Case Study on Uncertainty Triggers Overreaction: Evidence from Corporate Takeovers

Description
Case Study on Uncertainty Triggers Overreaction: Evidence from Corporate Takeovers:- A financial planner or personal financial planner is a practicing professional who prepares financial plans for people covering various aspects of personal finance which includes: cash flow management, education planning, retirement planning, investment planning, risk management and insurance planning, tax planning, estate planning and business succession planning

Case Study on Uncertainty Triggers Overreaction: Evidence from Corporate Takeovers
Abstract Behavioural finance models suggest that under uncertainty, investors overweight their private information and overreact to public signals. We test this prediction in a M&A's framework. We find that under high information uncertainty, when investors are more likely to possess firm-specific information, they generate highly positive and significant gains following the announcement of private stock, public cash and private cash acquisitions (positive news) while the market heavily punishes public stock (negative news) deals. On the other hand, under conditions of low information uncertainty when investors do not possess private information, the market reaction is complete (i.e. zero abnormal returns) irrespective of the type of acquisition.
Key Words: Information Uncertainty, Private Information, Investor Sentiment, Takeover Gains

1. Introduction

Extensive literature has investigated short-run bidder gains and possible factors which affect shareholders wealth following the announcement of a takeover deal1. Traditional studies that have examined short-term bidder performance have usually worked under the assumption that the market is efficient. Under this assumption, the short-term market reaction to a merger announcement is believed to depict the net present value of potential synergy gains that can potentially be created, minus any premium which may have paid for the target firm. Mergers are believed to be rational responses to economic disturbances (Mitchell and Mulherin, 1996; Harford, 2005; Gugler et al., 2006; Owen, 2006) such that the combination of two firms can become more attractive than remaining separate entities. In this neoclassical setting, mergers provide a vehicle for firms to unlock synergistic gains invoked by the economic shock. Alternatively, if we allow for behavioural theorists to enter the framework, then the returns experienced by both bidders and targets can reflect a wide range of psychological phenomena variations in investor recognition (Merton (1987); Foerster and Karolyi (1999)), misvaluation at the firm or market-level2 (Rhodes-Kropf and Viswanathan (2004); Shleifer and Vishny (2003)), information uncertainty (Daniel et al. (1998); Odean (1998)) and many more. In particular, DeBondt and Thaler (1985) provided seminal work in which gains could be unlocked from shorting past winners and buying past losers. The theoretical idea behind this strategy was that investors overreact to public news announcements. Good news led to a positive overreaction in a firm's stock price while bad news caused the reverse. However, at some point the market would become aware and a significant reversal would be found consistent with the idea of investor overreaction.

The behavioural literature itself continues to grow endlessly. Particular fields of interest of late have remained concerned with the resultant effects of information uncertainty. This paper takes note of this rising school of thought and offers a behavioural perspective to help explain short-term bidder abnormal returns, a topic which has attracted much debate (Jensen and

1

For evidence on announcement period gains to acquirers see Dodd and Ruback (1977) and Moeller, Schlingemann and Stulz (2004) for the US and Draper and Paudyal (2006) for the UK. Recent evidence shows that the announcement period gains to bidders are dependent on the listing status of targets: acquirers of listed targets tend to lose, while unlisted target acquirers gain (Faccio, McConnell and Stolin, 2006; Draper and Paudyal, 2006). 2 Fuller et al. (2002), and Draper and Paudyal (2008) claim that the short-run market reaction to bidder takeover announcements may reflect revaluation gains. Fuller et al. (2002) also claim that the gains in first-order deals may be higher because they incorporate revaluation gains as well as the potential synergy gains.

2

Ruback, 1983; Limmack, 1991; Chang, 1998; Shleifer and Vishny, 2003; Rosen, 2006; Bouwman, Fuller and Nain, 2009).

Experimental evidence shows that investors tend to overestimate the precision of their information, especially in cases where they have been personally involved in the collection of this information (Odean (1998)3). The theoretical model of Daniel et al. (1998) complements this predicting that investors are overconfident about the private information they hold. As a result, they attribute more weight to their private information and subsequently fail to react fully to public signals. In other words, they underreact to public information stimulus.

Additionally, Daniel et al. (1998, 2001) also claim that investors become even more overconfident under conditions of information uncertainty. A large part of the psychology literature4 suggests that individuals overvalue their own abilities in the decision making process whilst also overestimating the precision of the outcome of the decision made 5. Investors undoubtedly extract information from various sources (for example, from financial statements, the press and rumours amongst others). However, if they overestimate their own ability to extract this information, or they overweight the precision and significance of this information, then the resultant effect will be an overreaction due to the underestimation of the forecast error involved in the decision-making process. Daniel et al (1998) define overconfident investors as those which overestimate the precision of their private information as opposed to the public signals available. They find that overconfident investors who possess private information will overweight this information, leading to a stock price overreaction. When an investor trades on his/her private information/signals and subsequently receives a public signal which serves to confirm the trading strategy being executed, then the investor's confidence will rise. One of the advantages of the model of Daniel et al. (1998) when compared to previous behavioural models6 is that it assumes that investors become
3

Odean (1998) claims that there is excessive trading in equity markets. He explains this as a result of investors who are overconfident. Markets, in turn, become affected by this psychological bias as investors inevitably trade a lot because they repeatedly feel the gains they earn are not enough. Interestingly, securities purchased by overconfident investors are found to underperform those they sell supporting that overconfidence destroys value and leads to excessive trading volumes. 4 See, for example, Griffin and Tversky (1992), Greenwald (1980), Svenson (1981), Cooper et al. 1988, Taylor and Brown (1988), Alpert and Raiffa (1982), Fischhoff, Slovic, and Lichtenstein (1977), Batchelor and Dua (1992), Lichtenstein, Fischhoff, and Phillips (1982) and Yates (1990). 5 Hirshleifer (2001) suggests that psychological biases grow both under conditions of greater uncertainty, in the absence of accurate feedback about fundamentals. 6 Kyle and Wang (1997), Odean (1998) and Wang (1998) define overconfidence as overestimation of information precision regardless of whether the information is private or public.

