Exploratory Data Analysis of Stock Returns

• Activity Number: 133
• Type: Contributed
• Date/Time: Monday, July 30, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Business and Economic Statistics Section
• Author(s): Kimihiro Noguchi*+ and Alexander Aue and Prabir Burman
• Companies: University of California at Davis and University of California at Davis and University of California at Davis
• Keywords: financial time series ; conditional mean ; conditional volatility ; forecasting

Abstract

It has become increasingly popular to use GARCH-type models, which capture several important stylized features of financial time series, for predicting conditional volatility of stock returns. Although such models work reasonably well and their theoretical properties have been studied in depth, it is possible to find new properties of stock returns data with careful re-examination. Therefore, we conduct an extensive exploratory analysis of several recent daily individual stock returns data and propose a robust semiparametric approach of modeling conditional mean and volatility. Further, we investigate leverage effect and nonstationarity in conditional volatility. Finally, prediction performances of the newly proposed conditional volatility model is compared to the GARCH(1,1) and APGARCH(1,1) models.