Optimal Portfolio Construction on Russian Equity Market

This diploma work aims to construct an optimal investment portfolio on Russian Equity market. The performance of this portfolio is to be compared then with that of the market index (MICEX10 index) on out-of-sample interval and corresponding conclusions are to be drawn regarding the weak-form efficiency of the Russian Stock market. In order to complete this task, it was decided to form a portfolio consisting of assets that enter the MICEX10 index, since these represent 10 most liquid and stable assets traded on Moscow Stock Exchange. The whole optimal portfolio construction problem is divided into four stages. Firstly, an optimal ARMA-GARCH model is constructed for each of the 10 stocks entering the MICEX10 index and for the MICEX10 index itself. Based on obtained model results one-step-ahead forecasts are generated then for the period from 1st January 2017 to 31st May 2017. Forecasted values of stock returns, conditional variance and standardized residuals are extracted for further analysis. At the second stage, time varying conditional covariance matrix is estimated based on constant conditional correlation model. This is done in order to estimate forecasted portfolio variance used in portfolio optimization problem. At the third stage an optimization problem is solved on a daily basis on the interval from 1st January 2017 to 31st May 2017 based on forecasted values of stock returns and return variances in order to obtain optimal weights with which stocks enter the portfolio. Finally, based on realized values of stock returns, daily portfolio return is calculated on the interval from 1st January 2017 to 31st May 2017. Return on optimal portfolio is compared then with that of the MICEX10 index. Empirical results of this research paper showed that the optimal investment portfolio did not statistically outperform the market index on the interval from 1st January 2017 to 31st May 2017. The evidence of weak-form efficiency of the Russian Stock market was inferred.