Are Bonds Still Good Diversifiers? It Depends What Risk You're Hedging

In 2022 the S&P 500 lost 18.1% and the Bloomberg U.S. Aggregate bond index lost 13.0%, both including income. A 60/40 mix of the two, set at the start of the year, lost about 16%. Bonds fell alongside stocks instead of cushioning them, and the year produced a wave of obituaries for bonds and for the 60/40 portfolio.¹

The cause was specific. CPI inflation reached 9.1% in June 2022, the highest 12-month rate since 1981, and the Fed raised its policy rate by 4.25 points over the year, its fastest tightening since the early 1980s. Rising inflation and rising rates are the environment where nominal bonds are most exposed. Whether bonds still belong in your portfolio depends on what risk you are asking them to reduce, and whether the bond you own is built for that job.

Diversifier versus hedge

Most arguments about bonds blur two ideas that deserve to be kept apart.

A diversifier is something that does not move in lockstep with stocks. Lower volatility helps but is not required: an asset can pull down a portfolio’s total risk whenever its imperfect correlation with stocks more than offsets its own volatility and weight.
A hedge is something you expect to rise when stocks fall. That is a much stronger claim.

High-quality bonds are reliably the first and only conditionally the second. Portfolio risk depends on three things at once, not just correlation:

\sigma_p^2 = w_s^2\sigma_s^2 + w_b^2\sigma_b^2 + 2\,w_s w_b\,\rho_{sb}\,\sigma_s\sigma_b

As long as the stock-bond correlation $\rho_{sb}$ stays below +1, adding bonds can still lower total portfolio volatility, and lower bond volatility makes that easier without being required. Even two assets with the same volatility produce a less volatile portfolio when their correlation is below +1. A mildly positive correlation shrinks the benefit and leaves some of it in place. What broke in 2022 was the expectation that bonds rally in every crash; the diversification math still held. The same distinction applies in our guide on what actually hedges inflation.

Why the correlation flips between negative and positive

Stocks and bonds respond to two different kinds of news, and which one dominates decides whether they move together or apart.

Growth news pushes them in opposite directions. When recession fears rise, stocks fall, the Fed is expected to cut, yields drop, and bond prices rise. This is the classic flight-to-quality environment where Treasuries hedge stocks.
Inflation news pushes them the same direction. An inflation surprise raises discount rates and squeezes margins, so stocks fall, while rising yields drag down nominal bond prices at the same time.

Alfie Brixton and five AQR colleagues formalized this in a 2023 Journal of Portfolio Management paper. The sign of the stock-bond correlation, they find, tracks how volatile inflation news is relative to growth news, and their simple model explains around 70% of the long-term variation in the U.S. stock-bond correlation.² John Campbell, Adi Sunderam and Luis Viceira reach a similar result with a different model: the covariance of U.S. stocks and Treasuries was slightly positive on average from 1953 to 2014, unusually high in the early 1980s, and negative in the downturns of 2001 and 2007 to 2009.³

The negative correlation investors grew used to after 2000 is the exception in the modern U.S. record. Measured over 36-month windows, the average U.S. stock-bond correlation was 0.35 from 1970 to 1999 and −0.29 from 2000 to 2023.⁴ AQR’s rolling 36-month measure had climbed back to about +0.5 by mid-2026, its highest in decades.⁵ A positive correlation today fits the longer history.

Environment	Growth news	Inflation news	Stocks	Treasuries
Goldilocks (soft landing)	Better	Lower	Up	Flat to up: yield plus falling rates
Overheating	Better	Higher	Often up, on earnings	Down as yields rise
Recession	Worse	Lower	Down	Often up: rate cuts, flight to quality
Stagflation	Worse	Higher	Down	Down: yields and discount rates rise together

Four economic environments, and where assets have sat since 1972

Growth news and inflation news combine four ways, and AQR’s September 2026 paper Inflation Redux? places assets on that grid.⁵ For each asset it estimates two partial correlations over rolling 12-month windows from 1972 to mid-2026: one with U.S. growth news, holding inflation news fixed, and one with inflation news, holding growth news fixed. “News” blends the change in the year-over-year rate with the miss against the Survey of Professional Forecasters, so the measure tracks surprises rather than levels. The chart redraws 13 of the assets in AQR’s Exhibit 4.

