Regulation, Taxes, and Share Repurchases in the United Kingdom*
We examine share repurchase activity in the United Kingdom over a period when the tax and regulatory environment changed drastically. We find that the form and intensity of repurchase activity in the United Kingdom is influenced by the tax consequences for pension funds. We also find that firms announcing share repurchases earn smaller excess returns, both in the short run and the long run, than those earned by firms in the United States. This is because of regulatory provisions in the United Kingdom that make it less likely that firms can use superior information to buy back shares when their shares are undervalued.
I. Introduction
Share buybacks are extremely popular in the United States: from 1985 to 1999, U.S. corporations announced intentions to repurchase roughly $750 billion worth of stock (Grullon and Ikenberry 2000). Despite its growing popularity, the causes and consequences of share repurchase activity are still controversial. Moreover, the idiosyncrasies of the U.S. tax and regulatory system raise questions as to whether the results documented for the United States are unique to this country. Finally, the fact that repurchasing firms in the United States earn positive abnormal returns in the long run after the repurchase announcement raises important questions about market efficiency.
The purpose of this article is to examine the motivations and consequences of share buybacks in the United Kingdom. The United Kingdom is of particular interest for two reasons. First, companies and regulators face a different tax and regulatory environment than they do in the United States. In particular, companies are not allowed to buy back shares in periods when managers are most likely to have superior information about future earnings. Second, the tax treatment of dividends and share repurchases changed on several occasions in our sample period. This specific tax and regulatory environment enables us to test to what extent repurchase activity is driven by the tax code and the ability of managers to take advantage of an undervalued stock price.
We choose the United Kingdom rather than other European countries by default. Historically, share repurchases have not been very important in continental Europe. From January 1980 to June 1998, Securities Data Corporation (SDC) reports only 489 share repurchase announcements made by European firms. Approximately 60% of these announcements were made by companies listed in the United Kingdom (see fig. 1). The reasons for the lack of European share repurchase activity are fourfold. First, in some countries (e.g., France and Germany) share repurchases were illegal until recently. Second, in other countries with high taxes on dividends and low capital gains taxes (such as the Netherlands), specific tax provisions exist to discourage share buybacks. Third, in many European countries, companies do not have to disclose share buyback authorizations. Hence, (completed) share buybacks are only mentioned in the annual report. And, finally, as most continental European companies traditionally have emphasized maximizing “stakeholder value” rather than “shareholder value,” share buybacks do not fit the European corporate culture. Indeed, apart from shareholders, other stakeholders (such as managers, banks, and labor unions) are not interested in share buybacks. Managers prefer to maximize firm size, as the lack of incentive schemes tied to stock prices (such as stock options) makes firm size the most important determinant of executive compensation. Banks prefer the company to preserve its capital, and union leaders want the company to create more union members through additional capital expenditure. Hence, it is not surprising that, historically, European buyback activity mainly took place in the only European country with an Anglo‐Saxon corporate governance model. The gradual acceptance of this model may explain why share buybacks are becoming more popular in recent years in Europe.
Fig. 1.— Share repurchase announcements in Europe. This graph shows the number of common stock share repurchases for cash, announced in Europe between January 1985 and June 1998.
We analyze a sample of 264 substantial share repurchase announcements in the United Kingdom, reported by SDC between January 1985 and June 1998. Our sample includes all open‐market share repurchases, private repurchases, and tender offers reported by SDC in the sample period. The repurchases analyzed in this article are substantial in the sense that, on average, companies announce that they may buy back 9.8% of their own stock in the open‐market sample, while the corresponding percentages for tender offers and private purchases are 23.4% and 17.2%, respectively.
We can distinguish four major tax regimes during our sample period. Prior to September 1994, all open‐market share repurchases were unattractive for pension funds (relative to dividend payments and relative to repurchase tender offers). In September 1994, investment banks invented the “agency buyback,” an innovative open‐market repurchase mechanism that increased the tax attractiveness of stock repurchases in the open market. In October 1996, the tax authorities abolished the loophole, making all open‐market share repurchase programs tax inefficient again. Finally, in July 1997, the government changed the tax laws in such a way that pension funds, as in the United States, became indifferent between dividends and share buybacks.
We find that the tax changes have a major impact on the importance and method of share repurchase. The introduction of tax‐favored agency buyback programs in September 1994 caused a substantial increase in the number of open‐market repurchase programs. Simultaneously, tender offers essentially disappeared, with only one tender offer being announced between September 1994 and October 1996. When the tax credit given to tax‐exempt pension funds in agency buybacks was abolished in October 1996, dividends became more profitable as a method of returning cash to shareholders. We find that the number of open‐market repurchase completions fell significantly. Finally, when all tax credits, including those paid to dividends, were abolished in July 1997, the relative attractiveness of share repurchases rose again, with an increase both in the number of repurchases announced and repurchases completed.
The finding that corporate payout policy is sensitive to tax law changes in the United Kingdom seems to be inconsistent with the fact that, after the 1986 U.S. tax reform, buyback activity did not decrease (Bagwell and Shoven 1989). As the 1986 tax reform increased the capital gains tax, one would expect the volume of share repurchases to decrease. However, both results are consistent with the hypothesis that it is the tax treatment of important investors such as pension funds, and not individual investors, that determines company payout policy. We should note that, both in the United Kingdom and the United States, individual investors prefer share buybacks to dividends for tax purposes. However, unlike those in the United Kingdom, U.S. pension funds always have been tax‐indifferent.
We find that share repurchases intentions generate statistically significant average abnormal returns of 1.14% in the 11‐day window surrounding the announcement, which is significantly smaller than that reported in similar studies on U.S. data. This small announcement return is not an underreaction. When we examine long‐term returns, we find that firms making open‐market share repurchase announcements earn significant negative abnormal returns of −7% in the year after the announcement of the repurchase intention. However, this is almost solely driven by the performance of firms announcing so‐called agency buybacks. Agency buybacks are open‐market repurchases that are purely tax driven, where the seller is aware that he is selling to the corporation. These firms underperform up to 1 year after the announcement of the acquisition, earning significant negative cumulative abnormal returns of −13.35% in the year following the repurchase announcement. In contrast, firms announcing on‐market (nonagency) repurchase programs earn insignificant negative abnormal returns of −2.01% in the year after the announcement date. We reject the hypothesis that, unlike the typical U.S. or Canadian firm (see Ikenberry, Lakonishok, and Vermaelen 1995, 2000), U.K. firms are repurchasing shares to take advantage of an “undervalued” stock price. The fact that excess returns are much smaller in the United Kingdom, both in the short run and the long run, is consistent with the hypothesis that the U.K. regulatory environment discourages share repurchases designed to take advantage of an undervalued stock price. As this is a major motivation behind repurchase programs in the United States, it is not surprising that share repurchases remain relatively unpopular in the United Kingdom.
The rest of the article is organized as follows: Section II reviews the literature on share repurchases. Section III discusses the tax and legal environment of U.K. buybacks. Section IV summarizes the hypotheses, while Section V describes the data. Section VI describes our methodology and results. Finally, Section VII concludes.
II. Review of the Literature
As a share repurchase can be considered, at the same time, a change in the payout, capital structure, and ownership structure of the firm, there is no shortage of theories to explain the average positive announcement return of around 3.5% that is observed in the United States (see, e.g., Dann 1981; Vermaelen 1981; Comment and Jarrell 1991; Ikenberry et al. 1995). According to the personal tax savings hypothesis, repurchases are beneficial because they are more tax efficient than paying out dividends. However, Bagwell and Shoven (1989) report that buyback activity in the United States, as measured by the total dollars distributed, did not decrease following the 1986 Tax Reform Act, which abolished the differential taxation between dividends and capital gains. In contrast, Lie and Lie (1999) show that managers were less likely to choose open‐market share repurchase programs over regular dividend increases after the 1986 act. They also show that managers are more sensitive to the tax situation of the shareholders if a large fraction of the shares is held by institutional investors.
