UK managed funds trading around M&A announcements

Transcription

1 UK managed funds trading around M&A announcements By Raymond da Silva Rosa* Minh Huong To** & Terry Walter*** Abstract We test UK fund managers stock selection ability by investigating if they revise their daily holdings of firms involved in mergers & acquisitions (M&A) announcement to take advantage of the abnormal market performance of these firms. We find that fund managers are net buyers rather than sellers of firms involved in takeover activity, consistent with theories that they trade principally on perceived good news. However, there is no evidence that they have a comparative advantage in identifying value-increasing takeovers, nor do they take advantage of the systematic association of method of payment with abnormal return. Consistent with other studies, we find that returns to firms that made or received cash offers were higher than those to firms that offered equity. There is no evidence that UK fund managers exploit this association. * WA Centre for Capital Markets Research, UWA Business School, and SIRCA Limited. Ray.DaSilvaRosa@uwa.edu.au ** UWA Business School, University of Western Australia. HuongMinh_To@yahoo.com *** School of Banking and Finance, University of New South Wales and CMCRC Ltd. T.Walter@unsw.edu.au 1

2 1. Introduction Individuals invest via professionally managed funds partially for convenience and to minimize transaction costs (Edelen 1999) but also because fund managers claim to be able to deliver superior risk-adjusted returns. We test this claim by investigating if UK managed funds revise their daily holdings of firms involved in mergers & acquisitions (M&A) announcement to take advantage of the abnormal market performance of these firms. We focus on M&A announcements because they typically trigger substantial abnormal returns which, around the days surrounding announcements, are largely robust to model of risk-adjustment (Brown and Warner 1985) and so our test is relatively immune from the joint-hypothesis problem identified by Fama (1970) which is that assessment of abnormal returns rests on the assumption that risk has been appropriately controlled. Further, an impressive body of research reports that method of payment (e.g., cash or equity) in M&As systematically predicts long-term abnormal performance. Acquiring firms that offer equity as consideration subsequently under-perform, an association consistent with these firms being over-valued at the time they make their takeover offer (Shleifer and Vishny 2003; Rhodes- Kropf and Viswanathan 2003). Not all firms that offer equity display subsequently underperformance. We investigate whether fund managers are able to distinguish between overand under-valued firms after controlling for method of payment. Our study contributes to the literature as follows. Following Jensen (1968), most studies focusing on whole fund performance report that net returns to actively managed funds do not outperform passive benchmarks. A relatively recent research method innovation is to focus on the performance of individual stocks held in fund portfolios. Studies adopting this approach, which include Daniel, Grinblatt, Titman, and Wermers (1997), Chen, Jegadeesh, and Wermers (2000), and Wermers (2000), report results consistent with fund managers having the ability to choose stocks that outperform their benchmark. However, as Wermers (2000) and Pinnuck (2003) discuss, after transaction costs, management fees and poor market timing decisions are taken into account, net returns to unit holders do not outperform the benchmark. This result is consistent with economically sensible definitions of market efficiency but such tests are arguably biased against finding evidence of superior stock selection skills amongst fund managers. As Edelen (1999) observes, fund managers provide a great deal of liquidity to investors and thus engage in a material volume of uninformed, liquidity-motivated trading in an asymmetrically informed market with costly information production, equilibrium is attained only when liquidity-motivated traders sustain losses to informed investors (p. 440). One expects fund managers skill to be revealed most clearly in situations where the potential to earn profits is greatest and liquidity motivated trading is less pronounced. M&A announcements, which trigger substantial abnormal returns and the incidence of which is likely uncorrelated with fund managers need to trade for liquidity reasons, provide an appropriate setting to test fund managers stock selection skills. The strong empirical association of method of payment with abnormal returns also facilitates assessment of their ability to accurately interpret salient predictors of performance. Our results indicate that UK fund managers do not profit from any superior knowledge they may have. Overall, they are net buyers of equity in acquiring and target firms around the bid 2

3 announcement month but their trading is largely indiscriminate. Our sample of UK funds increased their net holdings in bidders offering cash, shares, or some mixture of the two in about equal proportions despite the method of payment being predictive of future abnormal returns, a finding borne out in our results; acquirers who offered cash as consideration performed best. As it happens, cash offers predominated during our sample period so the funds earned positive abnormal returns from their trading in firms involved in M&A activity. Shleifer and Vishny (2003) and Rhodes-Kropf and Viswanathan (2003) develop separate theories that propose market misvaluation drives much takeover activity. Our findings suggest that fund managers are not more expert than the typical investor in identifying mergers prompted by misvaluation. 2. Data and research method 2.1. Citywatch data Our Citywatch dataset contains the daily holding positions in equity stocks of all UK managed funds for the period from 21 September 2001 to 18 September For this period, we count in our data 2,581 different stocks, 4,264 different funds and over 78 million stock holding positions on the day that individual funds had a net increase or decrease in their holding of a given stock. The Citywatch dataset does not differentiate between, or otherwise identify, active and passive funds. Changes in the daily holdings of passive index funds presumably reflect attempts to align the fund with its benchmark index rather than being indicative of information-based trading and so passive funds should be omitted from the analysis. The distribution of funds number of trades shows that over 75% of the funds have less than 30 trades during the three years September 2001 through to September We exclude these funds so our study is based on investigation of 1,011 funds that engaged in at least 30 trades over the three-year period September 2001 through to September UK bidders data We review the International Mergers database to identify 1,297 bids by listed UK companies where deal value was equal to or greater than Euro 5 millions over the period 3 Sep 2001 to 31 Dec Of the 1,297 bids, 880 [68%] included just cash as consideration, 99 (8%) offered shares, 180 (14%) bids offered a combination of cash and shares and 138 (11%) bids involved combinations of cash, shares, loan notes, profit-contingent payments and options. Fund holdings of UK listed firms are identified by EPIC stock codes and we identify the EPIC codes of 1,179 bidding companies, i.e., 90.9% of the original sample of 1,297 UK bidders. Matching bidding companies with the Citywatch stock holding data of 1,011 funds by epic codes yields us 3,538,419 buy, sell and hold observations over the event-window, [- 30,+30] trading days relative to the bid announcement date. There are 1,709,154 observations prior to the announcement dates, and 1,829,265 on and after the announcement dates UK targets data 3