Three features of the map matter for a stock-and-bond investor.

The U.S. 60/40 sits next to U.S. stocks. Its growth sensitivity is about 0.39, against 0.43 for stocks alone. Measured by risk, a 60/40 portfolio is mostly a stock portfolio. Its inflation sensitivity, about −0.28, is a little more negative than the −0.20 for stocks, because Treasuries add inflation exposure in the same direction.
Treasuries sit in the Recession quadrant. They have done well on bad growth news and badly on inflation news. That combination is the hedge investors relied on from 2000 to 2020 and the risk that showed up in 2022.
No broad stock or bond index sits on the right side of the map. The assets that did better on inflation surprises are commodities, energy stocks, gold, breakevens (long TIPS, short nominal Treasuries) and AQR’s two trend-following strategies.

Why the cause of a downturn decides what bonds do

When demand falls, as in 2008, spending and inflation fall together. The Fed cuts, Treasury yields drop, and bond prices rise while stocks fall: the Recession quadrant. When supply falls, output weakens while prices rise. The 1970s oil shocks were the textbook case, and 2021 to 2022 combined supply-chain disruptions and an energy price spike with strong demand. The Fed tightens instead of easing, yields rise, and bonds fall with stocks: the Stagflation quadrant.

Stocks lose on inflation surprises through more than higher interest rates. Studying stock moves on CPI release days, Benjamin Knox and Yannick Timmer of the Federal Reserve Board find that investors expect firms’ nominal cash flows to stay flat while the equity risk premium and nominal yields rise. Higher prices do not show up as higher expected profits in the short run, and investors demand a larger premium to hold stocks. Firms with more market power fell less.⁶ The result fits a much older one: Eugene Fama and G. William Schwert found U.S. stock returns were negatively related to expected inflation, and probably to unexpected inflation, from 1953 to 1971.⁷ Our guide on inflation beta works through the Knox and Timmer estimates.

AQR made this argument before it played out. Its October 2021 paper When Stock-Bond Diversification Fails, by Ashwin Thapar, Thomas Maloney and Alfie Brixton, warned that stocks and nominal bonds could fall together on upside inflation surprises and named breakevens, commodities and trend strategies as diversifiers.⁸ The S&P 500 peaked ten weeks later, on January 3, 2022.

How much weight the map can bear

The map summarizes history and makes no forecast. Overlapping 12-month windows mean that five decades of quarterly data hold far fewer independent observations than the number of data points suggests. The correlations are contemporaneous, so they cannot separate inflation from the Fed’s response to it. Several dots are reconstructions: the trend strategies are AQR backtests, and the inflation-linked bond series is synthetic before TIPS were first issued in 1997. Positions also move between eras, as the correlation history above shows.

Which inflation an asset hedges is also contested. AQR finds that a diversified commodity basket helped against core inflation as well as energy inflation over its sample. Xiang Fang, Yang Liu and Nikolai Roussanov, using a different method and sample, find that commodities hedged energy inflation but not core.⁹ AQR also manages commodity and trend-following funds. Our review of the AQR paper covers that disagreement, the costs of trend following, and why its optimized portfolio weights should not be copied.

What you hire bonds to do

Nobody should own high-quality bonds only on the bet that they spike during stock crashes. Bonds hold several jobs at once, and most of them have nothing to do with crash timing.

Lower portfolio volatility. A 60/40 mix is usually calmer than 100% stocks even when correlations are not negative.
Income. Over a horizon of about twice a bond fund's duration, its annualized return converges to the starting yield (Leibowitz, Bova & Kogelman, 2014). Since 1954 the 10-year Treasury's starting yield has come within a point of its next-decade return about two-thirds of the time; see the rebalancing test.
Fund near-term spending. Retirees should not pay this year's bills out of assets that depend entirely on equity valuations.
Rebalancing dry powder. Bonds hold value you can sell to buy stocks after a drawdown.
Liability matching. A future tuition bill or retirement paycheck is a bond-like liability, best met with a bond-like asset. See our guide on matching bond duration to your goals.
Growth-shock hedge. When recession fears push rates down, Treasuries can still rally.
Behavioral ballast. A shallower drawdown helps people stay invested.

