In this paper, Professor Riordan has provided a model of lending that he relates to the desirability of a ‘single banking licence’ for Europe. The comments that follow stress the relevance of this model for drawing inferences about the effects of more competition in banking markets.

The bank objective function, maximizing the expected loan payoff, misses at least three trade-offs that are important in the single banking market context. First, the bank's price-setting mechanism must depend on the number of banks in the market in ways other than the effect on the conditional probability of borrower quality. In his conclusion, Riordan clearly recognizes this point. It is repeated here for emphasis and as a suggestion that explicit treatment of the direct price effects of more competition would be a valuable extension of the model.

A second advantage of opening European bank markets is the added scope for loan diversification that this could afford banks. Might not the value of an additional Spanish loan to a French bank be greater, all else equal, than the value of that loan to a Spanish bank? To capture this effect, the objective function must include an expected return/risk trade-off.

A third consideration is comparative advantage in information processing, which, presumably, is part of what Riordan is getting at in his section on the internal organization of banks. If some banks have superior information collection and processing skills, but limited capacity for exercising those skills (or diminished ability as a function of portfolio size), competition can help. The banks with best abilities will take on the best projects and increase the liquidity of good projects.