Modeling Multiple Irregularly Spaced Financial Time Series

Abstract

In this paper we propose univariate volatility models for irregularly spaced financial time series by modifying the regularly spaced stochastic volatility models. We also extend this approach to propose multivariate stochastic volatility (MSV) models for multiple irregularly spaced time series by modifying the MSV model that was used with daily data. We use these proposed models for modeling intraday logarithmic returns from health sector stocks data obtained from Trade and Quotes (TAQ) database at Wharton Research Data Services (WRDS).