Time Series Analysis: Improvement of the Accuracy for Forecasting Expected Return

Time series data often use by financial analysis for making an investment decision. The paper chooses US currency strength and interest rate as variable to use in APT method with S&P500 (US market stock index). The paper predicts US currency strength and interest rate in the next six months by using moving average (lag 1, 2, and 3) time series data to analyze. For US currency strength, lag 1 was the best time to predict the future for non-seasonality and lag 2 was the best time to predict the future for seasonality. For the interest rate, lag 1 was the best time to predict the future at non-seasonality and seasonality. Then using the new prediction number in APT method. The result showed that the different between the expected return and the actual return were smaller. By using the time series data analysis and moving average, it helps to improve the accuracy of predicting the expected return.