The Fed & the Bond Market
The Taylor Rule helps explain the recent failure of bonds to offset stock losses and when that might end
By: Johann Colloredo-Mansfeld, Greg Obenshain, and Dan Rasmussen
The US stock market fell in the first quarter of this year, but the US bond market fell further. This is unusual: when the stock market is down, the bond market has only been down more than stocks in 4% of months, most of which were in the 1970s. The Wall Street Journal is calling it "the worst bond market since 1842," noting that bonds have lost 10% this year, with long-term Treasurys down 18% on the year through April 30.
The cause of this simultaneous sell-off is the Federal Reserve reacting too late to a sharp increase in inflation. The Fed believed inflation would be transitory and relied on inflation expectations from surveys and the TIPS markets that turned out to be too anchored in the pre-COVID deflationary period.
The best way to understand and contextualize what’s going on in the bond market is through the work of Stanford economist John Taylor, the eponymous inventor of the Taylor Rule. The Taylor Rule starts with a target rate equal to the estimated long-term real rate plus current inflation and then adjusts it upward if inflation is above target or real growth is above trend and adjusts it downward if inflation is below target or real growth is below trend. Figure 1 shows the Taylor Rule versus the Fed funds rate over the last 12 months. Since March 2021, the gap between the rate specified by the Taylor Rule and the Fed funds target rate has widened more than 600bps. The graph also shows 10-year Treasury yields, which have risen more in line with the Taylor Rule.
Figure 1: Taylor Gap, 10Y Treasury Yield, Taylor Rule, monthly, LTM
Source: Bloomberg, FRED, Atlanta Fed
The Fed’s delayed reaction meant that the Fed began raising rates when the economy had already switched from Quadrant 2 (rising inflation, rising growth) to Quadrant 3 (falling growth, rising inflation). Usually, the Fed raises rates in Quadrant 2, which is when bonds typically have the worst performance. Raising rates when growth has already started to decelerate creates significant dangers for the economy, which is one of the reasons why stocks have also been selling off.
This divergence between the Taylor Rule and the Fed’s target rate is unusually large. Through most of the late 1980s and 1990s, the Fed target rate closely followed the Taylor Rule, according to Taylor’s own research. Taylor believes the only periods after the 1970s when the Fed again failed to react sufficiently to changes in growth and inflation were the mid-2000s and today. We can see this by looking at the historical gap between the Taylor Rule and the Fed funds effective rate.
Figure 2: Gap between the Taylor Rule and the Fed Funds Target Rate
Source: Bloomberg, Atlanta Fed
We can look in depth at the previous periods when the Taylor gap was as significant as it is today to try to understand what these historical periods might imply about today. Below we show five periods: the late 1960s, two periods during the '70s, the mid-2000s, and today.
Figure 3: Credit Spreads, Fed Funds Rate, Taylor Rule, and Select Asset Class Returns
Source: Bloomberg, FRED, Atlanta Fed
On average, these periods have lasted 32 months. The Fed funds target rate has started at 3.3% and risen by 5.6%. The yield on the 10-year Treasury has increased by 1.9%, and high-yield spreads have moved up by 1.4%. During these periods, Treasurys outperformed equities, and gold outperformed Treasurys. Within Treasurys, the shorter duration of five-year Treasurys outperformed.
Thinking of where we are today in the context of these prior periods, the Fed has a long road ahead to bring rates up to where they likely will need to be. The 10-year Treasury rate typically rises much less than the Fed funds rate, but Treasury yields have risen much more than the Fed funds rate and so may have more limited downside. Gold looks like an attractive hedge if past is prelude.
In the past two months, we see that Treasurys and equities have underperformed relative to gold. Figure 4 shows the returns of five-year Treasurys, 10-year Treasurys, the S&P 500, and gold over the past two months relative to the average annualized returns of our four comparable periods: 1965-1969, 1972-1974, 1977-1980, and 2004-2006.
Figure 4: Asset Class Returns in Current Hiking Cycle vs. Average Annualized Returns of Historically Comparable Periods
Source: Bloomberg, FRED, Atlanta Fed
While relative performance among asset classes roughly aligns with what has played out historically, it would appear that the market has oversold Treasurys, while gold has underperformed relative to history.
