The 60-40 Portfolio: Did It Ever Work?

The traditional 60-40 portfolio has served investors well over the past decades. Or at least so investors and fund managers like to believe. However, our research finds that investors would be better off investing in a market portfolio that includes alternative assets as opposed to a standard 60% equity and 40% bond portfolio. Further, if financial advisors focused more on measuring investors’ risk aversion and realized utility instead of on security selection, they would generate more value for their clients.

The origins of the 60-40 portfolio go back to the works of Harry Markowitz who researched the foundations of modern portfolio theory and found that the optimal asset allocation is some mix of the risk-free asset and the market portfolio depending on the investor’s risk aversion, and that risk could be reduced through diversification by holding several different asset classes. The 60-40 portfolio is an attempt commonly used in practice to capture this diversified asset allocation. With an annualized return of 10% since 1973 and a Sharpe ratio (return per unit of risk) of 0.42, the portfolio has done better than a 100% investment in global equities or in a pure bond portfolio which generated Sharpe ratios of 0.31 and 0.27 respectively while the annual risk-free rate (annualized T-bill returns) returned 5.1%. Based on these numbers the 60-40 portfolio looks like a good deal. What’s more, is that the industry classic is easy and cheap to implement; all you need is one well-diversified all-equity ETF and U.S. Treasuries, both of which are easy to purchase and require very low fees.

However, the 60-40 portfolio is merely a heuristic approach by practitioners to present investors with a rule of thumb portfolio that was created during the 1950s when calculating important metrics such as variances, correlations, covariance matrices and risk aversions could take days due to the computational restraints at the time. Second, this standard portfolio is an outdated one size fits all approach that gives different types of investors the same asset allocation regardless of their true ability and willingness to bear risk. But measuring risk aversion, utility and large-scale variance-covariance matrices are non-trivial tasks so we are not about to blame investors for using a heuristic solution to complex investment problems. As one fund manager confessed: “when choosing the asset allocation for my own retirement plan I should have calculated the efficient frontier of all available securities, measured my risk aversion and chosen the optimal weight in the market portfolio given my risk aversion. I opted for the 60-40 portfolio instead”. Third, the traditional 60-40 portfolio is often implemented in practice using a diversified equity index such as the ACWI combined with a global bond index such as Barclay’s Global Government Bond index. This, however, is a misrepresentation of the analytical solution to portfolio theory provided by Harry Markowitz. Instead, investors should be buying some combination of the market portfolio and the risk-free asset. In theory, as opposed to practice, the market portfolio should include all risky assets available, not just stocks but also commodities, currencies, real estate indices and other exchange-traded asset classes while the risk-free asset is most appropriately proxied by investing strictly in short-term U.S. Treasuries such as T-bills and not in long-term international government bonds.

To show that the 60-40 heuristic solution is not optimal, we found that both a 50-50 and a 40-60 portfolio returned a higher Sharpe ratio (0.440 and 0.441 respectively) than the 60-40 portfolio (0.42). This would imply that the traditional portfolio has failed risk-averse investors as they would have been better off investing in a 40-60 portfolio. Since bonds offer a lower standard deviation and a generally negative correlation to stocks, investors could benefit even further from selecting a strategic allocation tilted more heavily toward bonds than in their standard 60-40 portfolio if the risk-reward tradeoff is truly what they are most concerned with. An investor with a long investment horizon might not be as concerned about short term volatility from risky assets which would be expressed through different levels of risk aversion and utility that portfolio managers should take into account.

This lack of understanding of the investor’s true risk aversion and utility is the main reason why the 60-40 portfolio is a poor one-size fits all solution that does not adequately cover the needs of different investors. A young investor with a long investment horizon but low levels of wealth and income will require a significantly different asset allocation than an older but wealthier investor. The 60-40 portfolio is not able to meet the demands of either of those investors appropriately. Our calculations support the implications of modern portfolio that investors with low risk aversion prefer much higher allocations to stocks while investors with high risk aversion prefer higher allocations to bonds than the 60-40 would offer them. In general, it seems that the 60-40 portfolio only maximizes utility for moderate levels of risk aversion while investors with high and low levels of risk aversion do not maximize their welfare from the 60-40 allocation. We suggest two different solutions to find an allocation that better serves the client’s needs.

