Unless we are ready to settle for lower returns in our multi-asset portfolios, paying attention to the evolution of stock-bond correlations and knowing how to position around it is the key.
Global Market Strategist
- After a sharp sell-off this year on the back of the Federal Reserve’s (Fed’s) hawkish pivot, we believe the rates market has already reflected the risks of yields rising further
- Adding credit opportunistically is one way to enhance the “40” in the 60/40 portfolio. We like short-to-intermediate-duration U.S. investment grade bonds for their defensive qualities should we get a sharp slowdown
- Investors will be better served diversifying their U.S. equity allocation to other geographies like Asia ex-Japan where the policy backdrop, valuations and earnings expectations are looking more favorable/reasonable
The disappointment in the 60/40 portfolio has been palpable amongst multi-asset investors this year. After two decades of being the trusted diversifier to its more volatile other half, i.e., equities, investors were hurt by the breakdown in correlation between the two, with both government bonds and equities selling off at the same time. Indeed, the drawdown has been painful, with a 60/40 portfolio tumbling 20.4% in 1H22. But if history is any guide, the 60/40 has only registered negative returns in a calendar year over the past 30 years.
Is the 60/40 portfolio dead? Is it merely resting? Or does it require a makeover? These are just some of the many questions that underpin broader investor’s worries around possibly entering a new regime of positive stock-bond correlations. A few hopeful ones still believe that this is a function of the environment and should normalize once things settle down. We are inclined to believe that it is likely the latter rather than the former, and a closer look at the key macroeconomic drivers of stock-bond correlations will help to shed some light on our perspective.
Why do we even care about stock-bond correlations in the first place? Sure, positive stock-bond correlations may mean less diversification for a 60/40 portfolio but more importantly if it persists, it can mean structurally higher volatility. Assuming an investor’s risk tolerance has not changed significantly, this has implications for the stock-bond allocation in the form potentially decreasing one’s equity allocation to keep expected portfolio risk the same. Allocation changes biased to a smaller equity allocation will naturally also have an impact on expected returns. Unless we are ready to settle for lower returns in our multi-asset portfolios, paying attention to the evolution of stock-bond correlations and knowing how to position around it is the key.
We believe that growth and inflation are the two biggest macroeconomic drivers that feed through to policy decisions. This then has ramifications for the path of the risk-free rate and hence impacts risk asset pricing. Understanding how equities and government bonds behave across different growth and inflation regimes is a foundational step to understanding stock-bond correlations. To do that, we look at return/risk profiles of equities and government bonds during the periods of rising growth/inflation and falling growth/inflation versus their longer-term averages since the early 1970s in Exhibit 1a.
Exhibit 1a: U.S. Equities and Government Bonds Ann. Return/Volatility vs. Full Period Averages
Exhibit 1b: Rolling 12-month Inflation/Growth Volatility vs. Full Period Average
Source: Bloomberg, J.P. Morgan Asset Management. Analysis starts from 01/31/1973 – 06/30/2022. S&P500 returns are used for U.S. equities and Bloomberg U.S. Agg Treasury returns are used for government bond returns. Rising/Falling growth defined as monthly changes in PMI New Orders of more than +/- 0.1. Rising/Falling inflation defined as monthly changes in y/y CPI of more than +/- 0.1%. Data are as of June 30, 2022.
We use the U.S. Manufacturing New Orders as a high frequency proxy of our growth indicator and year-over-year changes in the Consumer Price Index as our inflation indicator. Three key things to note:
1) Across growth regimes, the relative return/risk profiles of U.S. equities and government bonds relative to full-period averages are directionally different. Equities much prefer a period of rising growth while bonds tend to sell-off on the expectation that an overheating economy will cause the Fed to tighten monetary policy
2) Across inflation regimes, the relative return/risk profiles of U.S. equities and government bonds relative to full-period averages are directionally the same. Both asset classes perform poorly (risk-adjusted) in periods of rising inflation, especially bonds as cashflows are nominal and fixed, which leads to an erosion of real return
3) Bonds hate a rising inflation environment marginally more than a rising growth regime. While equities thrive better in rising growth, the deterioration in risk adjusted returns relative to averages in times of rising inflation is a little less bad than bonds as ultimately some sectors are more well-positioned to pass on higher costs to the end-consumer
Exhibit 1b is an attempt to track if an inflation regime is more dominant relative to a growth regime across time. As you can see, inflation volatility relative to growth volatility has spiked versus how their historical relationship has been. The spike has been particularly sharp and has reached levels that are comparable with other periods where we observed high inflation volatility. The interesting observation across these episodes is that we also tend to see a relatively steep reversion as elevated inflation volatility recedes.
While shorter-term inflation volatility has spiked, it is important to note that longer-term inflation expectations have remained incredibly well-anchored with the 5-year-5-year inflation swap is trading between ~2.4% to 2.6%. Unless we expect inflation and accompanying monetary policy uncertainty to persist into the future, a more prolonged shift towards a regime with positive stock-bond correlations versus what we had experienced over the last two decades is probably unlikely. Hence, we think that the setback in the 60/40 is temporary.
Part of the thesis around the 60/40 regaining its rigor centers around government bond valuations. After a sharp sell-off this year on the back of the Fed’s hawkish pivot, we believe the rates market has already reflected the risks of yields rising further. Let’s also not forget that the economic outlook for the U.S. is rapidly dimming with cracks starting to form in the housing market, a large fiscal drag as well as a strong dollar.
As we move from a “There Is No Alternative” regime to “There Are Reasonable Alternatives” regime, select opportunities in credit are starting to look interesting. Specifically, we like short-to-intermediate-duration U.S. investment grade bonds for their defensive qualities should we get a sharp slowdown. While spreads can still widen in a recessionary scenario, the component of returns that stems from interest rate duration typically still dominates the overall return experience in investment grade. Adding credit opportunistically is one way to enhance the “40” in the 60/40 portfolio.
We believe that investors will also be better served diversifying their U.S. equity allocation to other geographies like Asia ex-Japan, where the policy backdrop, valuations and earnings expectations are looking more favorable/reasonable.
For the longer term, considering an allocation to alternatives is also a good way to maintain/enhance expected returns while achieving a desired volatility and diversification profile if stock-bond correlations do break down again.