Unveiling Hidden Risks: How Shifts in Stock-Bond Correlations Impact Investment Portfolios

When investors are contemplating how best to achieve their financial goals, the discussion generally centres around the required return needed to meet those goals, along with how much risk an investor is willing to bear. Once those two related topics have been decided, attention then turns to portfolio construction. While one might think this is a linear process, we at Viewpoint believe this process should be viewed as an iterative feedback loop since risk (as measured by the volatility of the portfolio) is not stable through time. The correlation between asset classes in the portfolio (e.g., stocks and bonds in a 60/40 balanced portfolio) plays a major part in the risk profile of the portfolio, and not only are correlations not static, but they can vary quite considerably over time.

A recent paper called Empirical Evidence on the Stock-Bond Correlation by Molenaar et al. (2023) looks at the historical evidence for what drives the correlation between stocks and bonds, as well as how a changing stock-bond correlation impacts the overall risk of a diversified portfolio. A section of the paper looks at the correlation structure between stocks and bonds during two different periods, 1970-1999 and 2000-2022. During the first period, the correlation between stocks and bonds was +0.3 versus -0.3 in the second period. This dramatic difference in how stocks and bonds behave together has large implications for investor portfolios. If an investor in a 60/40 balanced portfolio today was to witness a return to the correlation structure of the first period (positive correlation between stocks and bonds), that investor would need to reduce their equity position by 25% (i.e., a 35/65 portfolio) for the portfolio to have the same risk as it did in the second period (negative correlation between stocks and bonds). The key takeaway here is that as the correlation between stocks and bonds changes, investors may be unknowingly taking more risk than they originally determined they were willing to bear.

The paper also discusses that since 1952, countercyclical monetary policy administered by central banks has influenced a positive stock-bond correlation during periods of high inflation, high inflation uncertainty, and high real returns on Treasury bills. At Viewpoint, we recently published a research note analyzing how large swings in the correlation between stocks and bonds can impact the risk of 60/40 balanced portfolios, while noting the current rolling correlation between the two assets is trending into positive territory due to elevated levels of macroeconomic uncertainty. In our opinion, investors need to be looking at including alternative assets in their portfolio construction to help guard against a protracted period where the correlation between stocks and bonds is positive. The inclusion of commodities in a portfolio can help fortify investment portfolios, as commodities generally do well in periods where inflation is high, which is also when the correlation between stocks and bonds turns positive and portfolio risk increases. Viewpoint’s analysis shows that including an allocation anywhere from 10-20% in a 60/40 balanced portfolio can help to improve risk-adjusted returns when the correlation between stocks and bonds increases. Furthermore, large shifts in correlation have historically persisted for decades at a time. While we won’t know with certainty whether the most recent move in correlations will persist, it behooves investors to look at how to reinforce their portfolios should we be entering a new macroeconomic regime.