کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
4960565 | 1446501 | 2017 | 10 صفحه PDF | دانلود رایگان |
While stock prices and economic activity are interrelated in a nation, they “are not coincident” with each other. Stock prices are a leading economic indicator of the United States of America's (U.S.A.'s) economy. An economic variable that influences stock market prices is interest rates through an inverse relationship. The changes in stock prices (or stock returns) are generally caused by the demand for stocks. This paper reports on a study that investigates the underlying spectral and time-frequency characteristics of daily Standard and Poor's (S&P) 500, Dow Jones Industrial Average (DJIA), and National Association of Securities Dealers Automated Quotations (NASDAQ) composite stock returns, and changes in interest rate (namely, inverted 3-month Treasury bill). The study thereafter compared these findings with those obtained in a previous study by Joseph et al, which focused on monthly stock returns and interest rate data. Subsequent to studying stock returns and changes in interest rate that showed relatively similar spectral and frequency-time characteristics, this study investigated the forecastability of stock returns (in S&P 500, DJIA, and NASDAQ composite) by inverted interest rate (in 3-month Treasury bills) over prediction horizons of five and 30 days with the forecasting period covering the last 13 years. The measures of forecast accuracy used were root mean square error and correlation. The forecasts were favorable in all cases even with simpler neural network models.
Journal: Procedia Computer Science - Volume 114, 2017, Pages 481-490