透過您的圖書館登入
IP:18.118.1.158
  • 期刊

On the Predictability of Stock Market Returns: Evidence from Industry-Rotation Strategies

摘要


This paper evaluates historic, Bayes-Stein, Capital Asset Pricing Model (CAPM) and dividend-yield riskfree-rate estimators of asset means using statistical and economic criteria. None of the estimators exhibit much in the way of out-of-sample predictive ability when judged by statistical criteria. Yet, when combined with a discrete-time power-utility portfolio selection model, all the estimators generate economically significant returns judged in terms of compound return - standard deviation plots and accumulated wealth. Even so, the portfolios generated from dividend-yield riskfree-rate estimators perform by far the best and portfolios generated from traditional CAPM estimator perform the worst. For the most part, commonly accepted statistical measures of investment performance support these rankings.

關鍵字

無資料

延伸閱讀