Purchase prices and abnormal returns in United States banking industry mergers and acquisitions: 1989--1995

This study examines the purchase prices and the abnormal returns in mergers and acquisitions in the U.S. banking industry during 1989-95. It consists of two parts of empirical work. The first part, the comparison of deposit premiums paid in private branch sales with those in federally assisted failed institution sales, suggests that in the government-assisted failed branch sales, the FDIC and RTC subsidized the buyers of the failed-bank and thrift branches during 1989-91. After 1991, the deposit premium differences between the two types of deals were negligible, indicating that there may not be any government subsidization in the last four years of our observation period. Moreover, the two-stage ordinary least squares models show that the target's value is different for different buyers, which provides evidence for that the branch sales may be independent private value auctions.; In addition, this paper shows that industry level variables are statistically significant in determining the purchase premium, which also holds true in the second part of the empirical work dealing with private whole bank and thrift mergers.; The second part of the empirical study examines the merger prices and the abnormal returns (ARs) of the private whole bank and thrift transactions. Not surprisingly, the current study finds that the merger prices (the ratio of tangible book value premium to core deposit) are significantly higher in the deals involving publicly traded buyers or sellers, which indicates that the capital market plays an important role in determining the merger price. The shareholders of the target firms gain large positive ARs after the transaction announcement, whereas the shareholders of the acquiring firms gain much less, although still positive returns from the merger. On average, the overall returns of the buyers and sellers are positive. Similar results are also confirmed by the seemingly unrelated regression (SUR) models.; The study of ARs features a model that enables an analysis of the effects of consecutive deals made by the same buyers. The average AR resulting from the first deal is much higher than those of the second deal, implying that if a firm jumps to another deal before it can digest the previous transaction, the market shows its skepticism in the form of lower improvements in the buyer's stock price. However, once convinced of the buyer's ability in handling mergers and acquisitions, the market responds positively to its future deals, which is shown in the higher ARs in the third deals conducted by the same buyer. Nevertheless, we cannot draw a general conclusion on how previous deals affect the more recent transactions, because there is a wide dispersion in the measured ARs associated with "termed" deals.