Vanguard's research finds that global bonds have historically had lower volatility than equities and often delivered positive returns in equity down years, with 2022 as the painful exception. Morningstar's 150-year stress test reaches a similar conclusion: 60/40 portfolios usually had shallower drawdowns than all-equity, even though the bond side of 2022 was unusually brutal.

Do higher yields improve the odds of a hedge?

Yes, with a caveat. After the 2022-2023 repricing, Treasury yields are materially positive again. As of mid-June 2026, the curve ran roughly 3.8% on 3-month bills, 4.1% on the 2-year, 4.2% on the 5-year, 4.5% on the 10-year, and just under 5% on the 30-year. Higher starting yields help bonds in three ways: more carry if rates hold, more cushion to absorb a rate rise before the total return turns negative, and more room for rates to fall if a recession arrives, which produces price gains.

A useful one-year approximation:

\text{1-year bond return} \approx \text{starting yield} - \text{duration} \times \Delta\text{yield}

A bond fund with a 5-year duration yielding 4.5% can absorb about a 0.9 percentage-point rise in yields over a year before it loses money, since 4.5 ÷ 5 = 0.9. This first-order rule assumes a parallel shift in yields and ignores convexity, fund expenses, credit risk, and roll-down, so it works best for high-quality bond funds over short horizons and small yield moves. For moves above about a point, convexity adds back roughly ½ × D(D+1) × Δy²; for a 10-year Treasury in 2022 that was worth about 2.7 points. A 4.5% yield is far better than the sub-2% world of 2020 and 2021, when even a small rate rise produced losses. Higher yields still buy no guaranteed hedge. If stocks fall because inflation surprises higher and yields jump 1.5 to 2.0 points, an intermediate or long bond fund can still lose money. The cause of the stock decline still decides whether bonds help.

The calculator below runs your mix through the four environments and three historical years, before and after inflation. The 1974 replay shows a second way inflation hurts bonds. The 10-year Treasury gained 2.0% that year because it started with a yield near 7%, but CPI rose 12.3%, so a 60/40 portfolio lost about 15% in nominal terms and about 24% after inflation.

Bond substitutes that carry equity risk

The most expensive mistake in this debate is swapping high-quality bonds for things that look like bond replacements but quietly carry equity risk. A higher yield usually means more risk. AQR measured the five-year equity beta of several popular alternatives, and the gap is large:

Asset	Equity beta (5-yr)	What that means
High-quality bonds	0.19	Mostly its own risk, light on equity exposure
Buffer / defined-outcome funds	0.63	Roughly 3x the equity risk of bonds
Private credit	0.70	Smooth marks hide equity-like credit risk
Bitcoin	2.09	About 10x the equity risk of bonds

During a severe equity drawdown, credit spreads widen, private assets are often marked less frequently and model-valued so reported volatility lags the underlying economic risk, and high-yielding income products tend to behave more like stocks than like Treasury ballast. The smooth reported volatility of private credit does not make the underlying economic risk disappear. This is also the cleaner way to think about Treasuries versus corporate bonds: corporate bonds bundle in credit spread risk, which acts like equity risk in a recession, so Treasuries are the purer growth-shock hedge.

These can be reasonable holdings. Bought as bond replacements for diversification, though, they often add more of the risk you were trying to reduce. On the map above, high yield sits beside stocks in the Goldilocks quadrant. The same hidden equity concentration is what our guide on risk parity is built to expose.

So is 60/40 dead?

No, though the expectation that it never falls with stocks should go. The traditional 60/40 has always been dominated by equity risk, as its position on the map shows. Before 2022, in the two steepest equity declines of this century, the 10-year Treasury rallied: it returned 15.1% in 2002 and 20.1% in 2008.¹⁰ The bond sleeve shrinks drawdowns and gives you something to rebalance from; it was never designed to make the portfolio immune to losses.

Treat 60/40 as a risk level. The right bond weight comes from your time horizon, spending needs, risk tolerance, and tax situation, not from a headline about correlations. The same 60/40 label can be too conservative for one investor and too aggressive for another.

