The Impact of Serial Correlation on Efficient Portfolios
When it comes to investments, the assumption that returns follow a perfect “random walk” is not entirely accurate over time. This contrasts with traditional portfolio construction methods like mean variance optimization (MVO), which typically assume returns are independent and identically distributed (IID).
In a recent study by the CFA Institute Research Foundation, it was shown that serial dependence can significantly influence efficient portfolios for investors with different time horizons. The focus was on the optimal allocation to six risk factors: size, value, momentum, liquidity, profitability, and investment, and how it varies by investment horizon.
A Quick Visit to the Factor Zoo
Factors are designed to capture returns from specific investments while controlling for market risk. Factors like size, value, momentum, liquidity, profitability, and investment offer unique insights into market dynamics.
Among these factors, size and value become more attractive over longer periods, while momentum and profitability lose some appeal. The evidence for liquidity and investment factors presents a mixed picture, emphasizing the importance of considering serial correlations in portfolio construction.
Wealth Growth Over the Long Run
Analysis of rolling five-year cumulative returns for the factors from 1964 to 2023 reveals significant differences in performance. Historical returns for factors like SMB and HML show periods of both outperformance and underperformance, highlighting the potential diversification benefits of allocating across factors.
Portfolio Optimizations
To understand how optimal factor weights vary by investment horizon, portfolio optimizations were conducted using risk aversion coefficients that reflect different risk tolerances. Results indicate that factors like SMB and HML become more attractive over longer horizons, while momentum and profitability factors lose appeal.
Conclusions
Serial dependencies within factors have a significant impact on portfolio efficiency. Ignoring these dependencies could lead to suboptimal allocations. By considering serial correlations, investors can build more effective portfolios that adapt to changing market conditions.
References:
Fama, E., & French, K. (1992). “The Cross-Section of Expected Stock Returns.” Journal of Finance.
Fama, E., & French, K. (2015). “A Five-Factor Asset Pricing Model.” Journal of Financial Economics.
Feng, G., Giglio, S., & Xiu, D. (2020). “Taming the Factor Zoo: A Test of New Factors.” Journal of Finance.
Jegadeesh, N., & Titman, S. (1993). “Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency.” Journal of Finance.
Pastor, L., & Stambaugh, R. (2003). “Liquidity Risk and Expected Stock Returns.” Journal of Political Economy.
[1] Size and value factors were included for completeness.
[2] https://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html
[3] https://faculty.chicagobooth.edu/lubos-pastor/data