Applied Optimization in Global Efficient Portfolio Construction Using Earning Forecasts

The authors analyze the performance of an optimal global portfolio constructed from a 10-factor stock selection model that includes earning forecasts. Under the Markowitz mean–variance framework, applied optimization techniques are used to address the practical issues of risk tolerance, turnover, and tracking error. The impact of these practical constraints on portfolio performance is analyzed through extensive numerical experiments. The analysis finds that, subject to the same set of practical constraints, the risk-adjusted return of the efficient portfolio of the global equity universe outperforms that of the domestic equity universe. Active portfolio return increases with investors’ risk tolerance and the systematic tracking error allowed against the benchmark portfolio. TOPICS:Portfolio construction, factor-based models, performance measurement