Pierre Giot and Sébastien Laurent, "Value-at-Risk for Long and Short Trading Positions", Journal of Applied Econometrics, Vol. 18, No. 6, 2003, pp. 641-664. In the empirical applications, we consider daily return data for a collection of three stock indexes and three U.S stocks (Source: Yahoo! Finance): the U.K. FTSE stock index (FTSE, 1988-01-04 - 2000-12-21): ftse.data the U.S. NASDAQ stock index (NASDAQ,1984-10-11 - 2000-12-21): nasdaq.data the Japanese NIKKEI stock index (NIKKEI, 1984-1-4 - 2000-12-21): nikkei.data Alcoa common stock (AA, 1990-01-03 - 2002-05-03) : McDonald's common stock (MCD, 1990-01-03 - 2002-05-03) : dj.data Merck common stock (MRK, 1990-01-03 - 2002-05-03) : The corresponding files contain the date (except for the FTSE) and the daily returns in as percentages. The numbers in parentheses above are the start and end dates for our sample and the first symbol inside the parentheses is the short notation (or ticker for the U.S. stocks) for the series that are used in the tables and comments. All four data files, which are in DOS format, are zipped in the file gl-data.zip. Unix users should use "unzip -a". All computations concerning the univariate models were performed with "G@RCH", an Ox package with a friendly dialog-oriented interface designed for the estimation and forecast of various univariate ARCH-type models (see Laurent, S. and Peters, J.-P., 2002, G@RCH 2.0: An Ox Package for Estimating and Forecasting Various ARCH Models, Journal of Economic Surveys, 16, 447-485). G@RCH 2.2 (or a more recent version) is free for academics and can be downloaded from the web site http://www.egss.ulg.ac.be/garch/ as well as the code required to replicate the results. Computations for the multivariate application were performed with a preliminary version of "MG@RCH". Like its univariate brother, MG@RCH is an Ox package with a friendly dialog-oriented interface designed for the estimation and forecast of various multivariate ARCH-type models (developed by Laurent, S., Peters, J.-P., and Rombouts, J.). Please consult the web site of G@RCH for more information. As the packages follow the authors, it is very likely that the web site will move in the future. In this case, please contact the second author or consult the ox web site http://www.nuff.ox.ac.uk/Users/Doornik/. Pierre Giot and Sébastien Laurent pierre.giot@fundp.ac.be Sebastien.Laurent@ensae.fr