期刊名称：Bank of Finland Research Discussion Papers (früher: Bank of Finland Discussion Papers)
印刷版ISSN：0785-3572
电子版ISSN：1456-6184
出版年度：2015
卷号：2015
出版社：Suomen Pankki = Bank of Finland
摘要：This paper presents first steps toward robust early-warning models. We conduct a horse race of conven- tional statistical methods and more recent machine learning methods. As early-warning mo dels based upon one approach are oftentimes built in isolation of other methods, the exercise is of high relevance for assessing the relative p erformance of a wide variety of metho ds. Further, we test various ensemble approaches to aggregating the information pro ducts of the built early-warning models, providing a more robust basis for measuring country-level vulnerabilities. Finally, we provide approaches to esti- mating mo del uncertainty in early-warning exercises, particularly model performance uncertainty and model output uncertainty. The approaches put forward in this paper are shown with Europe as a playground
关键词：financial stability; early-warning models; horse race; ensembles; model uncertainty