To the extent that share buybacks are financed with debt, it can be argued that repurchases lower corporate taxes and, hence, lower the company’s cost of capital. However, this explanation seems less relevant for open‐market repurchase programs, which are often financed with excess cash.1
Another popular explanation has been that firms use share repurchase announcements to signal to the market that their shares are undervalued (see, e.g., Dann 1981; Vermaelen 1981; Hertzel and Jain 1991). The problem with considering open‐market repurchase programs as costly signals is that these programs are not firm commitments—it is costless to announce a repurchase and not carry it out later. Accordingly, Stephens and Weisbach (1998) study a sample of 450 open‐market share repurchase announcements in the United States between 1981 and 1990 and find that between 74% and 82% of the shares targeted at the time of announcement are later repurchased. They find that these actual share repurchases are negatively related to stock price performance after the buyback announcement. Similar findings are reported by Ikenberry et al. (2000), employing data on Canadian open‐market repurchase programs.
The strategic behavior reported by Stephens and Weisbach (1998) and Ikenberry et al. (2000) is consistent with the “good investment” hypothesis, which argues that companies buy back stock because their shares are undervalued. Note that, in contrast to the signaling story, this hypothesis assumes that the market’s response to the announcement is incomplete. Indeed, in a semistrong efficient market, after the announcement, the stock will no longer be undervalued, and it will be impossible for the average firm to buy back “undervalued” stock. Hence, the significant positive postannouncement excess returns reported by Ikenberry et al. (1995, 2000) in the United States and Canada are consistent with the good investment hypothesis.
An alternative explanation is that share repurchases reduce the agency costs of free cash flow: stock prices increase because the market is pleased that excess cash is not wasted in value‐destroying projects. Nohel and Tarhan (1998) examine a sample of 282 fixed price and Dutch auction share repurchases announcements between 1978 and 1991 in the United States and conclude that the positive investor reaction to share repurchases is better explained by the free cash flow hypothesis than the signaling hypothesis. However, unlike repurchase tender offers, open‐market programs are not firm commitments to distribute excess cash, so again it is not obvious that the agency‐cost hypothesis can be extrapolated to open‐market programs. It is also difficult to reconcile agency costs with the long‐term drift in share prices and strategic buying behavior. Specifically, the agency‐cost hypothesis (as well as other hypotheses assuming that share prices only increase if a buyback actually takes place) is inconsistent with the Ikenberry et al. (2000) finding that companies that do not repurchase any shares during the year after the repurchase authorization actually earn a much larger excess return than do firms that complete the buyback.
Besides saving taxes, correcting or exploiting undervaluation, or reducing agency costs, a number of other repurchase motivations, which are not necessarily expected to increase stock prices, have been proposed, such as fighting takeovers (Brown and Ryngaert 1991), getting rid of “temporary” nonoperating cash flows (Jagannathan, Stephens, and Weisbach 2000), or avoiding earnings per share dilution from executive stock option plans (Weisbenner 2000). The last explanation may not be very relevant for the United Kingdom, as repurchased shares have to be cancelled and cannot be held as treasury stock. In the United States, firms often take advantage of a low share price to buy shares that will be reissued to executives later on, after the stock has increased in value. In the United Kingdom such a “hedging” strategy is not possible: when executives exercise stock options, the company has to issue new shares.
In a study that considers all motivations simultaneously, Dittmar (2000) employs U.S. data to provide empirical support for these nonvalue creating motivations as well as for the “good investment” hypothesis. Of course, this is not surprising because regardless of why companies buy back stock, managers who care about the wealth of the long‐term investors (such as themselves) should only repurchase shares when the stock is correctly valued or undervalued.
To our knowledge, the only research on repurchases in the United Kingdom is by Rees (1996), who reports a 0.25% abnormal return in the 5 days surrounding the announcement of 882 open‐market repurchases made by U.K. firms between 1981 and 1990. His results, however, are not comparable with the 3.5% positive abnormal performance reported around the announcement date for share repurchases in the United States Unlike U.S. studies, which define the announcement date as the date when a company receives its buyback authorization, Rees defines the announcement date as the date when the company is effectively buying shares (the so‐called effective date). Unlike in the United States, U.K. companies have to notify the Stock Exchange before 12:00 noon on the day following a repurchase. As companies typically spread out an open‐market repurchase over several weeks or months, it is not surprising that, in his sample, companies repurchase only 0.5% of their shares, on average.2
III. Share Repurchases in the United Kingdom: Regulatory and Tax Framework
A. Regulation
In order to develop our working hypothesis, it is important to understand fully the legal and tax frameworks governing U.K. share repurchases and the difference, if any, between these frameworks and those in the United States. Although share repurchases were legalized in the United Kingdom by the Companies Act of 1981, various regulatory bodies have put restrictions on buyback activity. In general, these regulations reflect four main concerns.
First, share repurchases could allow a company to manipulate its stock price. Therefore, the Listing Rules of the London Stock Exchange specify that purchases within a period of 12 months of 15% or more of the company’s share capital must be made by a tender offer. Purchases below the 15% limit may be made in the open market as long as the price paid is not more than 5% above the average of the market value of those shares for the 10 business days before the repurchase is made.3
Second, by reducing the company’s capital, share buybacks could hurt creditors. Hence, the U.K. Companies Act states that only “distributable profits” or the proceeds of a fresh issue of shares (made for the purpose of the purchase) can be used to finance the purchase. A similar restriction exists in the United States. Moreover, debt covenants will often restrict the amount of buybacks, in the same way as do dividends.
Third, if managers will only buy back shares when their shares are undervalued, an open‐market purchase program where the seller is not aware that he is selling to the corporation essentially allows a company to engage in insider trading. The U.S. regulatory approach to this concern is to emphasize disclosure: companies should announce in advance that they have created an option to buy back stock. Moreover, U.S. regulators are generally not very concerned about insider trading as long as the insider trading profits flow back to the company. For example, section 16b of the Securities and Exchange Act of 1934 requires insiders to return all trading profits to the company if they buy and sell the company’s stock within a 6‐month period.
The U.K. approach is drastically different. Under the Model Code, companies are not allowed to buy back shares during periods when officers and directors are not allowed to trade in their company’s shares. In practice, this means that share repurchases are not allowed in the 2‐month period preceding the publication of annual earnings or semiannual earnings and in the 1 month before the publication of quarterly results. Moreover, under the Criminal Justice Act, the company cannot purchase shares when the directors are in possession of unpublished, price‐sensitive information.
Finally, U.K. regulators are concerned about the preemption rights of shareholders and therefore require a company to cancel all repurchased shares. So, the U.S. concept of “treasury stock” (i.e., shares that are repurchased but can be reissued without shareholder approval) does not exist in the United Kingdom, which means that repurchases are a much less flexible tool for managing the firm’s capital.
B. Taxation
The tax treatment of share buybacks in the United Kingdom is rather complicated and has been changed at various points in time. The complication arises from the fact that the United Kingdom has an imputation system, which is meant to reduce the double taxation of dividends. Under this system, shareholders receive credit for taxes paid by the company on earnings distributed as dividends and on the “distribution element” of share buybacks. The distribution element is defined as the difference between the market value of the repurchased shares and the book value of the corresponding paid‐in‐capital.
