The paper looks at the importance of the true business model in shaping the risk profile of financial institutions. We adopt a novel indirect clustering approach to enrich the classic bank business model classification on a global data set including about 11,000 banks, both listed and non-listed representing more than 180 countries over the period 2005-2014. A comprehensive list of global distress events, which combines bankruptcies, liquidations, defaults, distressed mergers, and public bailouts, is regressed against financial statement ratios (i.e. proxies for CAMELS) and controlling for macro and sectoral effects using a rare-event logit model. Our findings suggest that individual characteristics along with macro and sectoral factors contribute differently, sometimes with opposite sign, to the likelihood of distress and to the volatility of business models with the exception of liquidity whose contribution appears exogenous to business model choice. By capturing the switching behaviour across groups, we find that business model volatility exacerbates vulnerability and distress, especially if moving from wholesale-oriented to deposit oriented groups.