Modeling index tracking portfolio based on stochastic dominance for stock selection

Abstract We propose a three-step method using the stochastic dominance (SD) approach on stock filtering to determine the number and candidate stocks in a portfolio. We empirically prove that our model can be used to efficiently construct a partial tracking portfolio and replicate the return of the index. First, the low standard deviation feature is found in the proposed portfolio using SD for the risk avoider. Second, our model generates constituents for a portfolio and fills the gap in the index tracking strategy. Third, the portfolios chosen from the SD-based model outperform the FTSE index and traditional index trackers’ returns. Artificial intelligence algorithms of weighting constituents can be examined in future research.