Investing for the Long-Run in European Real Estate
FRB of St. Louis Working Paper No. 2006-028A
44 Pages Posted: 4 May 2006
Date Written: April 2006
We calculate optimal portfolio choices for a long-horizon, risk-averse investor who diversifies among European stocks, bonds, real estate, and cash, when excess asset returns are predictable. Simulations are performed for scenarios involving different risk aversion levels, horizons, and statistical models capturing predictability in risk premia. Importantly, under one of the scenarios, the investor takes into account the parameter uncertainty implied by the use of estimated coefficients to characterize predictability. We find that real estate ought to play a significant role in optimal portfolio choices, with weights between 12 and 44 percent. Under plausible assumptions, the welfare costs of either ignoring predictability or restricting portfolio choices to traditional financial assets only are found to be in the order of 150-300 basis points per year. These results are robust to changes in the benchmarks and in the statistical framework.
Keywords: optimal asset allocation, real estate, predictability, parameter uncertainty
JEL Classification: G11, L85
Suggested Citation: Suggested Citation