We start by first describing the tax regime for individual investors, as this regime has remained constant during our sample period. Next, we describe the tax regime for pension funds during four different tax regimes: (a) prior to September 21, 1994; (b) from September 21, 1994, to October 7, 1996; (c) from October 8, 1996, to July 1, 1997; and (d) after July 2, 1997.
1. Taxation of Individual Investors For the sake of clarity, we illustrate the taxation regime with a simple numerical example. The example, as well as the general principle, is summarized in table 1.
Assume a company with 100 shares outstanding and a share price of £1, which is considering distributing £10 either through a dividend of 10p per share or a repurchase of 10 shares at the current market price of £1. Assume that the original subscription price of the share is 20p per share. Assume further that the investor cost base (i.e., the price at which he purchased the shares plus an inflation allowance) is 50p and that his marginal tax rate is t. Hence, the investor cost base is higher in price than the subscription price, a reasonable assumption in an actively traded public corporation.
Dividend payment. If the company pays out £10 as a dividend, it has to pay ACT (advance corporation tax) on this dividend at a rate of 25%, or £2.5. This ACT can be offset against the company’s mainstream corporation tax. However, as ACT has to be paid within 14 days and other corporate taxes are paid much later, companies are effectively giving an interest‐free loan to the Inland Revenue.
The value of the dividend after personal taxes is 
where t is the investor’s marginal tax rate. In other words, the investor pays tax on the grossed‐up dividend (£12.5) but then can claim the ACT of £2.5 as a tax credit. Note that, for an investor with a tax rate of 20%, no additional taxes are due, and the dividend is worth £10. In this case, the imputation system avoids the double taxation of distributed profits. For the high tax bracket taxpayer (
%), the value of a £10 dividend after all taxes is £7.5.
Share repurchase. In order to calculate taxes, a distinction has to be made between an off‐market repurchase (such as a repurchase tender offer or a private repurchase) and an on‐market repurchase (or open‐market repurchase). Note that in the first case the seller is aware that he is selling to the corporation, and in the second case he is not.
Off‐market repurchase.— The difference between the repurchase price, £1, and the original subscription price, 20p, is defined as the “distribution element.” As in the case of dividends, the corporation has to pay ACT on this “dividend” at a rate of 25%. Hence, in our numerical example, the total distribution element is 
, and the corporation has to pay ACT of £2, which can be offset against the company’s mainstream corporation tax. The ACT can then be used as a tax credit: the investor who pays a tax of 
on the grossed‐up “dividend” can claim a tax credit of £2, so that the net tax paid on the distribution element is 
. However, this is not the end of the tax story: the difference between the investor’s cost base (50p) and the original issue price of the shares (20p) on his 10 shares is considered a capital loss of 
, which is subject to the same tax as ordinary income. However, capital losses can only be offset against capital gains.
Hence, for individual investors, the value of a £10 repurchase is worth (in £): 
Recall that, if, alternatively, the company had paid a £10 dividend, the after tax value would be equal to 
For 
%, the value of the repurchase alternative is £10.6, which exceeds the value of the dividend alternative (£10). For the taxpayer in the top bracket, with 
%, the value of the repurchase alternative is also higher: £9.2 versus £7.5. However, this results from the fact that we assumed: that the investor purchased the stock above the original subscription price (a reasonable assumption for most established companies) and that, without the repurchase, he would be subject to capital gains taxes. If the investor was not subject to capital gains taxes, the value of the repurchase alternative is 
. In this case, taxpayers at a low rate (
%) would be indifferent between a dividend payment and a repurchase, as in both cases they will receive £10. Taxpayers at a high rate (
%), however, would still prefer a repurchase (£8) to a dividend payment (£7.5). The reason is that, for high‐tax‐bracket individuals, the dividend tax credit does not entirely eliminate the double taxation of corporate income. So, as by definition, the “distribution element” is always smaller than the total payout, in general, we can state that individuals prefer off‐market share buybacks to dividend payments, as long as they are not liable for capital gains taxes (i.e., as long as the individual’s cost basis is significantly below the paid‐in‐capital per share.)
On‐market share repurchases.— When a shareholder sells on‐market, he will not know that the ultimate purchaser of the shares is the company itself. As he is selling to a market maker, his profit will be taxed as a capital gain, so no tax credit can be claimed. In this case, no general conclusions are possible as the attractiveness of the buyback program depends on the investor’s capital gains tax liability. Specifically, if the company repurchases 10/P shares at the market price P, and the investor purchased the shares at an average price B (the basis price), the total taxes (in £) owed after the repurchase are equal to 
If, alternatively, the company paid out the £10 as a dividend, the total tax would be 
A share repurchase is more profitable than a dividend payment if 
. For an investor in the 40% tax bracket, this means if 
is smaller than 62.5%. If the investor had been in the low tax bracket (20%), he would always prefer dividends to repurchases, unless if 
.
Summarizing, in the United Kingdom, high tax individuals tend to prefer share repurchase to dividends (unless the repurchase would generate substantial capital gains taxes), but, ceteris paribus, they would prefer an off‐market repurchase (tender offer or private purchase) to an open‐market buyback.
2. Taxation of Pension Funds Table 2 shows the value of £10 paid as a dividend or through a share repurchase during the four different tax regimes: regime 1 (prior to September 1994), regime 2 (September 1994–October 1996), regime 3 (October 1996–July 2, 1997), and regime 4 (after July 1997). Table 2 compares, during each regime, the cash flows received by the pension fund if the company decides to pay out £10 as dividends or through a share repurchase.
Dividend payments. Prior to July 2, 1997 (this means during three out of the four tax regimes), pension funds were entitled to a 25% dividend tax credit, despite the fact that they were exempt from taxes on dividends and capital gains (so, in this case, the name “tax credit” is somewhat misplaced). Hence, a £10 dividend was worth £12.5 after taxes. This law was changed after July 2, 1997, when U.K. tax authorities eliminated all tax credits for pension funds.
Taxation prior to September 1994.— For off‐market repurchases, pension funds are entitled to a tax credit of 25% on the distribution element of the repurchase. Recall that in our numerical example, the distribution element was 
, which entitles the investor to a £2 tax credit. Hence, for these investors, the £10 buyback will be worth £12. However, if the company had paid a £10 dividend, the tax‐exempt investor would have received £12.5. Hence, in contrast to individuals, tax‐exempt investors should prefer dividends to off‐market share buybacks for tax purposes.
However, pension funds could find an off‐market share buyback attractive for two reasons. First, for those pension funds that plan to sell their shares, selling to the company is more profitable than selling in the open market, for by selling to the company, they would receive a tax credit. Second, tax exempt investors could theoretically earn arbitrage profits by buying shares in the open market and tendering them to the company. The Inland Revenue has tried to prevent this type of arbitrage by specifying that a pension fund that buys shares after the announcement of a buyback and sells them to the company cannot claim a tax credit. Of course, this rule is ineffective if pension funds can predict which companies might shortly announce a buyback. It will also be ineffective if there is a long time lag between the announcement of buyback authorization and the actual repurchase. Discussions with bankers revealed that the Inland Revenue tries to counter such behavior by making the granting of any tax credit subject to antiavoidance rules. In particular, pension funds or other tax exempt investors have to convince the Inland Revenue that they are not simply selling the shares to the company in order to get the tax credit. For example, a pension fund that would sell its shares to the company and buy them back in the market a short period (less than 2 months) thereafter would not be able to claim the tax credit. In on‐market repurchases, no tax credits exist. Hence, as individual investors, pension funds prefer off‐market purchases to on‐market purchases.