Investor	What bonds are for
Young accumulator, stable income	Optional risk dial, emergency buffer, rebalancing fuel. A low bond weight can be rational.
High-income mid-career	Risk control, funding known future goals, tax-aware ballast.
Near-retiree	Managing sequence-of-returns risk before withdrawals begin.
Retiree	Spending reserve and liability matching, sized to the spending horizon.
Short-term saver (house, tuition soon)	Cash and T-bills; long-duration bonds are the wrong tool.

For the deeper allocation debate, see 60/40 vs. target-date vs. 100% stocks and the case for holding fewer bonds with a 90/10 mix. For what the bond sleeve itself is made of, and how much that changed 2008 and 2022 outcomes for identical allocations, see What a 60/40 Portfolio Really Owns.

What we recommend

Keep bonds even though correlations turned positive, and treat 60/40 as one risk level among several. A few principles hold up across the research:

Name the shock you are insuring against before buying a diversifier. Treasuries cover a demand-driven recession. A TIPS ladder covers future real spending. Commodities and trend following have done best in inflationary sell-offs, at higher cost and with less agreement in the research about which inflation they hedge.
Match the bond's job to the risk you are reducing. If you want a smoother ride and rebalancing fuel, bonds still do that. If you are worried specifically about an inflation shock, nominal bonds are the wrong tool.
Size bonds by spending horizon and risk tolerance, not by headlines about whether 60/40 is dead.
Prefer Treasuries over corporate or credit-heavy bonds for the growth-shock hedging job. Credit risk behaves like equity risk when you need diversification most.
For inflation worry, add inflation-aware assets rather than relying on nominal bonds: TIPS, short-term TIPS, T-bills, and a modest sleeve of real assets or trend-following, held alongside bonds. A TIPS ladder is the most direct tool for matching real spending. Note that a TIPS ladder and a TIPS fund are different instruments over short horizons: AQR’s inflation-sensitivity data puts a TIPS fund’s twelve-month behavior much closer to a nominal bond than most investors expect.
Mind asset location. Bond interest is taxed as ordinary income, so bond funds often belong in tax-advantaged accounts. Treasury interest is exempt from state income tax, which matters more in high-tax states than in a no-income-tax state.

Frequently Asked Questions

Are bonds still worth owning if stock-bond correlations are positive?

Yes, with realistic expectations. A positive correlation reduces the diversification benefit without eliminating it. Bonds do not have to be less volatile than stocks to reduce total portfolio risk; what matters is the bond sleeve’s volatility, its correlation with stocks, and its weight working together. In a positive-correlation regime, you simply should not count on bonds rallying in every equity drawdown.

Why did bonds fail in 2022?

The shock was inflation and rapidly rising interest rates, the one environment where nominal bonds are most vulnerable. Starting yields were near 1.5%, so there was almost no carry to offset the price losses from sharply higher rates. Bonds behaved as their exposure to inflation news predicts: a 2.4-point rise in the 10-year yield produced a 17.8% loss on the 10-year Treasury.

Which assets have done well in stagflation?

In AQR’s map of assets since 1972, the ones that did better on upside inflation surprises were commodities, energy stocks, gold, breakevens and trend-following strategies; broad stock and nominal bond indexes did worse. Each has a cost. Commodities and trend following can lag for years in calm markets, gold has been an unreliable inflation hedge over ordinary horizons,¹¹ and the trend results are backtests. Over a known spending horizon, a ladder of individual TIPS held to maturity protects purchasing power without depending on market prices along the way.

Are Treasury bonds better diversifiers than corporate bonds?

Usually yes. Corporate bonds carry credit-spread risk, which tends to widen in recessions and behave like equity risk exactly when you want diversification. Treasuries are the cleaner growth-shock hedge because they strip out credit risk.

Should I replace bonds with dividend stocks, REITs, high yield, or private credit?

Usually not as a core substitute. Those assets pay income, but they often carry meaningful equity or credit risk. AQR's equity-beta numbers show private credit and buffer products sitting much closer to stocks than to Treasuries. The extra yield pays for that risk.

Do higher yields make bonds safe?

