Abstract/Summary

This work makes three main contributions to the financial econometrics literature. In Chapter 3, we study the intraday volatility of European government bonds under the framework of the multiplicative component GARCH model (Engle and Sokalska, 2012). We suggest a flexible and effective procedure for jointly filtering mid-quote prices and estimating volatility models and show that intraday data contain relevant information for daily volatility forecasts. In Chapter 4, we show that a bond portfolio can reduce its intraday variance risk by including bonds from Italy and Spain. Furthermore, we demonstrate that the bivariate (scalar) DCC model is capable of computing an accurate VaR, providing correct conditional and unconditional coverage at lower than 1% (inclusive) confidence level and inducing lower losses. In Chapter 5, we demonstrate that liquidity measures, such as the bid-ask spread and quantity available for trading at the best quotes, improve across maturities and countries after EuroMTS has allowed every market participant to post limit orders and not just designated market makers. In particular, we show that the relative bid-ask spread for trading 10 million bonds decreases with the rule change. The proportion of time when the relative bid-ask spread stays low also increases. The results suggest that greater competition amongst liquidity providers improves liquidity.