Bonds Still Belong: What the Data Actually Says

The narrative surrounding fixed income has turned sharply negative. With US borrowing costs reaching their highest levels in nearly 25 years, concerns about inflation, fiscal sustainability, and the equity-bond correlation have converged into a broadly held view that bonds are finished as a portfolio tool. Peter Weidner, Tom Frith, and Tarek Abou Zeid at Man Group's Systematic team push back on that conclusion — rigorously, and with more than a century of data behind them. Their analysis, published October 6, 2026 1, makes a case that is harder to dismiss than the current consensus would suggest.

The Diagnosis: Elevated Is Not the Same as Rising

The crux of the Man Group argument is a distinction that most of the current commentary collapses: the difference between elevated inflation and rising inflation. Weidner et al. are direct on the point. "Elevated inflation alone is not bad for Treasuries," they write. "It takes rapidly rising inflation to drive bonds down."

To support that claim, the team conducted a rolling three-month analysis of US Treasury returns going back to 1900, sorting observations by both the level of inflation at the start of each period and how much inflation changed by the end. The findings disrupt the prevailing narrative. When inflation has been in the 2% to 4% range, as it is today, "10-year US Treasuries have historically delivered an average annualized return of 4% above inflation." That is a real return, above inflation, in the moderate band that currently describes the US economy.

Where the Real Risk Lives

The heatmap analysis does identify a danger zone. It is specific: inflation starting above 4% and rising by more than 50 basis points over three months. In that scenario, "real returns to 10-year Treasuries have averaged -11.7% annualized." 2022 was precisely that scenario. Deeply negative real yields at the start of that year compounded the damage, creating the historic drawdown that has since anchored the bearish case against bonds.

The team does not dismiss that episode. They use it to define the actual risk condition, which is not "inflation is elevated" but "inflation is high and accelerating." The distinction matters enormously for portfolio construction. "A 'higher for longer' world," they note, "does not require inflation to keep rising at pace." Sitting in the moderate inflation band today, with stable inflation implying a 4% annualized real return to 10-year Treasuries, the bear case requires an additional deterioration in inflation dynamics that is not yet present.

Even the high-and-stable category, where inflation starts above 4% but moves by no more than 50 basis points in either direction, "has historically delivered an average annualized nominal return of 8.3% and a real return of 2.3% for 10-year US Treasuries." The data is not equivocal on this.

Diversification Still Requires Bonds

Beyond the inflation mechanics, Weidner, Frith, and Abou Zeid anchor a second argument in the equity-bond correlation. Bonds have historically performed well when equities fall. That relationship is under scrutiny, but the team's position is that abandoning a strategic allocation based on recent volatility is premature. "We don't believe you should write off bonds just yet," they state plainly. "They have historically done well when equities fell, and far from being doomed in a 'higher for longer' world, they can still deliver positive real returns."

For managing inflationary tail risk, the team points to portfolio complements rather than replacements. Commodities alongside equities and bonds improve diversification and reduce inflationary exposure. Momentum strategies, which the team has documented in prior research as among the more resilient approaches in inflationary periods, offer another layer of protection. These are additions to a bond allocation, not substitutes for one.

Five Key Takeaways for Advisors and Investors

  1. The current inflation environment is not bond-hostile. At 3.4% CPI, the US sits in the moderate inflation band, where historical data shows average annualized real returns of 4% on 10-year Treasuries when inflation is stable.
  2. The real danger is rapid acceleration, not elevated levels. Bonds suffered in 2022 because inflation surged past 4% and kept rising fast. That specific dynamic, not elevated inflation per se, is the condition to monitor.
  3. Higher for longer does not mean worse for bonds. Sustained elevated inflation without a sharp upward move has historically still produced positive nominal and real returns in Treasuries.
  4. Commodities and momentum strategies complement, not replace, bonds. For advisors concerned about inflationary tail risk, layering in commodities and trend-following strategies alongside a bond allocation provides diversification without abandoning the asset class.
  5. Abandoning bonds based on recent narrative is a positioning error. The bearish case requires continued inflation acceleration from current levels. Absent that condition, a strategic bond allocation remains historically justified on both a nominal and real return basis.

 

Footnote:

1 Weidner, Peter, Tom Frith, and Tarek Abou Zeid. "Don't Write Off Bonds Yet." Man Group Insights: Views from the Floor, 6 Oct. 2026, https://www.man.com/insights/views-from-the-floor-2026-6-oct.

Total
0
Shares
Previous Article

Tech Is the Everything Cycle: Andreessen Horowitz Maps the State of Markets

Next Article

There's a Version of You Your Family Has Never Met

Related Posts
Read More

Four Stocks, One Yield, and the October 15 Reload

Twenty strategists on the U.S. equity market, and the two lines that decide the fourth quarter