We conclude that, prior to September 1994, unlike (high‐tax‐bracket) individuals, pension funds prefer dividend payments to share buybacks, regardless whether they are off‐market or on‐market. However, for pension funds that decided to sell their shares, selling to the company in an off‐market repurchase (tender offer or private transaction) was significantly more profitable than selling in the open market. Moreover, all investors prefer off‐market repurchases to on‐market repurchases, because of the tax credit on the implied distribution.
This analysis seems to imply that, before September 1994, a company that decides to repurchase shares and would like to ensure that its institutional investors are entitled to a tax credit should choose the tender offer route over announcing an open‐market share repurchase. The basic problem with the tender offer method, however, is that, without paying a large repurchase tender premium, there exists a significant probability that the stock price after the repurchase announcement, but prior to the offer expiration, will exceed the tender price. This will either make pension funds reluctant to tender or make them lose the tax credit because of antiavoidance rules. The Inland Revenue will not allow a tax credit if it can show that the pension fund tendered the shares only because of the tax credit. For example, if, after a repurchase tender offer at £20 per share, the stock price increases to £22, it is obvious that a pension fund only tendered (rather than sold in the market) because of the tax credit. This is effectively what happened in the 1993 Reuters tender offer. On July 23, 1993, Reuters announced a tender offer for 5.84% of its shares at a price of £14 per share. Although at the expiration date of the offer the stock price increased to £15.5, pension funds still tendered their shares at £14, because the tax credit attached to each share repurchased was worth £3.21. Because the Inland Revenue had given its guarantee prior to the offer that all tax‐exempt investors would be entitled to the tax credit, pension funds received tax credits for more than £20 million. After the Reuters buyback, the Inland Revenue announced that it would no longer give guarantees to investors that they would be entitled to the tax credit.
Selling shares to the company through a private transaction also does not guarantee that a pension fund will be entitled to a tax credit. The selling institution has to convince the Inland Revenue that it is not simply selling the shares to the company in order to obtain the tax credit. This will be difficult, unless the company buys back the stock at a premium. However, other investors are likely to object to a transaction that so obviously favors one investor over another.
Because of all of these legal uncertainties, we indicate in table 2 that, prior to September 1994, the expected value of a £10 off‐market repurchase was less than £12, which makes the case for dividend preference even stronger.
Taxation after September 1994: Introduction of agency buybacks.— Agency buybacks were invented by innovative bankers at Barclays de Zoete Wedd, who arranged the first agency buyback in September 1994 for Northern Electric. In an agency buyback the shareholders sell their shares to a broker who is acting as an agent for the company. Because the agent (i.e., a bank) would contact pension funds in advance and give them priority over individual shareholders, the open‐market repurchase effectively became a private repurchase directed at specific institutional clients. In contrast to a normal open‐market repurchase (as in the United States or in the United Kingdom prior to September 1994) where the buyback is completed in several small transactions over several months, agency buybacks were often completed in a few hours. For example, Anglian Water’s share buyback program for 10% of its shares was completed in less than a couple of hours by 10:00 a.m. on August 16, 1995.
Moreover, because an agency buyback appears to be a repurchase in which all shareholders participate, it is easier to convince the Inland Revenue that the antiavoidance rules do not apply. This made agency buybacks much more attractive than other off‐market mechanisms (such as tender offers and private purchases) or on‐market repurchases (see table 2)
Taxation after October 8, 1996.— On October 8, 1996, the tax treatment of share buybacks was modified significantly for tax exempt investors such as pension funds. These investors could no longer recover any tax credits associated with the distribution element of the off‐market repurchase. Hence, as table 2 shows, for pension funds, all off‐market repurchases and on‐market repurchases became equally unattractive.
Hence, if the tax consequences for pension funds is an important factor in deciding to repurchase shares, buybacks aimed at these investors (i.e., mainly off‐market share buybacks) should have fallen dramatically.
Taxation after July 2, 1997.— In July 1997, the U.K. tax authorities did not change the tax treatment of repurchases, but, as mentioned before, it eliminated all tax credits for dividends. Hence, after this, pension funds became indifferent between dividends and share repurchases, as is the case in the United States. Hence, we expect that share repurchases should become more popular after July 2, 1997.
IV. Hypotheses
In this article, we test three hypotheses regarding (1) the choice of repurchase mechanism, (2) the size of announcement returns, and (3) the difference in long‐term return between agency buybacks and other open‐market repurchase programs.
Hypothesis 1. Repurchase mechanisms in the United Kingdom are driven by tax considerations of pension funds.
As pension funds have a major influence on corporate decisions, we expect them to have a substantial impact on the method chosen by the company to repurchase shares. In particular, we expect that, during the period in which agency buybacks existed and pension funds were entitled to tax credits (September 21, 1994–October 8, 1996), agency buybacks will be the predominant form of repurchase, at the expense of other mechanisms such as tender offers or privately negotiated share repurchases. Because agency buybacks are open‐market share repurchases that provide tax benefits to pension funds, we expect open‐market repurchase activity to intensify after September 1994 and to decline after October 1996. Finally, the abolition of the imputation system after July 1997 should lead to an increase in share repurchase activity again, as dividends and buybacks now have no more tax consequences for pension funds, as in the United States.
Hypothesis 2. Announcement returns following U.K. buyback authorizations will be smaller than the ones typically observed in the United States.
Because of the restrictions on buying back shares prior to 2 months before earnings announcements and the need to disclose partial repurchases almost immediately, companies will find it more difficult to take advantage of an undervalued stock price. The announcement returns will be small, in particular for agency buybacks, as in these cases the buyback is tax driven and the seller is aware that he is selling to the corporation.
Hypothesis 3. Long‐term returns after agency buybacks will be much smaller than they are after other open‐market share repurchase programs.
One often mentioned reason for U.S. share buybacks is that the management believes that its shares are a “good investment.” Consistent with this hypothesis, positive long‐term excess returns are observed after U.S. and Canadian open‐market share repurchase programs of high book‐to‐market stocks (Ikenberry et al. 1995, 2000) and after the expiration of repurchase tender offers made by small firms (Lakonishok and Vermaelen 1990). Recall that agency buybacks (i.e., open‐market share repurchases between September 1994 and October 1996) were designed to ensure that pension funds could claim a tax credit. Hence, we expect these repurchases to be less motivated by the fact that the company’s shares were a good investment.
V. Data
Our sample is drawn from the SDC online Mergers and Corporate Transactions database, using the following criteria:
| • | The transaction is classified either as a self‐tender offer or a repurchase. | ||||
| • | The firm is listed in the United Kingdom. | ||||
| • | The consideration sought in the transaction is listed as common shares. | ||||
| • | The announcement date lies between January 1, 1985, and June 30, 1998.4 | ||||
Table 3 summarizes the sample screening process for the analyses. The SDC database lists 293 repurchases and 18 tender offers during the period from January 1980 to June 1998. Applying the screens above reduces the sample to 264 repurchase announcements and 14 tender offers. Of these repurchases, 198 are open‐market share repurchases and 68 are privately negotiated. Of the 14 tender offers, 12 are fixed‐price, self‐tender offers and 2 are Dutch auction self‐tender offers.
We analyze two types of announcements. First, we analyze announcements that the company intends to repurchase shares, “repurchase intentions.” This will typically be an announcement that the board has approved a stock repurchase program and will be seeking approval of this program at the next shareholders meeting. The second announcement is an announcement that the company has completed a stock repurchase (program), “repurchase completions.”5 Of the 198 open‐market repurchases, 144 were classified as repurchase intentions, while 101 were classified as repurchase completions.6 Similarly, we have 42 privately negotiated repurchase intentions and 45 privately negotiated repurchase completions.