When studying historically comparable periods, we found that Taylor gaps of this magnitude exert a significant force on markets, wreaking havoc with some common cause-effect relationships.
Treasurys tend to rally when high-yield spreads are rising. But when the Taylor gap is in the top quartile, Treasurys tend to fall despite rising spreads. The figure below shows us that forward quarterly 10-year Treasury returns approach zero when spreads are falling. Conversely, when spreads are rising, quarterly forward returns approach 1.0%. These values are bounded by 95% confidence intervals (the vertical bars), which provide ranges of historically probable returns. Because the two ranges are not overlapping, we can have confidence that credit spreads offer a robust signal for differentiating between environments that favor Treasurys.
Figure 5: Change in HY Spread vs. 10Y Treasury 3M FWD returns, Monthly, 1954 – Present
Source: Bloomberg, FRED
But when we look only at periods when the Taylor gap is in its highest historical quartile, we see a much murkier relationship, albeit with the introduction of hindsight bias.
Figure 6: Change in HY Spread vs. 10Y Treasury 3M FWD Returns with Taylor Gap in Highest Historical Quartile, monthly, 1954 – Present
Source: Bloomberg, FRED, Atlanta Fed
When the Taylor gap is in its highest quartile, changes in credit spreads offer little predictive information and Treasury returns are negative. In fact, when the Taylor gap is in its highest quartile, its effect overwhelms that of credit spreads.
So what is an investor to do when Treasurys are not the safe-haven asset because yields are rising, and when do these periods end? Within bonds, history shows that shorter-duration bonds outperform longer-duration bonds under these conditions. Figure 7 shows the relative performance of 10-year and five-year Treasurys in rising spread environments depending on the level of the Taylor gap.
Figure 7: 3M FWD 5Y Treasury Excess Returns over 10Y Treasurys with Rising HY Spreads by Taylor Gap, Monthly, 1962–2022
Source: Bloomberg, FRED, Atlanta Fed
When spreads are rising and the Taylor gap is outside its top quartile, 10-year Treasurys solidly outperform 5-year Treasurys. However, when spreads are rising but the Taylor gap is in its highest quartile (Taylor Gap = Q4), five-year Treasurys outperform 10-year Treasurys by an average of 0.5% per quarter. These results more broadly confirm what we see in the two comparable periods we highlighted above.
Beyond altering Treasury duration, investors could substitute gold for their credit exposure. Gold already performs well when spreads are rising: three-month forward returns to gold are ~1% higher when high-yield spreads are rising than when they are falling. The below chart shows the difference in excess returns of Treasurys over gold in rising spread environments, depending on the level of the Taylor gap.
Figure 8: 3M FWD Gold Excess Returns over 10Y Treasurys with Rising High Yield Spreads by Taylor Gap, Monthly, 1954–2022
Source: Bloomberg, FRED, Atlanta Fed
When spreads are rising but the Taylor Gap is in its highest quartile (Taylor Gap = Q4), gold outperforms Treasurys. In fact, Treasurys underperform gold by an average of ~4% per quarter. However, when spreads are rising and the Taylor Gap is outside its top quartile, Treasurys outperform gold.
With unemployment low and inflation high, the Fed has little choice but to react. But 25bps in March and 50bps in May is clearly not enough. Previous periods of easy Fed policy have ended with both Fed tightening and slowing inflation and growth, both of which narrowed the Taylor gap. Hoisington Investment Management, who have been right about bonds yields for decades, discussed the Taylor Rule in detail in their most recent letter. They see signs of impending recession as the Fed tightens but end their letter with a caveat: “should the Federal Reserve cease in their efforts to calm inflation before it has been fully restrained, bond investors should be wary.”
In sum, the Taylor gap highlights an important nuance for 60/40 investors. When stagflation strikes, the stock-bond approach to diversification fails. When the Taylor gap shows that the Fed is behind the ball on raising rates, it makes sense for investors to shift toward shorter-duration credit and to substitute some of their safe-haven asset exposure to gold, which has historically performed well in these environments. The past month has been no exception to this observation. Gold has outperformed Treasurys by a wide margin.