For most of the last two decades, professional portfolios were built on a dependable assumption: when stocks fall, bonds rise. Between 2000 and 2023, the correlation between US stocks and bonds averaged around -0.29 — a reliable, negative relationship that let a traditional 60/40 portfolio smooth out equity drawdowns almost automatically. That assumption is now in question.

In early 2022, as the Federal Reserve raised rates aggressively to fight inflation, stocks and bonds began falling together for one of the only sustained periods in modern market history — the correlation flipped sharply positive as both asset classes were driven by the same force. It has fluctuated since, but the era of dependable negative correlation appears to be over for now. A recent IMF financial stability analysis went further, warning that stock-bond diversification is offering less protection during selloffs, pointing to expanding government bond issuance and reduced central-bank balance-sheet absorption as structural reasons the relationship may stay unstable.

The clearest evidence isn't in the correlation statistic itself — it's in where capital has actually moved for protection. Gold has more than doubled since the start of 2024. Silver, platinum, and palladium have surged. The Swiss franc has strengthened. None of these are the textbook hedge for a stock portfolio; investors have simply been forced to look elsewhere because the textbook hedge stopped being reliable.

This doesn't mean 60/40 investing is dead, or that bonds are useless. Over multi-decade histories, the correlation between stocks and bonds has flipped sign more than once, including a sustained positive relationship through most of the 1970s and 1980s. What it does mean is that a portfolio built exclusively on the assumption of one stable relationship between two asset classes is making a bet on that relationship holding — whether or not the investor realizes they're making it.

The fair counterargument is that the 2022 episode was a rate-shock anomaly, not a new regime — and there is real evidence for that view. Barclays Private Bank's analysis shows the rolling equity-bond correlation has already fallen back from its mid-2024 peak as inflation cooled, and bonds have started regaining some of their historical diversification value as rate cuts have progressed. It's entirely possible the negative correlation of the 2000s returns as the norm rather than the exception. The honest position is that nobody knows which regime is coming next — which is itself the argument for not depending on either one exclusively.

For an allocator, the practical implication isn't panic. It's genuine diversification: exposure to return streams that don't depend on the stock-bond relationship holding at all. Strategies uncorrelated to both equities and duration, and alternative stores of value like gold or, increasingly, Bitcoin, aren't a hedge against one bad year. They're a hedge against the possibility that the relationship investors have relied on for twenty years doesn't come back the way it left.