4 From the International Mergers database, we identified 418 UK listed targets of takeover bids made over the period 3 September 2001 to 31 December 2004 where the bid value was equal to or greater than Euro 5 millions. Of the 418 target firms, 274 (65%) received an offer of cash as consideration, 43 (1) received an offer of just shares, 30 (7%) targets were offered a combination of cash and shares and 71 (17%) firms were offered a combination of cash, shares, loan notes, profit-contingent payment and options. We identify the EPIC codes of 353 of the listed target firms, i.e., 84.4% of the original 418 UK targets, were found. When EPIC code data are matched with the Citywatch stock holding data of our sample 1,011 funds we find 756,570 buy, sell and hold observations in the window 30 days prior to 30 days after the announcement dates. There are 372,067 observations prior to the announcement dates and 384,503 observations on and after the announcement dates. 2.4 Event windows and trade classification We focus on the trading decisions of fund managers over two event periods, expressed relative to each bid announcement date: [-30,-1] tradings days and [0,+30] trading days. [i] the pre-merger period of 30 days before the merger announcement date [-30, -1]; and [ii] post-merger period, which covers from the announcement date to 30 days after the announcement date [0, +30]. Buy, sell and hold decisions of each fund during the two periods 30 days prior to and 30 days after the merger announcement dates are aggregated, resulting in 105,771 and 108,028 observations respectively. We examine the trading behaviour of fund managers based on the buying and selling data derived from the stock holding positions. To differentiate between liquidity and informationbased trading we calculate the relative size of trading transaction, defined as the proportion of net buy or sell over the original number of shares held at the beginning of the studied period. We assume that the size of trade is a relative measure of the information content of the trade, basing our assumption on standard models of informed trading [e.g., Kyle 1985] that show that the more information investors have on an event and the more confident they are about the direction of future share price changes, the larger the trades they engage in. For each bidder and target, the relative trade size engaged by fund i is calculated as follows: Where: i Trade t = [Buy i t -Sell i i t ]/Beginning Number of Shares t-1 i Trade t is the relative size of net buys and sells engaged by fund i during the respective windows, [-30, -1] and [0, +30] trading days relative to the bid announcement date; and i Buy t is the number of shares that fund i bought during the respective event windows, [-30, -1] and [0, +30] trading days relative to the bid announcement date; and 4

5 i Sell t is the number of shares that fund i sold during the respective event windows, [- 30, -1] and [0, +30] trading days relative to the bid announcement date; and i Beginning Number of Shares t is the number of shares of a particular stock that fund i held prior to the studied period. For the window [-30,-1] trading days, the beginning number of shares is the number of shares that the fund on day [-31]. For window [0,+30] trading days, the beginning number of shares is the number of shares that the fund held on day [-1]. The relative size of trade, Trade ti, is sorted into seven categories: : i Trade t < -3 : i -1 < Trade t -3 : i 0< Trade t -1 No Trading: i Trade t = 0 : i 0 < Trade t < 1 : i 1 Trade t < 3 : i Trade t 3 3. Results and discussion 3.1 Trading by fund managers around M&A bid announcements Research shows that M&As generally increase bidder and (more dramatically) target firms shareholder value in the short-run around the announcement date. One implication is that informed fund managers engage in heavier bidder and target stock buying, especially target stock buying, around the merger announcement. Another reason to expect heavier buying by fund managers is that they generally hold long positions and so it is optimal for them to predominantly engage in searching for good news or stocks whose price is expected to rise (Saar 2001; Pinnuck 2003; Chan and Lakonishok 1993; and Keim and Madhavan 1995). Further, we can expect that net buying is associated with greater abnormal returns than net selling. This is because sales can be motivated by either bad news or to achieve liquidity and, if managers are selling for liquidity reasons they are restricted to the stocks they already own. In contrast, acting on good news does not depend on the relevant stock being in one s portfolio so buying is more reliably motivated by positive information. Another reason to expect buys to be more strongly associated with abnormal return is that institutional investors are more likely to be made aware of good rather than bad news. Fund managers often source information from analysts, who have greater incentives to issue buy recommendations than sell because the former generate greater trading volume and for fear of losing access to management as a source of information (McNichols and O Brien, 1997). In support, Chen, Jegadeesh and Wermers (2000) find aggregated buys but not sells realize abnormal returns. Chan and Lakonishok (1993) also find that buys but not sells have a permanent price impact in the context of institutional block trades and interpret this as consistent with traders having good but not bad news. At the market level, Hong, Lim, and Stein (1999) provide evidence that bad news is incorporated into prices more slowly than good news. They observe that this is consistent with economic agents such as fund managers gathering good but not bad news. 5

