We propose independence and conditional coverage tests aimed at evaluating the accuracy of Value-at-Risk (VaR) forecasts from the same model at different confidence levels. The proposed procedures are multilevel tests, i.e. joint tests of several quantiles corresponding to different confidence levels. In a comprehensive Monte Carlo exercise, we document the superiority of the proposed tests with respect to existing multilevel tests. In an empirical application, we illustrate the implementation of the tests using several VaR models and daily data for 15 MSCI world indices.

This is the accepted version of the paper.This version of the publication may differ from the final published version. This paper combines two closely related papers, one by Simona Bo¤eli and Giovanni Urga previously circulated under the same title and one by Vasiliki Skintzi previously circulated under the title "Dynamic Component Correlation Models and Macroeconomic Determinants". We are very grateful to the Editor, Eric Ghysels, for having sponsored and encoraged us to merge the two contributions to produce a comprehensive paper which fully explores a wide range of MIDAS models. We wish to thank participants in the 24th (EC) Permanent repository link2 Conference, The Econometrics Analysis of Mixed Frequency Data (13-14 December 2013, University of Cyprus, Nicosia, Cyprus) in particular Eric Ghysels and Rossen Valkanov; in the 6th MAF Conference (22-24 April 2014, Lloyd's Baia Hotel, Vietri sul Mare, Italy); and in the 7th Annual Society for Financial Econometrics-SoFIE-Conference (11-13 June 2014, Rotman School of Management and the Global Risk Institute in Financial Services, Toronto, Canada), in particular Andras Fulop, Eric Ghysels and Robin Lumsdaine, for very useful comments and suggestions. Special thanks to …ve referees, an Associate Editor, and to Jan Novotny for having provided us with very insightful and useful comments which greatly helped to improve the paper. The usual disclaimer applies. Special thanks to Morningstar, in particular to Richard Barden, for having made available the rich and unique data set used in this paper. Simona Boffelli acknowledges …nancial support from the Centre for Econometric Analysis, Cass Business School, City University London, UK. AbstractWe propose to adopt high-frequency DCC-MIDAS models to estimate high-and lowfrequency correlations in the 10-year government bond spreads for Belgium, France, Italy, the Netherlands and Spain relative to Germany, from 1-June-2007 to the 31-May-2012. The high-frequency component, re ‡ecting …nancial market conditions, is evaluated at 15-minute frequency, while the low-frequency component, …xed through a month, depends on country speci…c macroeconomic conditions. We …nd strong links between spreads volatility and worsening macroeconomic fundamentals; in presence of similar macroeconomic fundamentals relative spreads move together; the increasing correlation in spreads during the burst of the sovereign debt crisis cannot be entirely ascribed to macroeconomic factors but rather to changes in market liquidity.

This paper investigates the impact of macroannouncements, government bond auctions and rating actions on the 10-year government bond spreads for Belgium, France, Italy, the Netherlands, Spain with respect to Germany. Using a unique tick-by-tick dataset over 1/02/2009-05/31/2012, we identify the impact of the three drivers via jump and cojump detection procedures. Disentangling the pre-from the post-announcement effects, real economy and forward looking news releases from US and Euro area, country specific Spanish and German macroannouncements, and auctions hold in distressed countries such Italy and Spain have a statistically and economically significant effect.No role is played by rating actions.Keywords: Jumps, Cojumps,Government Bond Spreads, Macroannouncements, Government Bond Auctions, Rating Actions.J.E.L. Classification Numbers: C58, C12, H63, G24.Europe is under stress and integration among European countries seems more fragile than ever.Starting from the subprime crisis in 2007, markets are more aware of the differences between European countries, and this sentiment is reflected, amongst others, in increasing differentials of government bond yields. In 2008 and 2009, government bond spreads became sizable but it was in 5 2010 and 2011 that spreads substantially increase, reaching levels even higher than those experienced in the pre-Euro era. It was just after the famous Mario Draghi's "whatever it takes" in July 2012 that a more normal situation on government bond markets was restored. The European sovereign debt $ We wish to thank participants in the 12th Oxmetrics User Conference (Cass, 3-4 September 2012), in particular Siem Koopman, Sebastien Laurent, and Bill Lyons for useful suggestions. We thank Michael J. Flemming and Jan Novotny for having provided very useful comments on a previous version of the paper. The usual disclaimer applies. Special thanks to Morningstar, in particular to Richard Barden, for having made available the rich data set used in this paper. Simona Boffelli acknowledges financial support from the Centre for Econometric Analysis of Cass Business School.* Centre for Econometric Analysis, Faculty of Finance, Cass Business School, City University London, 106 Bunhill Row, London, EC1Y 8TZ, UK and Bergamo University, Italy. G.Urga@city.ac.uk, Tel: +44 (0)20 7040 8698, Fax: +44 (0)20 7040 8881. Preprint submitted to Journal of International Money and FinanceDecember 22, 2014 crisis involving, although at different extents, all the peripheral countries have questioned the much celebrated markets' self-regulatory power as well as the ability of policy makers and regulators to 10 adopt stability measures and stimulate economic growth. Thus, understanding which factors drive sovereign risk is particularly timely also for the macroeconomic consequences of the comovements associated to these factors. For instance, higher spreads deteriorate borrowing capabilities and market confidence which simultaneously impact on consumption and investment. The way to ameliorate the effects of th...

This is the accepted version of the paper.This version of the publication may differ from the final published version. This paper combines two closely related papers, one by Simona Bo¤eli and Giovanni Urga previously circulated under the same title and one by Vasiliki Skintzi previously circulated under the title "Dynamic Component Correlation Models and Macroeconomic Determinants". We are very grateful to the Editor, Eric Ghysels, for having sponsored and encoraged us to merge the two contributions to produce a comprehensive paper which fully explores a wide range of MIDAS models. We wish to thank participants in the 24th (EC) Permanent repository link2 Conference, The Econometrics Analysis of Mixed Frequency Data (13-14 December 2013, University of Cyprus, Nicosia, Cyprus) in particular Eric Ghysels and Rossen Valkanov; in the 6th MAF Conference (22-24 April 2014, Lloyd's Baia Hotel, Vietri sul Mare, Italy); and in the 7th Annual Society for Financial Econometrics-SoFIE-Conference (11-13 June 2014, Rotman School of Management and the Global Risk Institute in Financial Services, Toronto, Canada), in particular Andras Fulop, Eric Ghysels and Robin Lumsdaine, for very useful comments and suggestions. Special thanks to …ve referees, an Associate Editor, and to Jan Novotny for having provided us with very insightful and useful comments which greatly helped to improve the paper. The usual disclaimer applies. Special thanks to Morningstar, in particular to Richard Barden, for having made available the rich and unique data set used in this paper. Simona Boffelli acknowledges …nancial support from the Centre for Econometric Analysis, Cass Business School, City University London, UK. AbstractWe propose to adopt high-frequency DCC-MIDAS models to estimate high-and lowfrequency correlations in the 10-year government bond spreads for Belgium, France, Italy, the Netherlands and Spain relative to Germany, from 1-June-2007 to the 31-May-2012. The high-frequency component, re ‡ecting …nancial market conditions, is evaluated at 15-minute frequency, while the low-frequency component, …xed through a month, depends on country speci…c macroeconomic conditions. We …nd strong links between spreads volatility and worsening macroeconomic fundamentals; in presence of similar macroeconomic fundamentals relative spreads move together; the increasing correlation in spreads during the burst of the sovereign debt crisis cannot be entirely ascribed to macroeconomic factors but rather to changes in market liquidity.