Financial Market Forecasting Methods under Structural Break

Financial market forecasting particularly stock market forecasting is a considerable debate that confront to forecast failure and model break down when structural breaks in trends occur. This paper discusses the modeling to predict stock return under structural breaks and investigate new approaches of forecasting in this condition. This study proposes a taxonomy for research area in forecasting under structural breaks to suggest further studies. We use literature survey as methodology of the research and categorizes the methods, models, and results of the recent researches in stock market forecasting. Consequently, it provides three categories of strategies to forecast stock return under structural breaks. First strategy, called economically motivated model restrictions, uses financial theories as signs to adjust the parameters of models in out-sample periods. Second strategy, known as regime shift, uses a Markov chain transition matrix to model structural breaks in time series. Third strategy applies mix of quantitative models and qualitative surveys to predict future of financial markets. The proposed strategies are applicable in Tehran stock exchange under uncertainty conditions.