Panel B in table 3 reports the number of firms with data available to measure performance. For short‐term performance, we obtain daily returns data from Datastream. The panel reports the number of repurchase firms that announced (or completed) open‐market or privately negotiated repurchases or self‐tender offers listed on Datastream.
Long‐term performance is measured using monthly returns obtained from the London Share Price database (LSPD), which covers monthly returns up to December 1997. Since we measure performance for the 1 year after the announcement of the repurchase, we eliminate from our analysis all firms that made announcements after January 1997. Further, since we test abnormal performance after adjusting for market capitalization (from LSPD) and price‐to‐book ratio (from Datastream), panel B in table 3 also lists the number of firms that are listed on both LSPD and Datastream.
VI. Methodology and Results
A. Descriptive Statistics
Tables 4 and 5 report some descriptive statistics for the overall repurchase sample. Table 4, panel A, reports the distribution of the announcements and the value of the transaction by year for open market, privately negotiated share repurchases, and self‐tender offers. As is the case in the United States, open‐market share repurchases are more common than privately negotiated transactions and tender offers. As can also be seen, open‐market share repurchase activity increases significantly in 1994, when agency buybacks were introduced, quintupling from five and six open‐market announcements in 1992 and 1993 to 34 in 1994. Privately negotiated share repurchases also increase, though less markedly than open‐market repurchases. In contrast, self‐tender offers disappear in both 1994 and 1995. When off‐market repurchases lost their tax advantage in 1996, companies started announcing self‐tender offers again.
Table 4, panel B, expands on the information in panel A, recording the number and value by year of each of our four repurchase samples—open‐market and privately negotiated repurchase intentions and repurchase completion samples, respectively. Note that some announcements are mixed: some companies announce plans to repurchase shares in the open market, as well as in private transactions. Again, a similar pattern is observed as in panel A. Open‐market share repurchase activity increases substantially in 1994 and declines in 1996. Twenty‐eight and 31 open‐market share repurchases are completed in 1994 and 1995, respectively, while only 10 are completed in 1996 and nine in 1997. A less marked pattern is observed in privately negotiated purchases.
Table 5, panel A reports the percentage of shares sought and acquired by firms making repurchases. As can be seen, the repurchases that we study are substantial in contrast to those studied by Rees (1996); companies announce that they may buy back a median 6.70% of their own stock in the open‐market sample, while the corresponding percentage for privately negotiated purchases is 14%. Similarly, the percentage of shares actually acquired in completed repurchases is also significant; companies buy back a median 4.2% of their stock in open‐market transactions and 16% in privately negotiated share repurchases.
Finally, table 5, panel B, reports the industry groupings of the repurchasing firms. Note that 35 out of the 264 announcements (or around 13%) are made by banks and financial services companies. However, only two banks (National Westminster with two announcements and Barclays with six announcements) announced more than half of all the bank share buybacks. All other buybacks were made by investment or insurance companies. Hence, unlike the United States, where more than 40% of all buybacks in 1996 were announced by banks and savings and loans, share buybacks are not very popular in the U.K. banking sector. Note also the large importance of utilities: more than 15% of the announcements were made by privatized utilities (gas, electricity, and water). The importance of repurchase activity by utilities may have been driven by regulatory and political considerations. First, regulators put pressure on utilities to lower their cost of capital by increasing leverage. Second, the Labour (Socialist) party had promised (as soon as they would take over power from the Tory [Conservative] government) to impose a tax on “windfall” profits (which may be correlated with excess cash) made by privatized utilities.
B. Descriptive Statistics for Tender Offers
Table 6 shows some detailed information on the repurchase self‐tender offers in the sample. Panel A in the table shows descriptive statistics, where available. Panel B documents information on the price and return earned by the companies around the announcement date. It documents the percentage of shares sought, tendered, and repurchased. In addition, the panel also lists the cumulative excess return to the nontendering shareholders for a period from 15 days prior to the announcement to 5 days after the expiration. Excess returns are computed with respect to the FTSE All Share market index using the market‐adjusted model. The panel also lists the tender offer premium over the stock price 15 days prior to the announcement. Finally, it lists the weighted average abnormal return to all shareholders involved (INFO).7
The first repurchase tender offer was launched by International Business Communications (IBC) in January 1989; IBC made an offer to buy back 40% of its own shares at a premium of 56% above the preannouncement price (see Vermaelen [1995] for more details). Although IBC’s stock price increased by 37% as a result of the repurchase, the company’s example did not generate a large following. The reason, according to some practitioners, is that IBC subsequently (July 1990) defaulted on the debt issued to finance the repurchase. Moreover, the remaining tender offers in the sample involve small repurchase premiums and announcement returns. The small tender premiums, together with the fact that tender offers all but disappeared for 3 years after the July 1993 tender offer of Reuters, are consistent with the hypothesis that the choice of the repurchase mechanism is influenced by tax considerations. Recall that, after the Reuters tender offer, the Inland Revenue would no longer guarantee in advance that investors would be entitled to a tax credit. This, apparently, made the tender route less attractive, especially during the 1994–96 period when agency buybacks became a far cheaper tax‐effective mechanism. Note, however, that as the number of tender offers is always small, there must be other nontax factors that explain the lack of enthusiasm for tender offers in the United Kingdom.
One reason why tender offers never have been popular may be the almost obsessive preoccupation of U.K. regulators with ensuring equal treatment of stockholders. Note that, in a tender offer, unless everyone tenders, tendering and nontendering shareholders will earn different returns, depending on the tender price and the price after expiration of the offer. It is therefore quite common, unlike in the United States (see, e.g., Comment and Jarrell 1991), that insiders tender their shares in the same proportion as other investors. However, by tendering their shares, it becomes difficult to convince noninsider shareholders that there is a cost to false signaling (see Vermaelen 1984). Moreover, it should be noted that in the United States the number of tender offers is a small fraction of total repurchase activity. For example, Grullon and Ikenberry (2000) report that in 1999 there were 1,212 open‐market repurchases and only 40 tender offers. The real puzzle is not so much why the number of tender offers is small but, rather, why repurchase activity is so tiny compared with the same activity in the United States
C. Does Taxation Affect Share Repurchase Activity and Method?
Table 7 expands on the preliminary evidence discussed in tables 4 and 5 and the previous section. It shows the average number of repurchases announced and completed per month in the four tax regime periods. Period 1 analyzes repurchase activity before September 1994, that is, before agency buybacks were created. Period 2 analyzes repurchase activity between September 1994 and October 1996, that is, during the period when off‐market open‐market share repurchases (agency buybacks) had a tax advantage. Period 3 analyzes monthly share repurchase activity between October 1996 and July 1997, when dividends received a tax credit, but share repurchases did not. Finally period 4 analyzes the period after July 1997, when all tax credits were abolished, thus increasing the relative attractiveness of share repurchases over dividend payments.
Consistent with the hypothesis that taxation has a significant effect on share repurchase activity and method, table 7 shows clearly that the introduction of these agency buybacks led to a significant increase in buyback activity. Apparently, the ability of pension funds to recover dividend tax credits has a major impact on the popularity and format of buybacks. The fact that privately negotiated offers also increased after the introduction of agency buybacks is probably caused by the fact that SDC has classified certain agency buybacks as private purchases. Recall that, although in theory agency buybacks are open‐market share repurchases directed to all shareholders, in practice the agent typically gives priority to large institutional investors.
