An occurrence based regime switching model to improve forecasting

Purpose – The purpose of this paper is to propose an occurrence-based model to improve the forecasting of regime switches so as to assist decision making. Design/methodology/approach – This paper proposes a novel model where occurrences of relationships are taken into account when forecasting. Taiwan Stock Exchange Capitalization Weighted Stock Index is taken as the forecasting target. Findings – Due to the consideration of occurrences of relationships in forecasting, the out of sample forecasting is improved. Practical implications – The proposed model can be applied to forecast other time series for regime switches. In addition, it can be integrated with other time series models to improve forecasting performance. Originality/value – The empirical results show that the proposed model can improve the forecasting performance.