Market Investment Markets
Records and Warnings: The Market Is Doing Both at Once
The S&P 500 hit a record this week even as two of the financial world's most credentialed pessimists issued stark warnings. Michael Burry — whose short on mortgage-backed securities in 2007-2008 made him famous for being right when almost everyone was wrong — is warning of a 1987-style correction. The original Black Monday, October 19, 1987, saw the Dow fall 22.6 percent in a single session, driven by portfolio insurance strategies that created a self-reinforcing feedback loop. Whether analogous structural features exist today is genuinely debated. Ray Dalio, founder of Bridgewater Associates, which manages roughly $150 billion, is drawing parallels to both 1929 and 2000, arguing that AI is showing 'classic signs' of bubble behavior — a gap between current earnings and valuations premised on future potential. The 2000 analogy stings because the companies valued highest at the dot-com peak were real companies with real technology; the technology was not wrong, the price was.
The most concrete data point in the stress narrative is the collapse of the Situational Awareness hedge fund, which lost roughly two-thirds of its value in July and was forced to liquidate $16 billion in public equities to Citadel. Analyst Jim Cramer characterized the event as a 'clearing event' for tech stocks, arguing that forced selling at that magnitude creates buying opportunities and that markets will normalize once the liquidation pressure is absorbed.
A more structural concern involves roughly $100 billion in bonds from investment-grade issuers — including Oracle and Stellantis — now trading at spreads that imply junk-level credit risk. The 'fallen angel' dynamic describes what happens when bonds rated BBB or better are formally downgraded to junk: pension funds and insurance companies operating under investment-grade mandates must sell simultaneously, amplifying the very dislocation that triggered the downgrade. The driver of stress in several cases is AI-related borrowing — companies taking on substantial debt to finance data centers, compute capacity, and energy contracts, with markets repricing the paper before rating agencies formally act if AI-service revenues do not materialize on the timeline the borrowing assumed.
The macro backdrop adds further pressure. Multiple Federal Reserve officials signaled this week they are prepared to raise rates if inflation does not ease, following a July press conference by Chair Kevin Warsh — described as having a more hawkish reputation than his predecessor — that rattled markets. Higher rates make every leveraged position more expensive to maintain, and companies carrying heavy AI-related debt loads feel the effect first. Treasury Secretary Bessent separately led the first U.S.-Japan joint yen intervention since 1998, coordinating currency support that suggests the yen's weakness has reached a threshold both governments consider destabilizing for regional trade dynamics and carry-trade stability.