美國銀行業實際買回庫藏股計畫之研究：動機與管制因素之影響

Abstract

自1990年代初期開始，買回庫藏股在美國銀行業的股利政策中，重要性日 增，成長速度也很驚人。其在90年代初期處於還處在微不足道的水準，到1997年已經幾乎與股利的發放並駕齊驅。本篇研究的目的在於探討討美國銀行業實際買回庫藏股計畫如何受動機與管制因素之影響。由於銀行業的特殊性，過去關於庫藏股研究文獻多只針對非金融業。另外，由於銀行業處於政府法令管制的環境下，法令的因素也影響了銀行業買回庫藏股的政策。因此，本篇研究全面檢視美國銀行業實際買回庫藏股的動機。 根據我們的研究結果發現：當銀行業自由現金流量越多，成長機會越少，資本適足率越高，以及被併購的威脅越大時，其實際買回庫藏股的可能性越高。這些都與我們的預期相符合。另外，我們也發現，當銀行業的資產越多，風險越小，以及現金股利的支付率越高時，其實際買回庫藏股的可能性越高。這些與我們的預期結果相反。大致上來說，實證的結果大多支持了我們的假說。

Beginning from the early 1990s, there is a trend of dramatic growth in shareholder payout rates among bank holding companies. The rise in stock repurchases is particularly striking, with aggregate repurchases soaring from a negligible level in the beginning of 1990s to an amount almost equal to dividend payments in 1997. This thesis is to investigate how the motivation factors and the regulatory environment affect the intensity of bank holding companies’ stock repurchase activities in the United States. Although stock repurchases by non- financial firms have been extensively examined in the prior literature, there has not been an equivalent amount of research with regard to stock repurchases by bank industry. In addition, because of the industry uniqueness, banks are under a different regulatory environment from other industries. Repurchases of stock by banks are inevitably linked to the regulatory environment banks operate in. We then construct our own ten hypotheses and use Tobit model to test our hypotheses. We find that banks’ magnitude of shares buyback are obviously affected by the larger free cash flow they possess; the less growth opportunity they own; the greater first tier capital adequacy ratio and total capital adequacy ratio they have, and the greater potential takeover threats they face. These all meet our expectations. We also find that banks’ magnitude of shares buyback are strongly influenced by the larger assets they have; the less risk they face, and the more cash dividends to net income they pay. These are opposite to our prediction but have some reasonable explanations. Overall, we get satisfying evidences to support our hypotheses.