“…To address the potential self-selection bias arising from being a bailed-out bank, we employ a two-stage Heckman selectivity model (Heckman, 1979). Here, we first estimate a probit model, regressing the Recapitalized dummy on all control variables from our main specification, as well as on certain determinants discussed in the literature on bank recapitalizations (Beccalli & Frantz, 2016;Berger & Bouwman, 2013;Gerhardt & Vennet, 2017;Mariathasan & Merrouche, 2012;Papanikolaou, 2018). From this probit estimation, we obtain the inverse Mills ratio, which is used as one of the regressors in the second-stage equation to produce consistent estimates of underwriters' market shares.…”