A Novel Approach for Decomposition of Financial Time Series

Accurate forecasting of future events like decline in stock price of a company constitutes a captivating challenge in the field of time series analysis for research. For a long period, many researchers emphasis on the behavior of the stock market in long run but in this paper we are predicting the stock price (univariate) or NAV of mutual fund(multivariate) on daily basis or monthly basis because people are more interesting what's the closing price of next day or month. In this research, we are proposing a mathematical model for decomposition of time series and forecasting. Here we also compare with the classical decomposition method. Our main focus is to improve the accuracy of existing decomposition techniques to lift up the quality of management decisions