After the change in taxation policy in October 1996, the average number of repurchases completed per month dropped substantially from 2.7 to less than half this value for all repurchases. However, this decline is only observed for the completed sample, not for the sample of announced repurchases, where the repurchase frequency remains the same as it did prior to the tax law change. Apparently, many companies that announced buyback programs did not complete them. The cancellation may have been a response to the public criticism of companies who continued buyback programs rather than paying dividends. For example, in February 1997, the Financial Times stated that “yesterday’s buyback from TRJB mining suggests Britain’s pension fund managers are still letting value evaporate. How? Had RJB handed cash back through a special dividend instead, tax exempt shareholders would have received cash from the government on top. With a buyback, not only is this foregone but buy‐backs also take place at a pointless premium.”8 Another reason for the continuing popularity of share repurchases, relative to dividend payments, may be the preoccupation of many chief financial officers with “earnings per share enhancement” as well as the concern that the value of executive stock options will fall significantly when a (large) dividend is being paid (Jolls 1998; Fenn and Liang 2000).
Finally, after July 1997, when share buybacks became relatively more attractive, share repurchase completions rose again to an average of 1.67 per month until June 1998. The fact that the buyback intensity did not return to its pre‐1996 level is consistent with the hypothesis that many of the agency buybacks were arranged to generate tax credits for pension funds. Another reason may be the emergence of a new financial innovation: the “special dividend‐with share consolidation.” In this case, a company creates new “B” shares, which are distributed to the existing shareholders. These shares are then immediately repurchased. In order to avoid the price of the ordinary shares dropping by the payout per share (and hence lowering the value of executive stock options), the number of outstanding shares is reduced by a so‐called share consolidation (i.e., a reverse stock dividend). The repurchase of the B shares qualifies as a “capital repayment” rather than a “distribution.” The difference is that, on a capital repayment, no ACT has to be paid. Note that, as soon as dividend payments and repurchases were no longer generating tax credits, the main preoccupation of some companies became the avoidance of ACT.9 Although these transactions also involve a “repurchase” (i.e., of the B shares), they are not included in our sample 1.10
Table 8 presents logistic regressions of the probability that a firm chooses to announce or complete an open‐market share repurchase rather than a self‐tender offer or a privately negotiated repurchase.11 We use an indicator variable that takes the value of one if the repurchase or tender offer was announced (completed for repurchase completions) between September 21, 1994, and October 8, 1996, and zero otherwise, to proxy for the tax advantage to open‐market agency buybacks during this period. We also use the percentage of shares sought and the payout to the shareholders in the transaction (computed as the value of the transaction divided by the market capitalization of the firm) as controlling variables, since evidence from our univariate statistics in table 5, panel A, shows that these are different across open‐market and privately negotiated repurchases.
Consistent with hypothesis 1, the probability of a firm announcing (or completing) an open‐market share repurchase rather than a privately negotiated repurchase or a self‐tender offer is significantly higher in the agency buyback period even after controlling for the size of the firm, its payout, and its book‐to‐price ratio. The payout is also significantly negatively related to the probability of a firm using an open‐market repurchase, a result consistent with the evidence in table 5. Neither the size of the firm nor the book‐to‐market ratio is related to the probability of announcing open‐market share repurchase programs. The explanatory power of these regressions is also reasonably high, with a maximum generalized 
of 18% for repurchase announcements and 28% for repurchase completions.
D. Does the Market React Favorably to Share Repurchase Announcements in the Short Term?
Table 9 and figure 2 report the results of a short‐term event study around the dates of announcement for open‐market share repurchases and privately negotiated share repurchases for different event periods around the announcement date. Abnormal returns are computed with respect to the FTSE All Share market index using the market adjusted model. T‐statistics are calculated using a 20‐day holdout period, ranging from −35 to −16 days before the announcement of the repurchase (for repurchase intentions) or the completion of the repurchase (for repurchase completions).
Fig. 2.— Daily cumulative abnormal returns following open‐market share repurchase announcements and completions in the United Kingdom, 1985–98. This figure plots the cumulative abnormal returns (CAR) around the announcement date (for repurchase intentions) and the effective date (for repurchase completions). The CARs are computed using the market‐adjusted model with respect to the FTSE all share index from Datastream. Separate CARs are plotted for firms making repurchases within the agency buyback period from September 21, 1994, to October 8, 1996, and for firms making repurchases outside this period. a, Open‐market repurchase intentions. b, Open‐market repurchase completions.
Table 9 shows the cumulative average abnormal return during various subperiods around the announcement date, for both the open‐market repurchase intentions and the repurchase completions sample. In contrast to results reported by others on U.S. data (Dann 1981; Vermaelen 1981; Comment and Jarrell 1991; Ikenberry et al. 1995), open‐market share repurchase announcements are not preceded by significant negative excess returns. During the announcement period (which we define as the period from day −5 to +5, in order to allow for information leakage), stock prices increase significantly by approximately 1.14%, on average. Hence, announcement returns are significantly smaller than the ones reported in U.S. studies. We examine the difference between returns to firms announcing agency buybacks and firms announcing “normal” on‐market open‐market repurchase programs. Table 9 shows that the significant increase in stock prices resulting from the announcement of share repurchase programs is driven largely by the announcement of on‐market share repurchase programs: from day −5 to +5, on‐market repurchase announcements generate statistically significant excess returns of 1.59%, while the corresponding number for agency buybacks is an insignificant 0.08%. Similarly, over the −2 to +2 day period, firms announcing on‐market repurchase programs earn statistically significant excess returns of 1.38%, while the firms announcing agency buybacks earn an insignificant 0.38% over this period. This is also illustrated in figure 2. However, the excess returns are not significantly different from each other over any of these intervals. So the evidence that agency buybacks are perceived as less value‐creating is very weak. However, the results are broadly consistent with hypothesis 2: because of regulatory constraints, share buybacks are much less perceived as signals that the company’s stock is undervalued.
Table 9 also shows that the market does not react significantly to the completion of a repurchase either for an agency buyback or for an on‐market repurchase completion. Finally, the table also shows that privately negotiated repurchases carry no information content either. Again, this is not surprising, as it is unlikely that the company can buy undervalued shares from large, sophisticated investors. Moreover, as private purchases are generally considered as off‐market for tax purposes, the repurchase may have been simply motivated to capture a tax credit, as in the case of agency buybacks. Consequently, the results for these repurchases is not too surprising.
E. Do Firms Earn Excess Returns in the Long Term after a Share Repurchase Announcement? Is This Affected by the Tax Consequences of the Repurchase?
The small short‐term announcement returns may be misleading as the market may not capture all the effect of the announcement in the short‐term. Ikenberry et al. (1995) argue, for example, that firms announcing share repurchases continue to earn abnormal returns up to 3 years after the announcement date. In order to examine the long‐term abnormal performance for repurchasing firms, we analyze a subset of the repurchase sample consisting of repurchases announced between January 1985 and December 1996 (sample VA in table 3). We then compute abnormal returns for each type of repurchasing firm separately.
Fama and French (1998) show that, similar to those in the United States, value stocks tend to have higher returns than growth stocks around the world, suggesting that the return premium for value stocks is real and universal. We therefore compute long‐horizon abnormal returns with respect to a size‐ and price‐to‐book benchmark portfolio, formed using the sequential sort procedure employed by Ikenberry et al. (1995).
