Enhanced indexing using a discrete Markov chain model and mixed conditional value-at-risk

Enhanced indexing (EI) is a passive investment strategy that seeks to perform better than the benchmark index in the sense of higher return. The purpose of enhanced indexing is to determine optimal portfolios with the maximum excess mean return over the index return. The less efficient markets offer scope for enhanced indexing. The less (more) efficient the market is, the greater (lesser) is the chance of beating it. In this study, a two-step procedure is proposed for enhanced indexing of the Tehran Exchange Dividend and Price Index (TEDPIX). In the first step, a discrete Markov chain model is designed to filter stocks based on their high probability of gain over the benchmark index. In the second step, optimal weights are assigned to the filtered assets by maximizing the STARR ratio with MCVaR. The sample includes weekly data from March 2013 to March 2020. The data is divided into a 26-time frame, including 52 in-sample data and 12 out-of-sample data. The results of 26 window (containing a rolling data set of 52 weeks in- sample data 12 weeks out-of-sample) show that not only the portfolio return positively correlated to the TEDPIX return and could track it entirely, but also it could exceed and enhance the portfolio tracking. More precisely, our model portfolio could grow 13.65 times while the TEDPIX grows just 6.5 times simultaneously.