Credit Rationing in Markets with Imperfect Information

Stiglitz and Weiss build an equilibrium model of a competitive loan market in which banks cannot directly observe a borrower’s project risk, and show that this imperfect information alone is enough to generate credit rationing as a stable market-clearing outcome, not a sign of disequilibrium or price stickiness. Because raising the interest rate changes who applies for loans (the adverse-selection effect) and how the firms that do borrow behave (the incentive effect), the bank’s expected return on a loan is not necessarily monotonic in the rate charged; past a “bank-optimal” rate, a further increase actually lowers the bank’s expected return, so a profit-maximizing bank facing excess demand for loans may prefer to deny some observationally identical applicants rather than raise the rate to clear the market. The same logic can make the bank’s return a decreasing function of collateral demanded at some point, so collateral requirements cannot generally substitute for price as a market-clearing device either. The authors show the equilibrium need not allocate credit to the borrowers with the highest expected returns, and that distinguishable groups can be entirely “redlined” out of the market even when some of their members would be profitable to fund at the going rate. They conclude that the textbook presumption that prices clear markets is not a general law but a result specific to settings where price does not also affect the quality of what is being transacted.

Andrew Weiss | Macro Paper Warehouse

Credit Rationing in Markets with Imperfect Information