Statistical Arbitrage Strategy Based on Factor Models of Prices in Iran's stock exchange market

Statistical Arbitrage Strategies are looking for profitable opportunities using statistical methods. In this paper, we use a new approach to devise a statistical arbitrage strategy in Iran’s stock exchange market. In this new approach, instead of model and forecast stock prices independently, we take them as a whole and extract their common movement patterns, which can represent the general market movements, with principal component analysis. After that, we model and forecast these patterns (factors) and through them, we forecast the stock returns. Ultimately, we construct portfolios from chosen stocks in each period. Empirical result of this paper show the profitability of these strategies. chosen strategy, with time window of 100 days and forecasting horizon of 1 day could made the average annual return of 115%, without considering the transaction costs.