Markets, reserves and lenders of last resort as sources of  bank liquidity

We study the long standing issue of whether markets can supply banks with sufficient liquidity or whether markets should be complemented with a lender of last resort (LOLR).For this purpose, we develop an extended version of the recent model of Holmström and Tirole (1998) on the supply of liquidity to firms. H&Ts original model analyses liquidity supply to firms that are facing solvency shocks.We apply their framework to banking and extend the framework to admit the analysis of problems associated with transitory liquidity outflows, even absent any change in a bank's value.Our premise is that the scope for moral hazard may increase in connection with liquidity outflows.Moral hazard, which we interpret as the possibility of laxity in banks' monitoring of firms, may increase with liquidity outflows because banks need to increase their monitoring efforts in order to safeguard their own interests. The model illustrates many key aspects of the classical LOLR debate.The model shows how moral hazard limits of banks' ability to borrow from markets to cover liquidity outflows.It also predicts banks' demand for liquid reserves and the economies associated with centralization of reserves in a liquidity pool when the holding of liquid reserves entails opportunity costs.Finally, the model enables discussion of viable lending policies for the LOLR and contrast these with the 'Bagehotian principles', which are still widely used as benchmark criteria in evaluating LOLR operations. Key words: liquidity, lender of last resort, banking, central banking