The Impact of State Deposit Caps on Bank Merger Premiums

Abstract

Nineteen states set a ceiling on the percent of total deposits that any one banking organization may hold in that state. This ceiling is known as a deposit cap. Once a bank has reached a state's ceiling, either through growth or merger, it may not acquire additional banks in that state. Thus, the presence of a deposit cap may eliminate potential bidders for a target. We argue that fewer bidders for a target bank may reduce the ratio of purchase price to book value, or what has been termed the merger premium. In this short paper, we empirically address the impact of state deposit caps on bank premiums. We find, as expected, that the presence of deposit caps significantly reduces bank premiums. An analysis of the impact of deposit caps on bank premiums is particularly important for the following reason: On September 29, 1994, President Clinton signed into law the RiegleNeal Interstate Banking and Branching Efficiency Act [6] effectively striking down the McFadden Act of 1927 and its subsequent 1933 amendments. The new interstate banking law, which became effective on September 29, 1995, establishes a uniform state deposit cap of 30 percent for those states with no existing deposit cap and a national deposit cap of 10 percent. Although we cannot yet measure the impact of the new federally imposed deposit caps on premiums because limited data has accumulated since their imposition, we can measure the impact of existing state deposit caps. Our results may be useful as an indicator of the expected impact of the new federally imposed caps. In addition, our results may be immediately useful to various state legislatures as they seek to adjust their existing state deposit caps. Though previous studies employ financial, market structure, and regulatory data to explain the variation in bank premiums, no existing research has included deposit caps as determinants of bank premiums. We address the impact of deposit caps by employing them as additional explanatory variables in a model explaining premiums for the period of 1989 through 1994. Thus, while our results support many of the conclusions found in the existing lit-