Bernd Schwaab, Siem Jan Koopman, and Andre Lucas, "Global Credit Risk: World, Country, and Industry Factors", Journal of Applied Econometrics, Vol. 32, No. 2, 2017, pp. 296-317. The empirical part of our paper uses three separate sets of input data. Each set is proprietary or subject to licensing agreements. First, we obtained one-year-ahead expected default frequencies (EDF) data from Moody's Analytics. We accessed Moody's Analytics via their Credit Edge platform (https://www.creditedge.com) and obtained EDFs for listed financial and non-financial firms in the respective countries (for the U.S., U.K., and Japan) or region (the euro area). We consider country/region-level aggregate EDFs in the empirical analysis. These are the median EDF for financial firms, and EDF weighted by firm-level total assets (also from Moody's Analytics) for non-financial firms. The EDF data are available from Moody's Analytics for a monthly fee. Second, we used default and firm-count data from Moody's Default and Recovery Database (DRD). We accessed Moody's via their online platform (https://www.moodys.com/Pages/Default-and-Recovery-Analytics.aspx). We downloaded and combined three text files: MAST_DFLT.txt, MAST_ISSR_ISSR.txt, and SENR_RATG_STANDARD.txt. These data were matched based on Moody's firm id. The standard cleaning procedures are described in the main paper. Access to Moody's DRD is subject to a fee. Finally, we used macroeconomic and financial market data from Thomson Reuters/Datastream. The macro-financial data were obtained at the country level. Indexes on real estate property prices were taken from an ECB internal source, the Statistical Data Warehouse, SDW.