Long memory dynamics for multivariate dependence under heavy tails

• Published in: Journal of empirical finance Vol. 29; pp. 187 - 206
• Main Authors: Janus, Paweł, Koopman, Siem Jan, Lucas, André
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.12.2014
• Subjects: Correlation; Economic forecasts; Fractional integration; Memory; Multivariate analysis; Multivariate volatility; Student's t copula; Time series; Time-varying dependence; Volatility; Fractional integration; Correlation; Time-varying dependence; C22; C10; C32; Multivariate volatility; C51; Student's t copula
• ISSN: 0927-5398; 1879-1727
• DOI: 10.1016/j.jempfin.2014.09.007

We develop a new simultaneous time series model for volatility and dependence in daily financial return series that are subject to long memory (fractionally integrated) dynamics and heavy-tailed densities. Our new multivariate model accounts for typical empirical features in financial time series while being robust to outliers or jumps in the data. In our empirical study for daily return series of four Dow Jones equities, we find that the degree of memory in the volatilities is similar, while the degree of memory in correlations between the series varies significantly. The forecasts from our daily model are compared with high-frequency realized volatility and dependence measures. The overall performance of the new model is better than that of several well-known competing benchmark models. •A new multivariate dynamic model for the simultaneous analysis of volatility and dependence•We have long range dynamic properties for volatility and dependence in financial returns.•We have different tail shapes for volatility and dependence via dynamic copulas.•New model outperforms alternatives in fit when the copula dynamics have long memory.•Using real data sets, we find different long memory properties for correlations between equities.