Every year in January, we form five size quintiles on the basis of the market capitalization of firms listed on both LSPD and Datastream. Then we rank each firm listed on both LSPD and Datastream into one of five portfolios formed on the basis of these break points. This quintile break‐point formation and ranking procedure is repeated every January between January 1985 and December 1996. These are further sorted using price‐to‐book ratio into quintiles.12 Twenty‐five portfolio returns are formed every month by averaging the monthly returns for these 25 portfolios. These returns are then used as benchmarks to calculate abnormal performance. Abnormal returns are calculated for each firm relative to its size and price‐to‐book‐based benchmark (as the difference between its monthly return and that of its control portfolio) every month from 12 months before to 12 months after the month of the repurchase announcement (or completion). Cumulative abnormal returns (CARs) are then calculated by averaging across all repurchasing firms every month and then summing these averages over time. We test the statistical significance of these results using bootstrapping (as applied by Ikenberry et al. [1995]), a methodology that is robust to the problems that plague standard long‐horizon tests of statistical significance.13 (See Barber and Lyon [1997]; Kothari and Warner [1997]; and Rau and Vermaelen [1998] for details of these problems. Lyon, Barber, and Tsai [1999] find that the bootstrap method yields well‐specified test statistics and find, moreover, that this method is more powerful than the control firm method, a method also commonly used to detect abnormal performance in event studies.) Finally, since Kothari and Warner (1997, p. 332, n. 4) show that the empirical pseudodistribution has a mean different from zero, we subtract the mean CAR for the empirical distribution from the CAR value for the sample. This bias‐adjusted CAR (BCAR) value gives us a better idea of the economic significance of the results. We then examine these returns to see if they are different across different types of repurchase announcements. Since the abnormal return distributions are skewed, we use a nonparametric median test to test the null hypothesis that there is no difference between the abnormal returns earned by firms announcing agency buybacks and firms announcing on‐market buybacks, against the alternative hypothesis that firms announcing agency buybacks earn smaller abnormal returns than firms announcing on‐market buybacks. The results are reported in table 10 and illustrated in figure 3.
Fig. 3.— Monthly cumulative abnormal returns following open‐market share repurchase announcements and completions in the United Kingdom, 1985–96. This figure plots the cumulative abnormal returns (CAR) from 1 year before to 1 year after the announcement date (for repurchase intentions) and the effective date (for repurchase completions). The CARs are computed using a size and price‐to‐book ratio portfolio benchmark, with data from Datastream and London Share Price Database. Separate CARs are plotted for firms making repurchases within the agency buyback period from September 21, 1994, to October 8, 1996, and for firms making repurchases outside this period. a, Open‐market repurchase intentions. b, Open‐market repurchase completions.
Over the whole 1‐year period after the repurchase announcement, there is no evidence that share repurchase announcements (intentions or completions) are followed by significant positive excess returns. There is also no evidence, unlike in the United States, that buyback announcements are preceded by negative abnormal returns, on average. Firms announcing share repurchases earn insignificant negative abnormal returns of −2.47% over the year before the announcement and significant negative abnormal returns of −7.01% in the 1‐year postannouncement period alone. Closer inspection reveals, though, that this result is largely driven by announcements of agency buybacks, which earn statistically significant negative abnormal returns of −8.59% over the year before the announcement of the buyback and continue to earn significant negative cumulative abnormal returns of −13.35% in the 1‐year postannouncement period. In contrast, firms announcing on‐market share repurchase programs earn insignificant abnormal returns over both of these horizons. The difference between the abnormal returns earned by the on‐market and agency buyback firms is significant over the 1‐year postannouncement period but not over the preannouncement period. These results are consistent with hypothesis 3, the hypothesis that long‐term returns after agency buybacks will be much smaller than after other open‐market share repurchase programs. This result supports the hypothesis that the purpose of agency buybacks was mainly to allow pension funds to claim a tax credit by selling the shares to the company and, possibly, repurchasing them a few months later. Also, the fact that firms announcing on‐market share repurchases do not earn significant long‐horizon abnormal returns is consistent with the hypothesis that the U.K. regulatory environment discourages share repurchases designed to take advantage of an undervalued stock price.
Note that an alternative explanation could be that the excess returns reported in other papers for firms announcing share repurchases (see, e.g., Ikenberry et al. 1995, 2000) result from some type of misspecification. The misspecification story, however, is inconsistent with the relationship between strategic trading and postannouncement returns reported in Ikenberry et al. (2000). Moreover, each of these papers has 20 times as many observations as we have, so future research on the United Kingdom and other European countries will be necessary to settle this issue.
VII. Conclusion
The U.K. repurchase scene is different from the U.S. scene in many respects. Although the United Kingdom is the European country where buybacks are most popular, the extent of repurchase activity is tiny in comparison with that in the United States, where approximately 100 U.S. companies announce open‐market share repurchases each month.
We argue that this is because the U.K. tax and regulatory environment makes share buybacks much less attractive than they are in the United States. Unlike in the United States, pension funds have had a clear tax preference for dividend payments, at least until the U.K. government abolished the imputation system in July 1997. Prior to this abolition, pension funds were interested in buybacks only if they could be designed as off‐market for tax purposes. In that sense, the agency buyback was a major financial innovation that made open‐market share repurchases attractive for a while, that is, until the Inland Revenue disallowed them. Moreover, restrictions on buybacks prior to earnings releases or other company‐specific information reduce the value of a repurchase as a signaling mechanism or as an instrument to transfer wealth from selling shareholders to long‐term shareholders. In effect, U.K. regulators are much more concerned that share buyback programs effectively become legal mechanisms to engage in insider trading with company funds (Ikenberry and Vermaelen 1996). And finally, in the United Kingdom, repurchased shares have to be cancelled and cannot be held in treasury stock. Hence, buying back shares when they are “cheap” and reissuing them to executives when they exercise their stock options (presumably after a run‐up in the stock price) is not an option available to U.K. companies.
The market apparently agrees: announcement returns around buyback authorizations are smaller than the values reported in U.S. studies. This is even more surprising considering that share buybacks are more common in the United States, and in theory the market should react only to unexpected events. Unlike in the United States and Canada, this small announcement return is not an underreaction: long‐term (1 year) postannouncement returns are not significantly positive.
Finally, the data also show that the form and intensity of U.K. repurchase activity is influenced by the tax consequences for pension funds. When the Inland Revenue decided to no longer guarantee that pension funds were entitled to a tax credit in tender offers, tender offers virtually disappeared. Share repurchase activity took off again when investment bankers designed a clever way to allow pension funds to capture tax credits without violating the antiavoidance rules. After October 1996, when the Inland Revenue changed the rules, share repurchase completions declined and only restarted after the tax laws were changed again, making share repurchases more attractive. The importance of taxation in the United Kingdom is also seen in the short‐term and long‐term performance of firms announcing open‐market share repurchases. Firms announcing agency buybacks, which are driven by tax considerations, perform much more poorly, at least in the long run, than firms that announce on‐market buybacks, which are more likely to be driven by undervaluation or signaling concerns.
The finding that the tax system is an important determinant of the choice of payout mechanism is consistent with the hypothesis that the tax treatment of important investors, such as pension funds, determines the payout policy. While it is often argued that large stakeholders such as pension funds are beneficial for small shareholders because they improve monitoring and corporate governance (see, e.g., Del Guercio and Hawkins 1999), our results suggest that this benefit may reduced if large institutional investors are taxed differently from individual investors.
References
- Bagwell, L. S., and Shoven, J. B. 1989. Cash distributions to shareholders. Journal of Economic Perspectives 3 (Summer): 129–40.
- Barber, B. M., and Lyon, J. D. 1997. Detecting long‐run abnormal stock returns: The empirical power and specification of test statistics. Journal of Financial Economics 43:341–72.
- Brown, D. T., and Ryngaert, M. D. 1991. The mode of acquisition in takeovers: Taxes and asymmetric information. Journal of Finance 46:653–69.
- Comment, R., and Jarrell, G. A. 1991. The relative signalling power of Dutch‐auction and fixed‐price self‐tender offers and open‐market share repurchases. Journal of Finance 46:1243–71.