6 3.2 Trading by fund managers in bidding firms around the bid announcement period Table 1 and Figure 1 report information on trading by managed funds in bidding firms shares prior to the bid announcement date, classified by method of payment and magnitude of net change in holding position. The pre-bid period is defined as [-30,-1] trading days prior to the bid announcement date and each fund s holding as at [-31] trading day is used as a benchmark to calculate net change in holding. Table 1 and Figure 1 show that there are more pre-announcement trades of bidder stock in a cash merger compared to share, hybrid or other merger methods. In addition, regardless of method of payment, UK managed funds engage more in heavy buying than heavy selling, with over 1 of all holdings exhibiting a net increase over the pre-bid period of greater than or equal to 3 of the fund s original holding on day [-31] while between just 3.5%- 5.7% of all fund holdings experience a net decrease greater than or equal to 3 over the same period. Chi-square tests confirm these differences are statistically significant. Table 2 and Figure 2 report information on trading by managed funds in bidding firms shares after the bid announcement date. As before, trading intensity is summarized by method of payment and magnitude of net change in holding position. The post-bid period is defined as [0,+30] trading days prior to the bid announcement date and each fund s holding as at [-1] trading day is used as a benchmark to calculate net change in holding. In line with results for the pre-merger announcement period, UK fund managers engage more in heavy purchases than heavy selling over the immediate post-announcement period, a conclusion confirmed by Chi-square tests. The above results are for all bidding companies regardless of the merger outcome status. We run the same analysis for bidding companies involved in completed mergers only and obtain similar results. 3.3 Trading by fund managers in target firms around the bid announcement period Table 3 and Figure 3 report information on trading by managed funds in target firms shares prior to the bid announcement date. Trading intensity is summarized by method of payment and magnitude of net change in holding position. As for bidding firms, the pre-bid period is defined as [-30,-1] trading days prior to the bid announcement date and each fund s holding as at [-31] trading day is used as a benchmark to calculate net change in holding. The results are similar to those for bidding companies: prior to the bid announcement, there are more heavy purchases than heavy sales of target firms shares. Also, fund managers tend to engage in more heavy buying of target stocks that were offered cash compared to those offered shares. Chi-square test results indicate that the number of target stock purchases undertaken by UK fund managers is statistically significantly higher than the number of target stock sales. Table 4 and Figure 4 report findings that, again, are consistent with those found for acquiring firms. In the post-bid announcement period, UK managed funds made more heavy purchases than heavy sales. Chi square test confirm that both before and after M&A 6

7 announcements, UK managed funds purchases of target shares purchases are statistically significantly higher than their sales of bidding firms shares. 3.4 The association between net change in fund holdings and M&A firms performance Standard models of informed trade show that, ceteris paribus, there is a positive relationship between trade size and abnormal returns. We use net change in fund holdings as a proxy for trade size and investigate if the magnitude of change predicts abnormal return. We use a market adjusted returns model in light of Brown and Warner s finding that measures of daily abnormal performance are largely insensitive to method of adjustment for risk once marketwide return is controlled. Tables 5 and 6 show the market-adjusted return to individual bidder stock trades in the funds portfolio for three post merger announcement windows [0,+30], [+31,+60], and [+61,+90] classified by four payment methods. Table 5 shows that fund managers heavy buying of bidder stocks in the pre-bid announcement period is associated with subsequent significant positive abnormal returns across the window [0,+30] trading days for bids offering cash, share and other as consideration. However, funds that did not change their holdings in cash bidders earned superior returns than funds that increased their net holdings by 3 or more. In the hybrid payment category, the stocks that fund managers purchased heavily during the window [- 30,0] generated negative returns post-announcement. This suggests fund managers might have engaged in some informed trade prior to the announcement, but that does not consistently result in positive abnormal returns. Table 6 indicate fund managers post-announcement heavy buying of bidder stocks generates significant positive returns during the window [0,+90] only for acquisitions where cash is offered as consideration. For cash mergers, the returns from heavy buying outperform the returns from other trade size categories. For share, hybrid and other payment methods, the stocks that fund managers purchased heavily during the window [0,+30] generate negative returns after the announcement. This suggests that the level of informed trading by fund managers varies across different payment methods and fund managers appear to have more information, or have performed best, on cash mergers rather than hybrid, share and other mergers. On the other hand, the negative post announcement returns is consistent with the literature that the share price of bidder stocks declines after merger announcement. Another possible explanation is that information on bidder stocks has been reflected in bidder stock price prior to share, hybrid, and other merger announcement. Therefore, fund managers should sell off their position of bidder stocks for mergers not involved with purely cash payment soon after the announcement date. Figure 5 presents the Cumulative Market-Adjusted Returns or Cumulative Abnormal Returns [CARs] weighted by individual bidder stock trades in the funds portfolio. The reported CARs are classified by four payment methods and seven trade sizes of the window [-30,0] and [0,+30], respectively. 7

8 Figure 5 shows that fund managers buying, particularly heavy buying, prior to announcement, does not generate highest CARs. In fact, no trading category results in a [0,+90] window s CAR of 4% for cash mergers. Regarding post announcement, cash mergers bidder stocks which were heavily bought by fund managers significantly outperform the market. The CAR of cash mergers heavily bought bidder stocks for the window [0,+90] is over 25% compared with those of other trade sizes of between -7% to 5%. It is interesting to note that for no trading category which accounts for over half of the observations, CARs are relatively quite stable for all payment methods and in the range of 3% to 7% for the window [0,+90]. For other trade size categories, post announcement performance for the window [0,+90] of bidder stocks is mostly negative. Tables 7 and 8 report the trading portfolio-weighted Market-Adjusted Return of individual target stock trades in the funds portfolio three post merger announcement windows [0,+30], [+31,+60], and [+61,+90] classified by cash, hybrid, and other payment methods, and seven trade sizes of the window [-30,0] and [0,+30], respectively. The weighted Market- Adjusted Return for share payment methods is only reported for the no trading category, because of the insufficient number of targets with available stock price data in other trade categories. The positive Market-Adjusted Returns for the window [0,+30] from Tables 7 and 8 indicate that on average fund managers earn substantial positive returns by buying and holding, or just by holding on to target stocks for the period one month from the announcement. Fund managers pre-announcement heavy buying of bidder stocks only outperforms other trade size categories for cash mergers. This suggests fund managers might have preannouncement information advantage or have predicted most accurately the price trend in the case of cash mergers. Also, consistent with the literature, the size of target stocks abnormal returns are substantially higher than the corresponding of bidder stocks, especially for the interval [0,+30]. Table 8 shows that the post-announcement heavy buying category of cash and mergers target stocks generates significant positive abnormal return for the window [0,+30]. After the day +30, those target stocks no longer outperform the market. On the other hand, the target stocks of other payment method that fund managers purchased heavily on average generate negative abnormal return for the first 60 days after merger announcement. However, for the interval [+61,+90], the price of those stocks bounced back, generating an average abnormal return of 13.36%. Figure 6 presents CARs weighted by individual target stock trades in the funds portfolio. The reported CARs are classified by three payment methods cash, hybrid, and other - and seven trade sizes of the window [-30,0] and [0,+30], respectively. Figure 6 shows that share price of target stocks tends to pick during the period 30 days after merger announcement. Subsequently, target stock price either remains at the similar level, or 8

