Why do cross-asset correlations break down during crises?

Question

AcadiFi · Accepted Answer

Stock-bond correlation isn't a physical constant - it depends on the dominant macro driver. When growth shocks dominate (2008, 2020), stocks fall and bonds rally as investors flee to safety, producing negative correlation. When inflation shocks dominate (1970s, 2022), rising rates hurt both bonds (directly via discounting) and stocks (via higher discount rates and margin compression), making correlation positive. In 2022, CPI surged from 7% to 9%, the Fed hiked 425 bps, and a 60/40 portfolio lost 17% - its worst drawdown since 1937 - because both legs fell together. The regime shift wasn't random; it was a predictable response to the macro driver changing from growth to inflation. For risk management, assume cross-asset correlations are regime-dependent and stress-test inflation-driven scenarios where bonds don't hedge stocks. Alternative diversifiers during inflation regimes include commodities, TIPS, trend-following CTAs, and low-vol equity factors. Quantify with regime-switching models (Markov-switching) or conditional correlations (DCC-GARCH) rather than full-sample unconditional correlations. ```mermaid graph LR A[Growth shock] --> B[Stocks down
bonds up] C[Inflation shock] --> D[Stocks down
bonds down] B --> E[Diversify] D --> F[Correlated loss] ```

Why do cross-asset correlations break down during crises?

Master Part II with our FRM Course

Related Questions

Practice Questions