Using a Bootstrap to Measure Optimum Mixed‐Asset Portfolio Composition:
改进了Liang等人提出的自助法，发现低风险偏好的投资者应持有很少的房地产，与原有结论不同。
Liang, Myer and Webb (1996) have offered bootstrap simulation as a tool for quantifying the uncertainty in the optimum composition of portfolios. Unfortunately, the confidence intervals produced were so large, they were unable to provide any new insight to the question, “How Much in Real Estate?”. In this comment, adjustments have been made to the methodology they proposed and as a result have produced findings which lead to very different conclusions. More specifically, the results suggest that investors with a low risk preference should hold very little real estate.