Portfolio Implications of a Positive Stock-Bond Correlation World
While simultaneous large declines in stock and bond prices are likely temporary, a positive stock-bond correlation regime may persist.
The venerable 60/40 stock-bond portfolio has had a historically bad run thus far in 2022. The sharp year-to-date decline in stock and bond prices has pushed measures of correlation to 50y highs and well into positive territory for the first time in more than 20 years.1 This has prompted some to proclaim the demise of the 60/40 portfolio.
We disagree. Despite the challenges of a positive correlation environment, investors ought not learn the wrong lesson. Even in a positive stock-bond correlation world, diversified portfolios still have a critical role to play.
True, when correlation is positive, the loss of bonds as an implicit hedge for stocks leads to increased portfolio volatility and worse tail outcomes. But don’t forget Finance 101! The virtues of diversification remain intact even if assets are positively correlated, so long as they are not perfectlycorrelated (i.e., a correlation of +1).
Judging by the 1965-2000 period when US stock-bond correlation was last persistently positive, typical correlation readings are in the +0.25 range – a far cry from +1 and far more moderate than current correlation levels – leaving plenty of room for diversification benefits. In fact, the 60/40 portfolio delivered higher ex-post excess returns during the positive correlation regime of 1965-2000 than it did during the negative correlation regime of the last 20 years.
One other noteworthy historical market feature is that even during periods of positive correlation, the hedging properties of bonds came to the rescue just when they were needed most. Think about 1987 and 1998, both well ensconced within the long 1965-2000 positive correlation regime. Yet in both instances, when stock declines were deepest, bonds rallied and helped to hedge stock losses.
How should investors respond to positive correlation? To be sure, ex-ante portfolio performance will deteriorate, reflecting the riskier environment. Even so, a balanced portfolio remains optimal. For many investor types (e.g., a mean-variance optimizer) the optimal allocation when correlation is positive is not that dissimilar to when correlation is negative. As such, being wrong about future correlation is not much worse than being right. With no urgency in making a call about future correlation, a slow, steady, and even skeptical approach seems warranted. Assuming stock-bond correlation of zero may be the best multi-purpose strategy until correlation uncertainty resolves.
In a regime where the diversification properties of bonds are, admittedly, diluted, a popular remedy is to add additional assets to the portfolio mix. If included in small amounts (<10ppt), assets like commodities, energy, or precious metals push out the efficient investment frontier ever so slightly. However, improvements are modest at best because commodities offer returns that are equity-like but at significantly higher volatility. Private assets may be more promising on this score, offering equity-like returns with significantly lower reported volatility and less liquidity risk than is commonly thought.2
Even if changes in the macroeconomic landscape (unsustainable fiscal policy and less independent, less rules-based monetary policy are top-of-mind concerns) lead to persistently positive stock-bond correlation, the benefits of a balanced portfolio of stocks and bonds will endure. Negative correlation may be dead, but long live balanced portfolios!