both asset classes within minutes of one another.
The AI capex story has dominated equity-market discussion through the year. Estimated 2026 hyperscaler capital expenditure on AI infrastructure now exceeds $720 billion in the consensus estimate, up from $480 billion a year earlier. The buildouts have driven the order books at Nvidia, Vertiv, Eaton, and a handful of utility-equipment firms to multi-quarter highs, but the market debate has increasingly turned to whether end-customer demand can monetize the spending fast enough to justify current implied returns. Each significant data point on AI revenue or capex revisions has produced sharp moves in both AI stocks and Bitcoin in tightly coupled sequence.
The implication is that Bitcoin has, in effect, joined the "AI trade" as a closely correlated risk-asset complement. Multiple discretionary macro funds now openly group BTC inside their AI exposure when discussing portfolio risk, and several ETF-overlay strategies have begun explicitly hedging BTC with AI-stock baskets rather than broad equity indices. That positioning has been profitable through the late-2024 and early-2025 melt-up, when both asset classes climbed in tandem on a tide of capital rotating into long-duration risk. It has become a liability through October's reversal, when AI-stock weakness translated cleanly into BTC downside.
"For the first time, you can describe Bitcoin's day-to-day price action almost entirely with reference to the AI capex narrative and the Fed reaction function," wrote one strategist at a major macro hedge fund in a recent client note. "That doesn't change the long-term thesis, but it changes how you size and hedge the position in any given quarter." Other research desks have echoed the framing. Bridgewater's quarterly client letter referenced the BTC-AI correlation explicitly for the first time in its risk-factor analysis section.
Whether the correlation persists through a more meaningful AI-infrastructure correction is the central question for crypto allocators heading into 2026. Several scenarios have been mapped: a clean AI capex pause that takes mega-cap valuations down 25%-30% would, on current beta readings, drag BTC into a similar drawdown range; a soft landing in which AI revenue accelerates would likely produce continued BTC strength alongside; a true bubble-burst scenario remains a tail risk but one that more research desks are now modeling explicitly.
What desks are watching most carefully is whether BTC begins to decouple from the AI complex on idiosyncratic catalysts the way it briefly did during the 2024 ETF approval window. If a sustained altcoin rotation, or a new catalyst like a substantial sovereign-treasury allocation, can pull BTC out of the AI orbit, the correlation question becomes much more nuanced. If the asset stays tightly coupled through 2026, allocators will need to revise their risk frameworks accordingly — and the diversification narrative that drove much of the 2024 institutional adoption will need to be rewritten.