|
|
Outing insiders Last year set a record for mergers and acquisitions activity in Canada, with close to 2,000 takeover announcements valued at more than $250 billion Canadian dollars in 2006. This volume tops the recent peak in 2000 and represents a doubling of activity over the past two years. With all this takeover activity, it is not surprising that investors and regulators are concerned about the possibility of illegal insider trading. In August 2006 a study of U.S. takeovers found that 41% of the target companies experienced suspicious trading in the days and weeks before the deals became public.1While the source of these price and volume dynamics is not known, the suspicion is that illegal trading by insiders may be responsible. The alternative explanation is that sophisticated investors, such as merger arbitrageurs, are able to anticipate which companies will be the targets of a takeover, and this anticipation gets reflected in the target company’s share price. This paper examines these competing explanations using a sample of 420 Canadian takeovers that took place from 1985 to 2002. Existing studies of takeovers in Canada and abroad consistently document a run-up in the target firm’s shares before the takeover bid is made public (pre-bid run-up). We document the size and extent of these pre-bid runups for Canadian targets and describe the price and volume dynamics of the target firm’s shares ahead of the first public announcement. Illegal insider trading is an important policy issue facing Canadian capital markets. Insider trading is defined as trading by managers and board members in the stock of their own firms. Over the 1990s, regulators in 53 countries adopted securities laws restricting when and how corporate insiders may trade in a firm’s shares (Bhattacharya and Daouk 2002). These laws make it illegal for insiders to trade while in possession of material, non-public information, or to share this information selectively with other investors. Instead companies are required to disclose material information through a press release so that all investors have an equal opportunity to trade on this information. Illegal insider trading may undermine investor confidence, increase the rate of return demanded by less-informed investors, reduce liquidity in secondary markets, raise the cost of capital for firms, and ultimately hurt public welfare by reducing economic growth. Enforcement lacking Pre-bid run-ups may be caused by market anticipation, illegal insider trading, or some combination of both (Jabbour et al 2000). The market anticipation hypothesis argues that investors anticipate a takeover bid for a company based on rumours in the press, an analysis of industry trends, or factors specific to a company, such as financial distress. This market anticipation—whether accurate or not—becomes incorporated into prices through trades, leading to a run-up ahead of the takeover announcement. The alternative hypothesis is that pre-bid run-ups are due to information leakage associated with illegal insider trading. In this scenario, the increase in the stock price ahead of the bid announcement is caused by insiders—either the firm’s managers, its controlling shareholders, or its advisers—who trade illegally to profit from the price jump when the takeover is announced. U.S. studies of prosecuted cases of illegal insider trading support this view. These studies document that illegal insider trades are accompanied by abnormal returns and abnormal trading volume on the same day (Cornell and Sirri 1992; Meulbroek 1992). Court evidence further suggests that illegal insider trading typically takes place far ahead of the announcement, as insiders seek to avoid the period shortly before the announcement when regulatory scrutiny is highest. We examine 420 takeover bids of publicly listed Canadian firms from 1985 to 2002. We describe and quantify the extent of pre-bid run-ups, and benchmark the Canadian results against results from studies of U.S. takeovers. We propose a test to differentiate between competing explanations of run-ups based on the coincidence of abnormal returns and abnormal volume, and the timing of the pre-bid runup relative to the first public announcement. On average we find that the price and volume dynamics are more consistent with market anticipation for the average takeover target in our sample, although we cannot dismiss the possibility of illegal insider trading in any individual deals that we study. We also find that the timing and magnitude of pre-bid runups are similar in Canada and the United States, suggesting that markets are behaving in the same way on both sides of the border. Sample and methodology We used news searches of major newspapers and newswires to identify the “news-adjusted” announcement date, which represents the first date at which reliable information about the takeover bid was made public. About one-quarter of the bids have an announcement date prior to the formal takeover bid disclosed in a press release. We also identified media stories about rumoured takeover deals, where rumours are stories that name the target but do not provide any specific information about the acquirer or the terms of the transaction. On average about one in seven takeovers are rumoured in the press prior to the announcement, with the number increasing noticeably through the late 1990s, reaching a peak of one in four deals in 1999. The average market capitalization of a takeover target in our sample is $537 million, with a median value of $92 million, and a standard deviation of $1,589 million. The average share price is $9.60, with a median price of $5.40. Only 12% of the firms in our sample are penny stocks, with an average closing share price of less than $1. Our results are robust if we exclude these firms. In 77 out of the 420 takeovers, the acquiring firm purchased a toehold in the target firm’s shares. In these cases, the mean (median) toehold in the target firm’s shares was 22% (15%). To differentiate between the causes of pre-bid runups, we examine trading patterns ahead of the announcement consistent with the approach used by regulators when monitoring markets around major corporate events. We outline what price and volume dynamics we would expect to find if the pre-bid runup is due to market anticipation. Our expectations are based on a review of theoretical models of informed trading, and empirical studies of trading volume ahead of scheduled and unscheduled events. These studies find that trading volume increases ahead of earnings announcements, dividend changes, or takeovers, without generating a price reaction due to the contrasting expectations of investors. By comparison, studies of prosecuted U.S. cases of illegal insider trading consistently find illegal insider trades are associated with abnormal returns and abnormal volume on the days when insiders