One alternative is for advisors to approximate the client’s utility function to determine his level of risk aversion. Knowing the client’s risk aversion will in turn allow the advisor to determine the optimal asset allocation that maximizes the client’s welfare on a case by case basis. There are several different utility functions ranging from highly sophisticated regression models to simple one-factor models. We suggest using a simple concave function that measures utility in relation to wealth or income, or both combined with the assumption that risk aversion generally increases as absolute levels of wealth increase. This implies that wealthier investors prefer a higher allocation to bonds while less wealthy investors prefer stocks. Under this approach, we find that very wealthy investors (>$10m and above) satisfy their risk-return desires with a 40-60 stock-bond mix. At wealth levels of $1m to 10$m investors maximize utility at a 55-45 allocation. As wealth decreases below $1m, the optimal allocation to stocks surpasses the 60-40 mix and approaches a 70% weight to stocks. This leads us to the conclusion that the optimal allocation to stocks for investors of average wealth seems to be in the 65% to 70% range. However, wealth is not the only factor affecting risk aversion and thereby stock-bond allocation. Once the investor’s time horizon is taken into account as well, we find that the preferred allocation to stocks can go far above the 70% or below the 40% threshold. In general, the effect of time horizon on risk aversion is the opposite of that of wealth. As the time horizon increases risk aversion increases as well. Since younger investors are generally less wealthy our model suggests that those investors would benefit from an allocation to stocks that approaches 100% while wealthier investors that generally have a shorter time prefer an allocation to stocks in the 20% to 25% range. 

A second alternative closer to the original works of modern portfolio theory is to only measure the client’s risk aversion to derive the optimal portfolio weights between the maximum Sharpe ratio portfolio (market portfolio) and the risk-free asset (U.S. short-term T-bills). As mentioned above, most 60-40 portfolios currently invest in an all-equity index and a global government bond index but these choices are not ideal to proxy the true market portfolio and the risk-free asset. Under perfect circumstances, the market portfolio takes all tradeable risky assets into consideration, not just stocks. This means that alternative assets such as currencies, commodities and real estate should be added to stocks to create a portfolio with a higher Sharpe ratio than an all-equity index. Adding commodities as an alternative asset class, and gold as an inflation hedge to our analysis, we can create a market portfolio that generates a higher Sharpe ratio than the common 60-40 portfolio (0.52 vs. 0.42). Given a certain risk aversion level of a client, it is then possible to analytically solve for the optimal allocation weight to the new market portfolio. Allowing for short selling we find that the optimal allocation for investors with low or neutral risk aversion far exceeds 100%. This evidence suggests that the traditional 60-40 portfolio is oftentimes very different from the asset allocation that would generate the maximum welfare for the investor. We prefer this second method to the first alternative described above.

To conclude, the 60-40 portfolio has rarely been the optimal portfolio allocation for investors who seek to maximize their risk-return tradeoff. It can, at best, be viewed as a simple approximation for investors who do not have the time or capacity to find the optimal asset allocation based on their risk aversion. We provided two alternatives to improve the portfolio allocation and tailor it to the individual needs of the investor. Ideally, we believe that advisors and investors should focus more on finding the global maximum Sharpe ratio portfolio by including alternative asset classes and then decide what weight to assign to this portfolio based on the investor’s risk aversion rather than screening for individual securities to buy and sell. This would allow the investor to achieve the maximum risk-return reward while also maximizing his level of welfare.

Contact us if you want to learn more about our latest views on Portfolio Allocation or our other market insights.

March 2021