This was my second PhD project, and it took far less time than my first project, as I had already learned much of the coding and theory, so it was a matter of reading papers on the subject, while trying to find something no one else had done before.
This paper builds upon my first project on realised volatility, and uses the RV GARCH as marginals as inputs into a Realised DCC Skewed Student-t Copula. The Realised DCC component using intra-period data to to construct a Covariance Matrix which will hopefully add a valuable information source into the Copula. Student-t and Skew Student-t (Fernandez & Steel – 1998) Marginals are used together with Normal, Student-t and Skew Student-t (Bauwens & Laurent – 2005) Copulas for both RV and traditional GARCH models. 5 assets (indexes, equities, currencies, commodities & bonds) are used, and weekly returns are calculated, with daily returns forming the ‘high frequency’ component.
The reason for this is twofold. Firstly, it becomes near impossible to build covariance models with high frequency data in an international market, as markets are only open at the same time for a small overlap. To overcome this, I used daily data, and generated 5-day returns (the week), while using the daily returns as the high frequency component to calculate realised volatility. The second reason is because high frequency intra-daily data is not available for many assets, so to simulate a realistic portfolio with a variety of asset classes, this approach was the only way.
This paper has been further extended by using intra-day data for USA only assets, which overcome the problems stated above.
This paper was presented at the Second International Conference on Mathematics and Statistics (AUS-ICMS ’15) at the American University of Sharjah, United Arab Emirates in April 2015.
If you have questions about any aspect of this project (code, theory etc), please shoot me an email and I will do my best to get back to you as soon as possible.