trade. Provided that the Canadian and the U.S. markets have similar market structures and institutions, illegal insider trades may be expected to generate the same effects in Canada as they do in the United States. These characteristics may allow us to differentiate between market anticipation and illegal insider trading. We conduct a standard event study of abnormal returns and trading volume in the target firm’s shares ahead of the first announcement of a takeover. The aim is to determine how the event affected the stock by comparing actual stock price movements to what changes might have been expected if the event had not taken place. For each takeover in our sample, we set the date of the first public announcement of the takeover as day 0. We then study trading over the 250 trading days prior to and the 60 trading days following the takeover announcement, denoted [-250,60]. For each day, we calculate a daily abnormal return (AR), which represents the portion of the daily stock return that is not related to movements in either the S&P/TSX Composite Index or the industry sub-index to which a company belongs. While any firm may exhibit abnormal returns on a given day as news reaches the market, these abnormal returns should follow a random walk and exhibit both positive and negative returns. Abnormal returns for a given bid are aggregated and averaged across the 420 takeovers in our sample to generate an average abnormal return (AAR). By summing these average abnormal returns over different windows, we calculate a cumulative average abnormal return (CAAR) that can be used to identify whether this group of takeover bids systematically outperformed or underperformed the stock market. We test to see whether these average and cumulative average abnormal returns are statistically different from zero using both the average and the median of the sample distribution. We conduct a similar analysis of trading volume using average abnormal share turnover for each of the 420 takeover announcements in our sample. Magnitude of run-ups
As a sensitivity analysis, we divide our sample into various sub-samples to investigate the impact of industry membership and the time period when the takeover bid occurred. Previous studies suggest that a clustering of takeovers in one sector or during one time period increases the ability of the market to anticipate future potential takeovers. Given the high number of takeovers in the natural resource sector, and a clustering of bids over a few key years, we expect that takeover bids in this sector might be easier for the market to anticipate. If this were the case, we would expect to find that the CAARs for natural resource takeovers should be higher than for nonresource bids that are more heterogeneous. Table 1 shows the results with oil and gas, metals, mining and forestry takeovers in column (2) and manufacturing and service sector bids in column (3). Contrary to our expectations, the run-ups for natural resource firms are much smaller—almost half the comparable run-ups for non-resource firms.
We also consider the impact of institutional changes on pre-bid run-ups. If illegal insider trading is the source of pre-bid run-ups, increased resources devoted to supervision and enforcement, together with advances in technology, should discourage this behaviour by making it easier to detect illegal trading activity ex-post. The resources devoted to monitoring and enforcement increased significantly in 1998 after the Ontario Securities Commission became self-funded. At the same time, the TSX closed its trading floor and moved all stocks to an electronic trading system. Both changes lead us to expect that pre-bid price run-ups may be smaller post-1997 than during the earlier period. Table 1 splits the sample between deals announced from 1985 to 1997 in column (4) and deals announced from 1998 to 2002 in column (5). We find that both the pre-bid price run-ups and the price jump over the event window were larger for takeovers announced after 1997. This finding, together with the finding that more media rumours are observed over this period, suggests that market anticipation has increased—possibly due to improvements in market transparency and information dissemination. Canada versus U.S. Conclusion Based on the pattern of run-ups, the absence of abnormal volume on days with abnormal returns, and the timing of the run-up shortly before the announcement date, we conclude that pre-bid runups for the average Canadian takeover in our sample are consistent with market anticipation, and reject an explanation based on illegal insider trading. While this conclusion applies to the average takeover bid in our sample, we cannot dismiss the possibility of illegal insider trading in any of the individual takeovers in our sample. Our results depend on the assumption that illegal insider trading in Canada follows patterns that have been documented using prosecuted cases of illegal insider trading from the United States. We have not studied prosecuted Canadian cases, but plan to look at this question in future research. If the Canadian institutional setting or behaviour of insiders in Canada is very different than from the United States, then the most our results can show is that price and volume dynamics ahead of an announcement are not a reliable mechanism for identifying illegal insider trading. As one knowledgeable Canadian institutional investor wrote us, “To believe that the run-up of stocks two to three weeks before an announcement is just the result of intelligent investors, private or institutional, strikes me as naïve. Sometimes it may possibly occur, but I would rather call that an exception”. This scepticism may be warranted given the enforcement record in Canada, but this research has not found concrete evidence to support it. Endnotes References —Michael R. King, research adviser, International Department,
at the Bank of Canada, and Maksym Padalko, For a PDF version of this article, click here.
|
|
Table
1 provides results for different event windows and different groupings
of firms. Column (1) covers the whole sample, and shows that the
average takeover has an abnormal return of 9.76% on the day of the
announcement, with an abnormal return of approximately 2% on the
day before and the day after the announcement. When we accumulate
abnormal returns over the window [-60,20], we find that an investor
with perfect knowledge could have outperformed the market by 15.90%.
About one-third of this pre-bid run-up occurs in the final two weeks
before the announcement, as shown in the graph in Figure 1. The
final row in Table 2 (see below) presents the “runup index,”
which measures the size of the run-up prior to the announcement
to the period including the announcement. The run-up index confirms
that about one-third of the pre-bid run-up occurs prior to the takeover
announcement.