9 reduce slightly in the case of cash or hybrid mergers, or reduces substantially in the case of other mergers. Fund managers buying, particularly heavy buying, prior to announcement, generates the highest CARs for cash mergers, but underperforms in the case of hybrid and other mergers. Regarding post announcement, cash mergers bidder stocks which were heavily bought by fund managers only perform well for the first 30 days after the merger announcement. Meanwhile, heavy buying of hybrid mergers target stocks appears to generate a stable average abnormal return of 18%. Overall, the results indicate that UK fund managers seem to be able to predict the stock price direction of bidder and target stocks for cash mergers. For mergers with share, hybrid, and other payment methods, the results are mixed. 4.0 Concluding comments This paper investigates whether UK fund managers possess superior information in relation to target and bidder stocks around merger announcement. We approach this through analyzing the bidder and target stock trading patterns and their abnormal returns. There are two main findings. First, as expected, fund managers engage significantly in more buying then selling, which indicates that fund managers tend act on good rather than bad news. Second, UK fund managers appear to possess information on cash mergers where the stocks that they bought heavily generally generate positive abnormal returns. For mergers with hybrid, share and other payment methods, the results are mixed, suggesting that it might be more difficult for fund managers to identify the implications of merger events on bidder and stock share price for hybrid, share and other mergers. This is consistent with the existing literature abnormal returns from share mergers are positive, while those of share and mixed mergers are inconclusive. REFERENCES 1. Agrawal, A., Jaffe, J., and G., Mandelker [1992]. The post-merger performance of acquiring firms: a re-examination of an anomaly. Journal of Finance. 47[4] Andrade, G., Mitchell, M., and E. Stafford [2001]. New evidence and perspectives on mergers? Journal of Economic Perspectives Asquith, P. [1983]. Merger bids, uncertainty, and stockholder returns. Journal of Financial Economics Chan, L. K., and J. Lakonishok [1993]. Institutional trades and intraday stock price behavior. Journal of Financial Economics Chen, H., Jegadeesh, N., and R. Wermers [2000]. The value of active fund management: an examination of the stockholdings and trades of fund managers. Journal of Financial and Quantitative Analysis Chen, H., Jegadeesh, N., and R. Wermers [2000]. The value of active fund management: an examination of the stockholdings and trades of fund managers. Journal of Financial and Quantitative Analysis

10 7. Cici, G. [2005]. The impact of the disposition effect on the performance of mutual funds. Working Paper. Wharton Research Data Services. 8. Daniel, K., Grinblatt, M., Titman, S. and R. Wermers [1997]. Measuring mutual fund performance with characteristic-based benchmarks. Journal of Finance. 52[3] Draper, P., and K. Paudyal [1999]. Corporate takeovers: Mode of payment, returns and trading activity. Journal of Business Finance and Accounting. 26[5] and [6] Emery, G. W., and J. A. Switzer [1999]. Expected market reaction and the choice of method of payment for acquisitions. Financial Management. 28[4] Franks, J., Harris, R., and S. Titman [1991]. The post-merger share price performance of acquiring firms. Journal of Financial Economics Hong, H., Lim, T., and J. Stein [2000]. Bad news travels slowly: size, analysts coverage, and the profitability of momentum strategies. Journal of Finance Huang, Y., and R. A. Walkling [1987]. Target abnormal returns associated with acquisition announcements. Journal of Financial Economics Jensen, M. [1968]. The performance of mutual funds in the period Journal of Finance Jensen, M. and R. S., Ruback [1983]. The market for corporate control: The scientific evidence. Journal of Financial Economics Jensen, M. C. [1993]. The modern industrial revolution, exit, and the failure of internal control systems. Journal of Finance. 48[3] Jensen, M. C. [2000]. A Theory of the Firm: Governance, Residual Claims, and Organizational Forms. 1 st edition. Harvard University Press. Cambridge. Massachusetts. 18. Keim, D. B., and A., Madhavan [1995]. Anatomy of the trading process: empirical evidence on the behavior of institutional trades. Journal of Financial Economics Leeth, J. D., and J. R., Borg [2000]. The impact of takeovers on shareholder wealth during the 1920s merger wave. Journal of Financial and Quantitative Analysis. 35[2] Manne, H. [1965]. Mergers and the market for corporate control. Journal of Political Economy McNichols, M., and P.C., O Brien [1997]. Self-selection and analysts coverage. Journal of Accounting Research Myers, S.C. [1984]. The capital structure puzzle. Journal of Finance. 39[3] Myers, S.C., and N.S., Majluf [1984]. Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics Nelson, R. [1959]. Merger movements in the American industry. NBER. New York. 25. Pinnuck, M. [2003]. An examination of the performance of the trades and stock holdings of fund managers: further evidence. Journal of Financial and Quantitative Analysis. 38[4], Rappaport, A. and M.L., Sirower [1999]. The trade-offs of buyers and sellers in mergers and acquisitions. Harvard Business Review. Nov-Dec Rau, P. and T., Vermaelen [1998]. Glamour, value and the post-acquisition performance of acquiring firms. Journal of Financial Economics Roll, R. [1986]. The hubris hypothesis of corporate takeovers. Journal of Business. 59[2]

