Due to regulatory changes, share repurchases have become increasingly common around the world in the last 15 years. As such, in our paper, Buybacks Around the World, which was recently made publicly available on SSRN, we first examine whether the findings based on U.S. data hold up in an international setting, and whether examining non-U.S. data can change the way we think about buybacks. Second, we examine whether the original concerns about managers using buybacks to prop up the share price were somewhat warranted in countries outside the U.S.

Past research covering buybacks prior to 2002 shows that, in the U.S., open market share repurchase authorization announcements are accompanied by positive announcement returns of about three percent and long-run abnormal returns in the order of 30 percent over three to four years. Our study is based on a global sample of 17,487 buyback authorization announcements from 32 countries between 1998 and 2008. We find that buyback announcements in the 7,394 non-U.S. buybacks generate a significant positive average abnormal return of 1.3%, as well as average long-term excess returns of about 25% over three to four years.

These results are consistent with a variety of non-mutually exclusive explanations, which can be grouped under two broad hypotheses: undervaluation and agency cost. The undervaluation hypothesis posits that firms buy back their stocks when they are temporarily undervalued. The agency cost hypothesis argues that buybacks, by returning cash to the shareholders, mitigate the agency costs of free cash flow. The findings, however, are inconsistent with the hypothesis that managers systematically use buybacks to support the share price in the short run. We use the regulatory differences that govern share repurchase authorizations as well as differences in corporate governance quality across countries to test these hypotheses.

Our global approach provides a unique laboratory to test for the relevance of the agency cost vis-à-vis the undervaluation hypothesis, by looking at variation in corporate governance quality at the country and firm level. If firms buy back shares to increase shareholder value by reducing agency costs of free cash flow, we expect a negative relation between corporate governance quality and excess returns. Indeed, firms with low corporate governance quality should benefit more from the reduction in agency costs of free cash flow. The undervaluation hypothesis would argue the opposite: firms buy back stock when the shares are undervalued, but they are more likely to do so if they care about shareholder value, i.e. when their corporate governance quality is high. Thus the undervaluation hypothesis predicts a positive relation: the better corporate governance, the higher excess returns.

We find that announcement returns are positively associated with governance quality, at least at the country level. However, at the company level this positive relation is only significant in the non-U.S. sample. One interpretation of this is that in a high quality governance country like the U.S., buybacks are typically driven by good (shareholder value maximizing) reasons. In non-U.S. countries this is not always the case, and therefore investors pay attention to company-specific governance quality when assessing whether the repurchase is good for shareholders. This suggests that some of the reasons for announcing a buyback in low governance quality firms may actually be driving by non-value maximizing reasons—something earlier regulation was trying to prevent. Interestingly, some countries outside the US have decided to require a shareholder approval of a buyback to protect shareholders against buybacks driven by non-value maximizing reasons. However, requiring shareholders to approve the buyback may diminish the information content of buyback announcements.

We find significantly higher average short-term announcement returns in countries where board approval is sufficient, a result consistent with the undervaluation hypothesis: board approval allows a more timely response to undervaluation. Again, the data is inconsistent with managers systematically announcing buybacks for non-maximizing reasons, which are better monitored by requiring shareholder approval of buybacks.

Similar to the U.S., open market buybacks globally generate significant, positive long-term excess returns, larger and more persistent than in the U.S. Indeed, we present some evidence that in recent years long-term excess returns have declined in the U.S, while no such decline is observed in the foreign sample. As in the U.S., long-term excess returns in foreign countries are significantly higher for beaten up small value stocks. However, unlike in foreign countries, U.S. long-term excess returns can be partially explained by takeover activity: undervalued firms get subsequently taken over and/or their stock price reflects a compensation for takeover risk. As in the case of short-term returns, long–term excess returns are positively related to country-level governance quality and board approval, but there is only a positive relation between company-specific governance and long term returns in foreign countries, not in the U.S.

In sum, we conclude that there are a lot of similarities between global buybacks, i.e. significant positive excess returns in the short and long run, especially for small beaten up value stocks. But there are some major differences: First, company-specific corporate governance quality matters abroad, but not in the U.S. Second, the lack of an active takeover market abroad means that market inefficiencies or agency costs are less likely to be corrected through the market for corporate control. Third, as U.S. markets have become more efficient in the recent decade it has become more difficult for U.S. firms than for foreign firms to use buybacks to take advantage of undervaluation. Fourth, we conclude that, on the net, non-value increasing reasons for a buyback are not dominant outside the US despite some evidence that announcement returns of low governance quality firms outside the U.S. are lower compared to high governance quality firms.

The full paper is available for download here.