Title: Financial Integration, Liquidity and the Depth of Systemic Crises
Abstract: We investigate how financial integration a¿ects banks liquidity holding and the depth of systemic crises. We model a multiple regions economy, where each region is characterized by idiosyncratic liquidity shocks. The liquidity shocks are not necessarily asymmetric, then liquidity coinsurance is not always possible. The first-best allocation is reached when the regions are completely integrated since in this case the perfect liquidity coinsurance is achieved. Under complete integration no crisis would occur. However, when financial integration is only partial the perfect liquidity coinsurance is no longer possible. Nevertheless, two (or more) regions will find it optimal to coinsure the liquidity shocks in order to reduce their liquidity uncertainty and to increase their expected utility. Accordingly, under partial integration, the regions could find it optimal to reduce the liquidity holding and to increase the level of long term investment. In this case, partial integration opens up the opportunity for the occurrence of extreme events. That is, the cost of liquidity can become unusually high and the optimal consumption can display both higher volatility than in autarky and negative skewness. When complete financial integration is not achievable, then extreme events can be the optimal outcome of partial integration.