11 29. Saar, G. [2001]. Price impact asymmetry of block trades: an institutional trading explanation. Review of Financial Studies Wermers, R. [2000]. Mutual fund performance: an empirical decomposition into stock-picking talent, style, transaction costs and expenses. Journal of Finance

12 Table 1 changes in the holdings of managed funds in shares of bidding firms over the pre-bid announcement event-window [-30,-1] trading days. changes are summarised separately by method of payment offered in bid. Each fund s holding is measured at trading day [-31] relative to the bid announcement date. changes in each fund s holding over the pre-bid period are standardized as follows: Trade t i = (Buy t i -Sell t i )/Beginning Number of Shares t-1 i where beginning number of shares is each fund s holding in the bidding firm as at trading day [-31]. decreases to 3 of the fund s holding on day [-31] are classified as heavy net selling, net decreases 1 but < 3 are classified as moderate net selling, while net increases in holdings 1 but < 3 are classified as moderate net buying. net buying includes any net increase in holding 3 of the fund s holding on day [-31]. The sample data are based on 1,011 UK managed funds respective holdings in 1,179 UK listed companies that made a bid worth at least five million euros over the period September 2001 to September Panel A Type of Payment Offered in Bid Total Number of Holdings No Trading Cash 86,364 (82%) 4.5% 4.7% 12.4% 52.5% 11.7% 4.1% 10.1% Hybrid 7,955 (8%) 4.1% % % 3.5% 12.4% Other 7,169 (7%) 3.5% 3.4% 7.2% 63.8% 6.8% 3.7% 11.6% Shares 4,283 (4%) 5.7% % % 13. Panel B No of Sells No of Buys Total Number of Trades Percentage of Sells Percentage of Buys Cash 18,649 22,347 40, % 54.51% Hybrid 1,330 1,852 3, Other 1,005 1,587 2, % 61.23% Shares 718 1,007 1, % 58.38% Total 21,702 26,793 48, % 55.25% 12

13 Table 2 changes in the holdings of managed funds in shares of bidding firms over the postbid announcement event-window [0,+30] trading days. changes are summarised separately by method of payment offered in bid. Each fund s holding is measured at trading day [-1] relative to the bid announcement date. changes in each fund s holding over the pre-bid period are standardized as follows: Trade t i = (Buy t i -Sell t i )/Beginning Number of Shares t-1 i where beginning number of shares is each fund s holding in the bidding firm as at trading day [-1]. decreases to 3 of the fund s holding on day [-1] are classified as heavy net selling, net decreases 1 but < 3 are classified as moderate net selling, while net increases in holdings 1 but < 3 are classified as moderate net buying. net buying includes any net increase in holding 3 of the fund s holding on day [-1]. The sample data are based on 1,011 UK managed funds respective holdings in 1,179 UK listed companies that made a bid worth at least five million euros over the period September 2001 to September Panel A Type of Payment Offered in Bid Total Number of Holdings No Trading Cash 87, % 5.2% 13.1% 49.2% 11.8% 4.5% 11.3% Hybrid 7, % 4.6% 8.2% % 4.1% 12.1% Other 8, % 4.4% 9.2% 51.5% 8.1% 4.1% 18.3% Shares 4, % 7.6% % 3.1% 10.5% Panel B No of Sells No of Buys Total Number of Trades Percentage of Sells Percentage of Buys Cash 20,249 24,282 44, % 51.33% Hybrid 1,438 1,880 3, % 51.33% Other 1,483 2,543 4, % 57.34% Shares , % 51.16% Total 23,870 29,528 53, % 51.74% 13

14 Table 3 changes in the holdings of managed funds in shares of target firms over the pre-bid announcement event-window [-30,-1] trading days. Each fund s holding is measured at trading day [-31] relative to the bid announcement date. changes in each fund s holding over the pre-bid period are standardized as follows: Trade t i = (Buy t i -Sell t i )/Beginning Number of Shares t-1 i where beginning number of shares is each fund s holding in the bidding firm as at trading day [-31]. decreases to 3 of the fund s holding on day [-31] are classified as heavy net selling, net decreases 1 but < 3 are classified as moderate net selling, while net increases in holdings 1 but < 3 are classified as moderate net buying. net buying includes any net increase in holding 3 of the fund s holding on day [-31]. The sample data are based on 1,011 UK managed funds respective holdings in 353 UK listed companies that received a takeover bid worth at least five million euros over the period September 2001 to September Type of Payment Offered in Bid Total Number of Holdings No Trading Panel A Completed, withdrawn and pending mergers Cash 13, % 3.9% % 6.2% % Hybrid 1, % 4.5% 12.5% 52.5% 12.2% 4.8% 8.7% Other 4, % 4.4% 7.5% 65.4% 6.3% % Shares 1, % 4.8% 6.2% 68.1% 5.4% 4.1% 6.9% Panel B Completed mergers only Cash 7, % 4.2% 7.4% 62.6% 5.6% 2.7% 13.1% Hybrid % 5.8% 13.3% 40.4% 13.3% 6.5% 13.5% Other 2, % 4.1% 7.9% 62.9% 6.3% 3.4% 9.9% Shares 1, % 6.1% 7.9% % 8.5% Panel C Method No of Sells No of Buys Total Number of Trades Percentage of Sells Percentage of Buys Cash 2,051 2,745 4, % 57.24% Hybrid % 54.02% Other , % 50.68% Shares Total 3,409 4,207 7, % 55.24% 14