- Cox, D. R., and Snell, E. J. 1989. The Analysis of Binary Data. 2d ed. London: Chapman & Hall.
- Dann, L. Y. 1981. Common stock repurchases: An analysis of returns to bondholders and stockholders. Journal of Financial Economics 9:115–38.
- Del Guercio, D., and Hawkins, J. 1999. The motivation and impact of pension fund activism. Journal of Financial Economics 52:293–340.
- Dittmar, Amy K. 2000. Why do firms repurchase stock? Journal of Business 73, no. 3:331–55.
- Fama, E., and French, K. R. 1998. Value versus growth: The international evidence. Journal of Finance 53:1975–99.
- Fenn, G. W., and Liang, N. 2001. Corporate payout policy and managerial stock incentives. Journal of Financial Economics 60:45–72.
- Grullon, G., and Ikenberry, D. 2000. What do we know about share repurchases? Journal of Applied Corporate Finance 13:31–51.
- Hertzel, M., and Jain, P. C. 1991. Earnings and risk changes around stock repurchase tender offers. Journal of Accounting and Economics 14:253–74.
- Ikenberry, D.; Lakonishok, J.; and Vermaelen, T. 1995. Market underreaction to open market share repurchases. Journal of Financial Economics 39:181–208.
- Ikenberry, D.; Lakonishok, J.; and Vermaelen, T. 2000. Stock repurchases in Canada: Performance and strategic trading. Journal of Finance 55:2373–97.
- Ikenberry, D., and Vermaelen, T. 1996. The option to repurchase stock. Financial Management 25:9–24.
- Jagannathan, C.; Stephens C. P.; and Weisbach, M. S. 2000. Financial flexibility and the choice between dividends and stock repurchases. Journal of Financial Economics 57:309–54.
- Jolls, C. 1998. Stock repurchase and incentive compensation. Working Paper no. 6467. Cambridge, Mass.: NBER.
- Kothari, S. P., and Warner, J. B. 1997. Measuring long‐horizon security price performance. Journal of Financial Economics 43:301–39.
- Lakonishok, J., and Vermaelen, T. 1990. Anomalous price behavior around repurchase tender offers. Journal of Finance 45:455–77.
- Lie, E., and Lie, H. J. 1999. The role of personal taxes in corporate decisions: An empirical analysis of share repurchases and dividends. Journal of Financial and Quantitative Analysis 34:533–52.
- Lyon, J. D.; Barber, B. M.; and Tsai, C.‐L. 1999. Improved methods for tests of long‐run abnormal stock returns. Journal of Finance 54:165–201.
- Nohel, T., and Tarhan, V. 1998. Share repurchases and firm performance: New evidence on the agency costs of free cash flow. Journal of Financial Economics 49:187–222.
- Rau, P. R., and Vermaelen, T. 1998. Glamour, value and the post‐acquisition performance of acquiring firms. Journal of Financial Economics 49:223–53.
- Rees, W. 1996. The impact of open market equity repurchases on U.K. equity prices. European Journal of Finance 2:353–70.
- Stephens, C. P., and Weisbach, M. S. 1998. Actual share reacquisitions in open‐market repurchase programs. Journal of Finance 53:313–33.
- Vermaelen, T. 1981. Common stock repurchases and market signaling: An empirical study. Journal of Financial Economics 9:139–83.
- Vermaelen, T. 1984. Repurchase tender offers, signaling and managerial incentives. Journal of Financial and Quantitative Analysis 19:163–81.
- Vermaelen, T. 1995. International Business Communications. In P. Stonham and K. Redhead (eds.), European Casebook on Finance. Hertfordshire: Prentice‐Hall International.
- Weisbenner, S. 2000. Corporate share repurchases in the 1990s: What role do stock options play? Working paper. Urbana‐Champaign: University of Illinois.
-
* We would like to thank Kimberly Rodgers and Orlin Dimitrov for able research assistance and Caroline Goodall for providing information on the U.K. regulatory and tax treatment of share buybacks. We would also like to thank an anonymous referee, Wolfgang Aussenegg, Jay Dahya, David Denis, David Douglas, David Reeb, Doug Emery, and seminar participants at Case Western Reserve University, Purdue University, the University of Miami, the University of Strathclyde, the European Finance Association meetings (Vienna, 1997), the Financial Management Association meetings (Chicago, 1998), and the European Financial Management Association meetings (Paris, 1999) for their helpful comments.
-
1. Unless, of course, the cash is invested in bonds that generate taxable income for the corporation. In this case, excess cash can be considered negative debt.
-
2. For example, between October 20, 1989, and August 1, 1990, B.A.T Industries made 94 repurchases of ordinary shares at prices ranging between 581p and 840p and amounts between 100,000 to 7,850,000 shares at a time. Rees (1996) treats these as 94 different events, while we consider the event to be the announcement of the program itself on September 26, 1989.
-
3. The U.S. antimanipulation provisions are summarized in Securities and Exchange Commission rule 10b‐18, introduced in 1982. The rule specifies that (1) trades should be made through one broker; (2) none of the trades should be made at the opening transaction during the last half hour of trading on a given day; (3) none of the trades may be completed at a price exceeding the highest current bid price or the last independent sale price, whichever is higher; and (4) the total of such purchases in a day should not exceed 25% of the average daily trading volume for the preceding 4 weeks. Hence, by putting limits on price movements and trading volume, regulators have significantly reduced a company’s ability to manipulate its stock price.
-
4. The SDC database does not contain any announcements by U.K. firms before January 1, 1985.
-
5. Note that, in many cases, these repurchase completions will actually be only partial completions, i.e., the company may continue with a repurchase program even after announcing that the current program is over.
-
6. We classified a repurchase as a repurchase announcement if it was an intended announcement or if the completion date differed from the announcement date in the SDC database. A repurchase was classified as a repurchase completion if the completion date was listed in the database, regardless of whether the announcement date differed from the completion date. For these repurchases, the event date was the effective date. Six intended announcements are counted both as an open‐market transaction as well as a private transaction.
-
7. INFO is computed as the
of shares 
of shares repurchased. -
8. February 6, 1997, p. 25.
-
9. ACT can only be offset against tax paid on U.K. corporate profits; this creates a problem for international companies that generate most of their profits abroad, as well as for companies that have no operating profits.
-
10. Examples of such transactions include payouts made after July 2, 1997, by Diageo, BTR, British Gas, EMI, and Bass.
-
11. We classify a repurchase as an open‐market repurchase if SDC reports that any part of the repurchase was carried out in the open market. In an alternative specification, we classified repurchases as open market if no part of it was privately negotiated. We excluded self‐tender offers in a third specification. We also classified privately negotiated and self‐tender offers separately in a fourth specification. Our results are robust to these alternative specifications.
-
12. Price‐to‐book ratio is obtained from the PTBV variable from Datastream.
-
13. For each repurchase firm in the sample, we randomly select with replacement a firm listed on LSPD and Datastream that has the same size and price‐to‐book ranking at that point in time. This matching firm is treated as though it had announced a repurchase at that point in time. We carry out this process for each firm in the repurchase sample, ending up with a pseudoportfolio consisting of a set of randomly drawn firms, matched in size, book‐to‐market, and time to the firms in the sample. We repeat this process until we have 1,000 pseudoportfolios and, thus, 1,000 abnormal return observations. This gives us an empirical distribution for the abnormal returns drawn under the null model specific to our hypotheses. The significance levels that we report represent the probability that a randomly drawn portfolio from the empirical distribution will have an abnormal return greater than the sample.