We conducted the portfolio optimization on the selected benchmarks for nine asset classes with a time range starting from January 2007 to December 2016 in Canadian Currency, to prove whether the mean-variance approach by Markowitz (1952) combined with a covariance matrix blended from a quiet time and a turbulent time as introduced by Chow, G., Jacquier, E., Kritzman, M., and Lowry, K. (1999) is still valid with recent years’ data.As a result, the optimal portfolios with different covariance matrices blended from turbulent and quiet periods have shown sensitivity of optimal weights to both possibilities of occurrence for the turbulent and quiet periods, and different risk aversion to turbulent and quiet periods. The outlier-sample optimal portfolio is the most conservative one and provides a lowest expected return. Besides, the optimal weights of Cash are much higher due to the higher volatilities of ours benchmarks for US equity, emerging market equity, US bonds, high-yield bonds, and commodities.
MSc in Finance Project-Simon Fraser University.
Copyright is held by the author(s).
Member of collection