Bonds Are a Safety Net Only in the Right Kind of Selloff

The conclusion: bonds are conditional protection

Fixed income can be a powerful diversifier when a market selloff is driven by weakening growth. It is much less reliable when the shock comes from interest rates.

In a growth shock, recession fears usually pull expected policy rates lower, Treasury yields fall, and bond prices rise. That pattern helped bonds cushion equity losses in periods such as 2000–02, 2008, and 2020.

In a rate shock, the problem is rates themselves. Inflation, Federal Reserve tightening, heavy Treasury supply, or forced bond selling can push yields higher into the panic. Because bond prices fall when yields rise, stocks and bonds can decline together. The same bond that hedges one crisis can become a liability in another.

2022 showed what happens when the 40 falls with the 60

The clearest recent example is 2022. Inflation forced the fastest repricing of interest rates in four decades, and assets priced off those rates fell together.

• The S&P 500 fell 18% on a calendar-year total-return basis and about 25% peak to trough.
• The Bloomberg US Aggregate Bond Index fell 13%, its worst calendar year on record.
• A hypothetical 60/40 portfolio of 60% S&P 500 and 40% Bloomberg US Aggregate, rebalanced annually, fell roughly 16%.

The backdrop explains the breakdown. CPI peaked at 9.1% in June 2022. The Fed raised its policy-rate target from near zero to 4.25%–4.50% in nine months. The 10-year Treasury yield rose from about 1.5% at the end of 2021 to more than 4% at its October 2022 closing high. Bonds entered the shock with low yields and high duration, leaving little income cushion against rising rates.

The stock-bond relationship changes with the regime

Stock-bond correlation is not a fixed law of markets. It is a readout of the shock dominating the cycle.

Across the Bloomberg US Aggregate Bond Index’s history beginning in 1976, stock-bond correlation was positive for much of the first quarter century. The reliably negative relationship that shaped the modern 60/40 playbook came later, roughly from 2000 through 2021, when major crises were largely growth shocks and rates fell into each one.

As of September 25, 2026, rolling stock-bond correlation had reached some of its highest levels since the late 1990s. That does not guarantee another rate shock, but it shows why the hedge should not be treated as automatic.

Higher yields help, but they do not remove rate risk

As of September 25, 2026, the Bloomberg US Aggregate Bond Index yielded about 5%, compared with under 2% at the start of 2022. That higher income gives bonds more cushion against rising rates.

The cushion is not the same as immunity. If yields keep climbing, bond prices can still fall, and they can fall at the same time as stocks. As of that date, inflation remained above the Fed’s target, the Fed had resumed raising rates for the first time since 2023, the 10-year Treasury yield had reached its highest level since 2007, and equities had spent much of the year near record highs. Those conditions do not ensure a rate shock, but they keep the risk alive.

Diversification should not rest on one correlation

Cash, short-term Treasuries, and TIPS each help in some conditions and struggle in others. Cash and short-term Treasuries can provide stability while rates rise, but inflation can erode purchasing power and they do not typically rise enough to offset losses elsewhere. TIPS can help against inflation, but they can still lose value when real rates rise; in 2022, the Bloomberg US TIPS Index lost roughly 12% during the worst inflation episode in 40 years.

Bonds still have a role: income, stability in many environments, and meaningful protection when the next crisis comes from the growth side. But diversification that works only when stock-bond correlation is negative is conditional. We view stronger diversification as structural rather than purely statistical: protection should be designed into what a portfolio owns, not depend entirely on which shock arrives.

Past performance and past correlations do not guarantee future results. Indexes are unmanaged, do not reflect fees, expenses, or taxes, and cannot be invested in directly. A hypothetical 60/40 portfolio is for illustration only and does not represent any actual account or strategy.

Downloads