Finance: Research, Policy and Anecdotes

A lot has been written about the negative effects of bank failures on the real economy and about the adverse effects of bank bail-outs on banks’ incentives to take aggressive risks.  But what about bank bail-in, i.e., a resolution of a failing bank that involves maintaining the “good” part of the bank, while shareholders and junior bondholders have to bear the losses?   The resolution of the Portuguese Banco Espirito Santo in 2014 offers a good example to study, as it was unexpected, not related to loan losses and as its resolution did not affect negatively the rest of the financial system.  Andre Silva, Samuel Da-Rocha-Lopes and I just finalized the first version of a working paper gauging the effect of this bank resolution, also summarized in this Vox column.

Using credit and corporate registry data, we find that (i) the supply of credit from banks more exposed to the bail-in (i.e., BES itself and banks that had to come up with additional funding to the resolution fund) declined significantly after the shock when compared to less-exposed banks, (ii) firms exposed to the shock expanded borrowing from other, less-exposed banks, both on the intensive margin (existing banking relationships) and on the extensive margin (establishing new relationships), (iii) despite not suffering a credit supply shock, SMEs exposed to the bail-in experienced a relative reduction in investment and employment, (iv) SMEs exposed to the bail-in had higher cash holdings after the shock, pointing to precautionary cash hoarding as explanation for the discrepancy in credit supply and real sector effects, and (v) large firms seemed to have achieved a similar effect by increasing outstanding trade credit from their suppliers, thus not having to cut investment and employment.

In summary, credit supply effects were contained by the bank bail-in and good-bank-bad-bank split, but there were real sector effects.  While immediate panic and contagion effects were thus avoided, resolution of a failing bank, however well done, may not completely eliminate negative effects on the real economy. 

28. May, 2017