3 hours ago
3
The 30-day correlation between Bitcoin and the S&P 500 has dropped to -0.17, its lowest level since 2021. At the same time, stock-to-stock correlations across the equity market are sitting near record lows.
The great AI debt binge
AI-related capital expenditures are expected to hit roughly $400 billion in 2025. And here’s the part that should make your antennae twitch: an increasing share of that spending is being funded not from cash piles, but from corporate bond sales.
Meta recently issued a $30 billion bond offering. Amazon raised $25 billion with yield incentives to attract buyers. These are companies that used to sit on mountains of cash. Now they’re borrowing.
The shift shows up clearly in cover ratios, the metric that measures how easily a company can service its debt from earnings. Those ratios have fallen from around 5x to below 2x for some of these hyperscalers. In English: these companies used to earn five dollars for every dollar of debt payments. Now it’s closer to two.
Most of these firms are still keeping overall leverage below 1x. But companies that were net-cash are transitioning to modest net-debt positions, and they’re doing it fast.
Why low correlation is the perfect camouflage
Equity markets are experiencing unusually high dispersion among AI-related stocks, meaning these companies are moving in different directions on any given day despite sharing the same fundamental bet on artificial intelligence.
This environment, characterized by low correlations and high gamma, creates a false sense of security. Risk models that rely on correlation to detect concentrated exposure are essentially blind to what’s building in the AI space.
The bond market is telling one story: tech companies are borrowing aggressively to fund AI buildouts. The equity market is telling another: everything is fine, look how uncorrelated we are.
Where crypto fits into this picture
Bitcoin’s -0.17 correlation with the S&P 500 means Bitcoin has been moving in the opposite direction of the S&P 500 more often than not over the past 30 days. If corporate credit conditions tighten because investors get nervous about the pace of AI debt issuance, that affects interest rates, liquidity, and risk appetite across every asset class, including crypto.
If the AI revenue story doesn’t materialize fast enough to justify $400 billion in annual spending, companies that borrowed to build data centers and train models need those investments to generate returns. The leverage that looked manageable at below 1x starts to feel heavier.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
English (US) · 