CAN MEGAMERGERS IMPROVE BANK EFFICIENCY

This paper uses a fitted multiproduct translog cost function to simulate mergers between pairs of US commercial banks with assets exceeding $1 billion and studies the predicted impact of such mergers on total costs. Thick frontier analysis reveals that variations in X-efficiency dominate scale, product mix, and branching efficiencies. Ex ante financial data imply that costs could potentially be reduced in nearly half of the possible mergers except those between banks each larger than $10 billion. Most cost-efficient mergers are across state lines, apparently a result of variations in product mix.