View article

This study addresses several aspects of stock selection, portfolio construction, and data mining corrections and hypothesis testing of excess returns. Mean-variance, equally actively weighted, tracking-error-at-risk, and 130/30 portfolios are created and tests are conducted to find out whether the excess returns of these portfolios are statistically different from the average models that could have been used to build portfolios. Knowledge of earnings forecasts is an important input to the portfolio construction process. The portfolios constructed fulfill a global growthmandate, and the strategies work in EAFE plus Canada and the U.S. universes. The excess returns produced by the models are statically different from the average models used. The 130/30 strategy dominates the long-only strategy. Moreover, evidence is presented to show that these portfolios can be implemented and produce excess returns in the …