15 Table 4 changes in the holdings of managed funds in shares of target firms over the post-bid announcement event-window [0,+30] trading days. Each fund s holding is measured at trading day [-1] relative to the bid announcement date. changes in each fund s holding over the pre-bid period are standardized as follows: Trade t i = (Buy t i -Sell t i )/Beginning Number of Shares t-1 i where beginning number of shares is each fund s holding in the bidding firm as at trading day [-1]. decreases to 3 of the fund s holding on day [-1] are classified as heavy net selling, net decreases 1 but < 3 are classified as moderate net selling, while net increases in holdings 1 but < 3 are classified as moderate net buying. net buying includes any net increase in holding 3 of the fund s holding on day [-1]. The sample data are based on 1,011 UK managed funds respective holdings in 353 UK listed companies that received a takeover bid worth at least five million euros over the period September 2001 to September Type of Payment Offered in Bid Total Number of Holdings No Trading Panel A Completed, withdrawn, pending mergers Cash 13, % 4.9% % 3.2% 10.4% Hybrid 1, % 3.8% 7.4% 67.1% 4.7% 3.8% 9. Other 3, % 4.2% 6.4% 63.9% 4.9% 2.6% 11.8% Shares 1, % 4.6% 7.3% 56.4% 6.1% 3.8% 15.4% Panel B Completed mergers only Cash 6, % 4.9% 8.4% 59.9% 6.3% 3.1% 11.2% Hybrid % 1.7% 89.4% 0.4% 1.7% 3.5% Other 2, % 4.1% 6.7% 63.5% 5.2% 2.8% 12.6% Shares 1, % 9.6% 45.6% 7.6% 4.5% 19.2% Panel C Method No of Sells No of Buys Total Number of Percentage of Percentage of Buys Trades Sells Cash 2,555 2,731 5, % 51.66% Hybrid % 53.26% Other , % 53.49% Shares % 58.13% Total 3,812 4,258 8, % 52.76% 15

16 Table 5 The association between pre-bid changes in funds holding of bidding firms shares and the subsequent performance of those shares over various post-bid periods The pre-bid period is defined as [-30,-1] trading days relative to the bid announcement date. Market adjusted returns are reviewed for three post-bid announcement periods: [0,+30], [+31,+60], & [+61,+90] trading days. decreases to 3 of the fund s holding on day [-31] are classified as heavy net selling, net decreases 1 but < 3 are classified as moderate net selling, while net increases in holdings 1 but < 3 are classified as moderate net buying. net buying includes any net increase in holding 3 of the fund s holding on day [-31]. t-statistics test the null hypothesis that the portfolio-weighted mean of market-adjusted returns equal zero. ***, * Statistical significance in 2-tailed tests at the 1% and 5% levels, respectively. Trading days relative to the bid announcement date No Trading Cash Payment Method [0,+30] -0.14% -0.81% -1.76% 1.63% -1.95% -0.97% 0.61% (-1.14) (-6.81) *** *** (-25.65) *** (-29.67) *** (-8.56) *** (6.90) *** (-27.51) [+31,+60] 1.71% -0.02% -0.22% 1.16% -1.28% 0.59% 0.13% (8.89) *** (-0.08) (-2.31) * (-15.66) *** (-13.33) *** (3.50) *** (1.04) [+61,+90] % -0.61% 1.17% -1.96% % (-3.11) *** (-1.41) (-4.20) *** (-12.02) (-12.20) *** (0.86) (-2.53) * Share Payment Method [0,+30] -2.27% -1.13% 3.15% -0.33% -7.93% 0.51% 5.09% (-1.76) (-1.01) (2.47) * (-0.66) (-9.85) *** (0.38) (6.92) *** [+31,+60] -4.45% -0.99% % % 1.64% (-2.47) * (0.57) (4.09) *** (0.70) *** (-9.78) *** (1.23) (1.91) [+61,+90] -4.45% -2.63% 6.31% -0.79% -8.46% -1.63% -2.58% (-2.67) * (-1.80) (3.80) *** (-1.30) *** (-7.79) *** (-0.80) (-2.83) *** Hybrid Payment Method [0,+30] % -4.94% -8.52% 0.53% -7.88% -7.09% -4.5 (-13.40) *** (-10.55) *** *** (2.61) *** (-17.11) *** (10.55) *** (-9.29) *** (-18.55) [+31,+60] -9.85% -4.83% -9.24% 1.55% -7.22% -7.85% -3.57% (-8.90) *** (-6.41) *** *** (5.31) *** (-11.03) *** (-7.94) *** (-3.52) *** (-14.35) [+61,+90] % -9.38% % 0.38% % % -5.54% (-10.38) *** (-9.96) *** *** (0.91) (-16.60) *** (-9.05) *** (-4.47) *** (-20.08) Other Payment Methods [0,+30] 1.03% % 0.94% 0.81% 0.58% 2.08% (2.32) * (1.63) (4.99) *** (-5.23) *** (2.85) *** (2.02) * (9.43) *** [+31,+60] 1.45% 1.39% 2.35% 0.46% 1.05% -0.04% 2.36% (1.86) (2.10) * (4.94) *** (1.69) (2.02) * (-0.07) (6.47) *** [+61,+90] -0.95% -1.58% -1.34% 0.97% -2.75% -3.26% -1.78% (-1.19) (-2.52) (-3.52) *** (2.90) *** (-5.50) *** (-5.94) *** (-4.55) *** 16