They make bonds more attractive and provide a larger income cushion, but they do not remove duration risk. A bond fund can still post a negative year if yields rise far enough. Higher yields raise the breakeven rate rise; a larger move still produces a loss.

Are long-term bonds better than short-term bonds?

It depends on the job. Long bonds offer more upside if yields fall and much larger losses if yields rise. Short-term bonds and T-bills are better for near-term spending. Intermediate bonds are often a reasonable compromise for the core of a portfolio.

Is 60/40 still a good default?

It can be, especially for moderate-risk investors and retirees, but it is a risk level rather than a universal answer. Depending on your horizon and spending needs, 90/10, 80/20, 70/30, 60/40, or a liability-matched mix may all be reasonable.

In a hurry? See the quick BND vs TIPS comparison for the decision on one scannable page.

For the same risk taken to its limit, see Austria’s 100-year bond, where a government with impeccable credit lost 70% of its market value on a 2.4 point move in yields.

Whether rising government debt changes the case for owning Treasuries is covered in Will Government Bonds Blow Up?, which splits the 2026 sell-off into real yields and inflation expectations.

Sources

2022 returns: S&P 500 total return −18.11% and Bloomberg U.S. Aggregate −13.01%, from the benchmark rows of the iShares IVV and AGG fact sheets; the 60/40 figure is 0.6 × −18.11 + 0.4 × −13.01. CPI: BLS, Consumer Price Index news release, July 13, 2022. Fed policy rate: FOMC statement, December 14, 2022.
Alfie Brixton, Jordan Brooks, Pete Hecht, Antti Ilmanen, Thomas Maloney and Nicholas McQuinn, “A Changing Stock–Bond Correlation: Drivers and Implications,” The Journal of Portfolio Management 49(4), 2023. doi.org
John Y. Campbell, Adi Sunderam and Luis M. Viceira, “Inflation Bets or Deflation Hedges? The Changing Risks of Nominal Bonds,” Critical Finance Review 6(2), 2017, 263–301. doi.org
Roderick Molenaar, Edouard Sénéchal, Laurens Swinkels and Zhenping Wang, “Empirical Evidence on the Stock–Bond Correlation,” Financial Analysts Journal 80(3), 2024, 17–36. doi.org
Pete Hecht, Antti Ilmanen, Thomas Maloney and Nick McQuinn (AQR Portfolio Solutions Group), “Inflation Redux? Real Solutions for Real Returns,” Alternative Thinking 2026 Issue 3, September 2026, Exhibits 1 and 4. aqr.com
Benjamin Knox and Yannick Timmer, “Stagflationary Stock Returns,” Finance and Economics Discussion Series 2025-056, Board of Governors of the Federal Reserve System, 2025. Working paper. federalreserve.gov
Eugene F. Fama and G. William Schwert, “Asset Returns and Inflation,” Journal of Financial Economics 5(2), 1977, 115–146. doi.org
Ashwin Thapar, Thomas Maloney and Alfie Brixton, “When Stock-Bond Diversification Fails: Managing Inflation Risk in Investor Portfolios,” AQR white paper, October 27, 2021. aqr.com
Xiang Fang, Yang Liu and Nikolai Roussanov, “Getting to the Core: Inflation Risks Within and Across Asset Classes,” The Review of Financial Studies 39(3), 2026, 702–743. doi.org
Aswath Damodaran, “Historical Returns on Stocks, Bonds and Bills”, NYU Stern. Annual S&P 500 and 10-year Treasury total returns, used for the 1974, 2008 and 2022 replays in the calculator.
Claude B. Erb and Campbell R. Harvey, “The Golden Dilemma,” Financial Analysts Journal 69(4), 2013, 10–42. doi.org

Editor’s note

Updated October 14, 2026, to add the four-environment map from AQR’s 2026 data, the Knox and Timmer evidence on why stocks fall on inflation surprises, real returns and historical replays in the calculator, and a source list. Updated June 17, 2026. Thanks to Ryan Gavin for helpful feedback on the diversification math. An earlier version overstated the idea that bonds must be less volatile than stocks to reduce portfolio volatility; what matters is covariance, correlation, and allocation size.