An Empirical Case Against the Use of Genetic-Based Learning Classifier Systems as Forecasting Devices

摘要: We adapt a genetic-based learning classifier system to forecast evaluation exercise by making its key parameters endogenous and taking into account the need of convergence algorithm, an issue usually neglected in literature. Doing so, we find it hard for algorithm beat simpler ones based on recursive regressions random walk forecasting stock returns. then argue that our results cast doubts plausibility using systems represent agents process expectations formation, approach commonly found agent-based computational finance