(ProQuest: ... denotes formulae omitted.)IntroductionThe global financial system experienced a serious crisis in July 2007. This crisis, initially affecting the American housing market, gradually spread to the entire global financial system. The crisis not only caused the default of some of the world's largest banking institutions, but was also at the root of a worldwide financial crisis comparable to that of the Great Depression of 1929.The 2007-2008 banking crises have been the subject of numerous controversies as to their similarities and differences to past banking crises. Some claim that the recent crises are different in every aspect. They are chiefly due to a global savings glut and the absence of shadow banking system regulation (Adrian and Shin, 2009). Others maintain that the recent episodes of banking distress are not so different from the previous ones and that the latter show remarkable similarities to the former. According to Claessens and al. (2010b), these similarities are: First, the price of real and financial assets rising considerably in a number of countries before the crisis, notably in the United States and Europe. These prices reached 60% before the start of the crisis, which strongly recalls the price spike observed during major financial crises of the '90s, notably the Japanese crisis of 1997 (Caballero, 2010). A second similarity is the occurrence, in a number of major economies, of credit booms before the crisis, estimated at over 150% of GDP (Claessens and al., 2010b). Third, international financial integration facilitated large capital inflows, which contributed to the acceleration of GDP growth and massive credit growth, which in turn led to a strong fluctuation of global demand and a strong deterioration of current bank balances during the period preceding the crisis (Cardarelli and al., 2010). Fourth, the inadequacy of the regulation and prudential supervision framework (Bair, 2009).In light of these findings, the goal of this study is twofold. First, to determine if the banking crises of the 2000s have shared causes with the crises of the 90s, and second, to determine if aggregated accounting indicators are robust banking crisis indicators.Thus, in this study we propose first to identify banking crisis indicators by means of a limited dependent logit approach for a cross-sectional view of advanced economies during the period preceding the 2007-2008 banking crises, namely 1990-2006, and the period 2007-2012. Second, we propose to test the robustness of the results derived from the multivariate logit approach, by means of Bayesian statistics (BMA). Indeed, according to Cuaresma and Slacik (2009) and Babecký and al. (2012), the BMA approach has the advantage of reviewing different model combinations and of weighting them according to their adjustments in the model.This paper will be organized as follows: The first section being an introduction, in the second section we briefly present a review of the literature on banking crisis indicators. In the third section, we present our methodology, namely: our country sample and our main data sources, our endogenous and exogenous variables, and our two econometric approaches. In the fourth section, we present a brief descriptive analysis of our data. We describe and discuss in the fifth section our empirical results. The last section is the conclusion.Financial Crisis Indicators: Review of the LiteratureBanking crises are not limited to the 21st century. Indeed, during the past four decades, the global economy was marked by an increase in banking crises. According to Reinhart and Rogoff (2013), banking crises represent a threat to equal opportunity amongst emergent and advanced economies: most countries have had at least one banking crisis during the period of 1945-2008.The reoccurrence of these crises, their magnitudes and their surprising and unpredictable character, and the financial costs associated with these episodes explain the research communities' interest in these events. …