Measurement Matters — Alternative Input Price Proxies for Bank Efficiency Analyses
Most bank efficiency studies that use stochastic frontier analysis (SFA) employ each bank’s own implicit input price when estimating efficient frontiers. But at the same time, most studies are based on cost and/or profit models that assume perfect input markets. Traditional input price proxies therefore contain at least substantial measurement error. We suggest here two alternative input market definitions to approximate exogenous input prices. We have access to Bundesbank data, which allows us to cover virtually all German universal banks between 1993 and 2003. The use of alternative input price proxies leads to mean cost efficiency that is significantly five percentage points lower compared to traditional input prices. Mean profit efficiency is hardly affected. Across models, small cooperative banks located in large western states perform best while large banks and those located in eastern states rank lowest.