17 Table 6 The association between post-bid changes in funds holding of bidding firm s shares and the subsequent performance of those shares The post-bid period over which the net change in each funds holding is [0,+30] trading days relative to the bid announcement date. Market adjusted returns are reviewed for three periods: [0,+30], [+31,+60], & [+61,+90] trading days. decreases to 3 of the fund s holding on day [-1] are classified as heavy net selling, net decreases 1 but < 3 are classified as moderate net selling, while net increases in holdings 1 but < 3 are classified as moderate net buying. net buying includes any net increase in holding 3 of the fund s holding on day [-1]. t-statistics test the null hypothesis that the portfolio-weighted mean of market-adjusted returns equal zero. ***, * Statistical significance in 2-tailed tests at the 1% and 5% levels, respectively. Trading days relative to the bid announcement date No Trading Cash Payment Method [0,+30] -0.11% -0.64% -2.04% 1.68% -1.86% -1.37% 8.06% (-1.09) (-5.40) a (-31.80) a (26.92) a (-28.21) a (-10.50) a (87.22) a [+31,+60] -1.89% 0.05% -1.89% % -1.25% 10.51% (-12.90) a (0.34) (-20.72) a (17.31) a (-21.14) a (-6.66) a (91.89) a [+61,+90] -3.95% 0.55% -3.32% 1.79% % 7.72% (-17.89) a (2.35) a (-22.51) a (18.19) * (-23.26) a (-9.06) a (63.56) a Share Payment Method [0,+30] 2.29% 0.62% -5.12% 0.34% -0.83% % (1.50) (0.48) (-8.01) a (0.60) a (-1.13) (-3.25) a (-2.85) * [+31,+60] -0.43% 0.92% -8.54% 2.04% -0.84% -3.66% -2.4 (-0.29) (0.54) (-9.98) a (3.38) a (-0.90) (-2.57) * (-2.38) [+61,+90] -0.65% 0.99% -5.53% 0.07% -2.12% -2.29% -1.51% (-0.31) (0.61) (-7.60) a (0.09) a (-1.99) (-2.09) * (-1.69) Hybrid Payment Method [0,+30] -3.67% % -9.74% 0.66% -6.55% % (-5.83) *** (-13.29) *** (-20.11) *** (3.49) *** 12.88) *** (-8.94) *** (-8.65) *** (- [+31,+60] -4.58% % % % -4.76% -3.36% (-5.47) *** (-11.41) *** (-15.80) *** (12.12) *** 10.31) *** (-4.96) *** (-4.40) *** (- [+61,+90] % % % 2.85% 12.22% % -5.69% (-7.99) *** (-13.37) *** (-20.42) *** (6.53) * 15.77) *** (-9.35) *** (-5.46) *** (- Other Payment Method [0,+30] -1.24% % 1.36% -0.25% -1.51% -6.47% (-2.90) *** (-2.28) * (2.11) *** (8.15) *** (-0.89) (-3.18) *** 26.72) *** (- [+31,+60] % -0.02% % -0.75% -2.47% 11.02% (-5.93) *** (-0.03) (-1.86) (6.25) *** (-1.60) (-3.81) *** 27.78) *** (- [+61,+90] -5.87% -2.06% -1.87% 2.14% -3.61% -5.43% -9.72% (-7.18) *** (-2.31) * (-4.49) *** (6.35) *** (-6.20) *** (-7.12) *** 21.91) *** (- 17

18 Table 7 The association between pre-bid changes in funds holding of target firms shares and the performance of those shares in the post-bid announcement period decreases to 3 of the fund s holding on day [-31] are classified as heavy net selling, net decreases 1 but < 3 are classified as moderate net selling, while net increases in holdings 1 but < 3 are classified as moderate net buying. net buying includes any net increase in holding 3 of the fund s holding on day [-31]. t-statistics test the null hypothesis that the portfolio-weighted mean of market-adjusted returns equal zero. ***, * Statistical significance in 2-tailed tests at the 1% and 5% levels, respectively. Trading days relative to the bid announcement date No Trading Cash Payment Method [0,+30] % 5.38% 7.54% 5.09% 4.26% 13.26% (7.21) *** (8.70) *** (13.98) *** (22.30) *** (14.63) *** (5.37) *** (10.70) *** [+31,+60] -3.91% -3.05% -1.75% 0.11% -0.93% -0.78% 1.06% (-8.71) *** (-7.63) *** (-4.81) *** (0.76) (-2.47) * (-1.03) (4.06) *** [+61,+90] -0.63% -1.61% -1.15% -1.68% 1.48% -2.06% -3.52% (-1.58) (-4.72) *** (-4.68) *** (-9.40) *** (5.33) *** (-6.35) *** (-11.86) *** Share Payment Method [0,+30] n/a n/a n/a 5.08% n/a n/a n/a (3.45) *** [+31,+60] n/a n/a n/a % n/a n/a n/a (-19.72) *** [+61,+90] n/a n/a n/a % n/a n/a n/a (-25.12) *** Hybrid Payment Method [0,+30] 17.54% 12.61% 13.02% 5.58% 11.72% 12.61% 4.85% (8.50) *** (7.31) *** (13.63) *** (6.48) *** (12.41) *** (5.75) *** (2.82) *** [+31,+60] -0.71% -1.54% -1.47% -1.72% -1.69% -1.54% -2.84% (-2.06) (-5.33) *** (-9.18) *** (-14.18) *** (-10.67) *** (-4.19) *** (-9.86) *** [+61,+90] -0.16% -0.48% -0.46% -0.55% -0.54% -0.48% -0.99% (-1.22) (-4.32) *** (-7.36) *** (-11.77) *** (-8.83) *** (-3.39) *** (-8.85) *** Other Payment Method [0,+30] 11.27% 10.41% 17.03% -2.12% 12.24% (7.62) *** (7.36) *** (26.97) *** (-2.57) *** (10.92) *** (13.13) *** (7.59) *** [+31,+60] -7.65% -6.42% -8.16% -4.25% -5.95% -8.45% -8.85% (-12.61) *** (-9.63) *** *** (-13.59) *** (-10.57) *** (-20.02) *** (-22.73) *** (-28.65) [+61,+90] % -5.56% 5.56% -2.05% -3.83% -2.67% (-5.69) *** (-4.76) *** (-6.93) *** (6.69) *** (-2.17) * (-3.20) *** (-1.95) 18

