Three essays on banking crises

This dissertation analyzes the phenomena of systemic banking crises. The first two chapters deal with the international dimension, exploring international factors that may increase the probability of crises, and the third chapter studies the medium-term evolution of productivity growth after crises.;Chapter 1 aims at answering the question of whether or not bonanzas (surges) in net capital inflows increase the likelihood of crises. The chapter fits a binary-outcome model for the probability of crises in the period 1973--2008. The methodology focuses on separating the effect of windfalls of capital from that of lending booms. The findings indicate that bonanzas are associated with an increased likelihood of crises, and that this effect is present even in the absence of a lending boom. A baseline bonanza raises the likelihood of a crisis from an unconditional probability of 4.4% to 12%. Decomposing flows into FDI, portfolio equity, and debt indicates that bonanzas in all flows increase the probability of crises when taking place jointly with a lending boom. However, surges in portfolio-equity flows seem to be the only type of flow with an independent effect.;Chapter 2 explores the involvement of banks in the global economy, using network statistics to proxy for financial integration of banks. The analysis is based on a count-data model for the number of banking crises in the period 1973--2007. The results indicate that the level of integration of the average bank, as measured by its international borrowing, is a robust determinant of the incidence of crises. However, the more important the average bank is as a global intermediary (proxied by betweenness), the lower the incidence of crises.;Chapter 3 aims at shedding light on the medium-term effects of crises on productivity growth, bringing the 'creative destruction' hypothesis to the data. The chapter studies the differences in mean productivity growth before and after a crisis using a differences-in-differences methodology. The results indicate that during the seven years following a crisis, productivity grows faster than during the seven years before the crisis. Furthermore, sectors that rely relatively more on external finance exhibit, on average, comparatively faster post-crisis productivity growth.