19 Table 8 The association between post-bid changes in funds holding of target firms shares and the subsequent performance of those shares decreases to 3 of the fund s holding on day [-1] are classified as heavy net selling, net decreases 1 but < 3 are classified as moderate net selling, while net increases in holdings 1 but < 3 are classified as moderate net buying. net buying includes any net increase in holding 3 of the fund s holding on day [-1]. t-statistics test the null hypothesis that the portfolio-weighted mean of market-adjusted returns equal zero. ***, * Statistical significance in 2-tailed tests at the 1% and 5% levels, respectively. Trading days relative to the bid announcement date No Trading Cash Payment Method [0,+30] 6.92% 4.99% 9.67% 6.95% % 9.68% (5.76) *** (5.65) *** (12.24) *** (22.40) *** (6.77) *** (2.20) * (9.55) *** [+31,+60] 0.77% -1.53% % -0.83% -1.31% (2.02) * (-5.16) *** (-3.63) *** (-4.50) *** (4.85) *** (-1.47) (-6.49) *** [+61,+90] 2.28% -2.74% -2.62% -0.22% 0.02% -0.97% -3.12% (3.71) *** (-7.71) *** (-10.48) *** (-1.36) (0.08) (-2.29) * (-9.29) *** Share Payment Method [0,+30] 6.05% (2.53) * [+31,+60] % (-10.91) *** [+61,+90] % (-14.29) *** Hybrid Payment Method [0,+30] 18.83% % 3.03% 3.58% 15.64% 18.83% (18.27) *** (5.56) *** (19.24) *** (4.48) *** (3.66) *** (9.71) *** (27.96) *** [+31,+60] -0.49% -2.25% -0.77% -2.66% -3.02% -1.01% -0.49% (-2.77) *** (-8.87) *** (-5.33) *** (-26.11) *** 18.22) *** (-3.57) *** (-4.27) *** (- [+61,+90] -0.08% -0.76% -0.19% -0.92% -1.06% -0.28% -0.08% (-1.12) (-7.71) *** (-3.32) *** (-23.24) *** 16.48) *** (-2.54) * (-1.74) *** (- Other Payment Method [0,+30] -5.13% -4.45% % 5.28% % (-2.62) * (-3.20) *** (-8.12) *** (7.62) *** (3.22) *** (0.49) (-2.17) * [+31,+60] -5.95% -0.76% 1.32% -7.49% -3.22% -1.26% -6.51% (-7.27) *** (-1.03) (2.74) *** (-31.12) *** (-4.34) *** (-1.28) (-10.69) *** [+61,+90] 10.89% -2.92% -5.39% % -3.15% 13.36% (4.19) *** (-1.48) (-4.29) *** (1.11) *** (-4.18) *** (-1.42) (6.31) *** 19

20 Figure 1 Distribution of Managed Fund Trading prior to Cash and Stock Mergers % 4.5% 4.7% % % 59.7% 11.7% % 3.5% No Trading purchase as a % of beginning stock holdings % Cash Shares Figure 2 Distribution of Managed Fund Trading on and after Cash and Stock Mergers % % 4.5% 13.1% 7.6% 49.2% % 6.4% 4.5% 3.1% No Trading 11.3% 10.5% Cash Shares purchase as a % of beginning stock holding 20

21 Figure 3 Distribution of Managed Fund's Target Stock Trading Prior to Cash & Stock Mergers Cash Shares 2 1 No Trading Figure 4 Distribution of Managed Fund's Target Stock Trading After Cash & Stock Mergers Cash Shares No Trading 21

22 Figure 5 Prior and post announcement trading of bidder stocks and CARs 6% Prior announcement tradingof bidder stocks w ith cash payment method and CARs 3 Post announcement trading of bidder stocks w ith cash payment method and CARs 4% 2% 25% 2 15% -2% -4% 1 5% -5% -6% -1 5% Prior announcement trading of bidder stocks w ith hybrid payment method and CARs 1 Post announcement trading of bidder stocks w ith hybrid payment method and CARs -5% % % % -5 5% 4% 3% 2% 1% -1% -2% -3% -4% Prior announcement trading of bidder stocks w ith 'other' payment method and CARs 1 5% -5% -1-15% -2-25% -3 Post announcement trading of bidder stocks w ith 'other' payment method and CARs 22

23 Prior announcement trading of bidder stocks w ith share payment method and CARs No Trading Post announcement trading of bidder stocks w ith share payment method and CARs 5% -5% -1-15% -2-25% No Trading 23

24 Figure 6 Prior and post announcement trading of target stocks and CARs Prior announcement trading of target stocks with cash payment method and CARs Post announcement trading of target stocks with cash payment method and CARs 2 15% 15% 1 1 5% 5% Prior announcement trading of target stocks with hybrid payment method and CARs Post announcement trading of target stocks with hybrid payment method and CARs % 15% 1 1 5% 5% -5% Prior announcement trading of target stocks with 'other' payment method and CARs Post announcement trading of target stocks with 'other' payment method and CARs % 5% 1 5% -5% -5% -1-1 No Trading -15% No Trading 24