3

overconfident about private signals and therefore allows for both over- and under-reaction effects. Furthermore, the authors claim that since the model is mainly based on both private information and subsequent under or overreaction, its predictive power will be more evident for firms with higher information uncertainty.

Zhang (2006) also suggests that investor overreaction should be more prominent under conditions of information uncertainty since investors become more overconfident for firms that are hard to value. He finds that under conditions of information uncertainty, announcements of good news generate relatively higher abnormal returns while announcements of bad news generate relatively lower abnormal returns. While Zhang (2006) controls only for information uncertainties, he does not include private information into his analysis, proposing that further investigation is required.

This paper is motivated by the theoretical behavioural finance models of Daniel et al (1998, 2001) and Hirshleifer (2001) who conclusively suggest that investors are overconfident about their private information and due to this psychological bias, they subsequently overreact to their private information. The psychological bias of overconfidence increases under conditions of information uncertainty when the firm's value is difficult to predict. Zhang (2006) empirically shows that under conditions of uncertainty, good (bad) news generates relatively higher (lower) abnormal returns while when uncertainty is low, there is less market predictability.

Merger announcements undoubtedly convey private information of the firms involved to the public market. The corporate finance literature shows that various types of takeovers convey a message to the market, signalling positive or negative news regarding the intrinsic value of the bidding firm. There is substantial evidence which suggests that the target firm's listing status and the method of payment used to finance the takeover signal different news about the valuation conditions of the bidding firm.

Derived from the seminal work of Myers and Majluf (1984), the signalling literature suggests that managers who believe that their firm's stock price is undervalued will prefer to finance a potential acquisition with cash while when they consider that their stock price is overvalued,

4

they will prefer to conduct equity transactions to capitalise upon this overvaluation7. The signalling hypothesis, as proposed by Travlos (1987), suggests that investors will perceive the announcement of an equity offer for a public target as bad news since, in the Myers and Majluf (1984) setting, they interpret that the bidding firm must be overvalued. Otherwise, why would the manager wish to use his/her stock? On the other hand, cash offers are perceived as good news regarding the acquiring firm's intrinsic value as the manager must believe his firm is not overvalued, potentially even undervalued. Mergers are a way for managers to convey their beliefs over the value of their firm to the market. In this way, the private information of the manager enters the public spectrum of the market at the time of a merger announcement, predominantly via the manager's financing choice.

On the other hand, Chang (1998), and Draper and Paudyal (2006) suggest the opposite effect for the market's reaction to the acquisition of private targets to be financed using equity. Investors interpret such announcements as good news and this for several reasons. Most importantly, because unlisted firms tend to be owned by a small number of owners, then these individuals are portrayed as having a stronger incentive to carefully examine the true value of the bidders stock. If they believe for it to be overvalued, it would be an irrational act for these owners to accept the bidder's equity as payment for their firm. Therefore, it is highly unlikely that the owner of the privately held firm will accept stock if they believe it to be overvalued as they will effectively 'lose-out'.

Considering this, the signalling effect of private stock acquisitions can be classified as a positive one as the acceptance of the bidder's stock by the unlisted target conveys the news that the bidder's stock price must not be overvalued. A cash acquisition for a private firm is usually a positive announcement also but in truth, does not reveal a lot of information about the bidder's intrinsic value in this setting. A reasonable assertion to make is that an acquirer paying for an unlisted target with cash may be less uncertain regarding the level of potential synergy gains which can be extracted from the proposed combination and as such is confident enough to offer cash. This loosely infers that the bidder is confident as they may be motivated to avoid the issuance of equity so as to avoid sharing potential synergy gains with the

7

Shleifer and Vishny (2003) provide a model in which acquisitions are driven by firm-misvaluation. They support the idea that overvaluation provides an incentive to acquire a less overvalued target using equity.

5

ownership of the target firm 8. Therefore, a cash acquisition does not directly reveal information about the bidder's stock value but can, in general, be classified as a relatively positive piece of information.

Taking heed of the models and empirical research developed to date, we examine the shortterm market reaction following takeover announcements for UK bidding firms under two conditions - firstly when there is high and low information uncertainty9 surrounding bidding firms and secondly, when investors are more or less likely to possess private information. Under high information uncertainty conditions investor overconfidence is expected to be much higher. When this is interrelated with investors who are more likely to possess private information, then this can lead to a high overreaction following takeover announcements. Investors will overreact and generate highly positive abnormal returns following the announcement of acquisitions which signal 'good' news - i.e. private targets financed with cash or stock and public targets financed with cash. Under the same conditions, the market reaction will be highly negative following announcements of takeovers which signal 'bad' news - i.e. public targets paid for with stock. On the other hand, when information uncertainty conditions are expected to be low so there is a low level of uncertainty regarding the bidder's intrinsic value, coupled with investors who are less likely to have collected private information, then the market reaction is expected to be complete (i.e. zero abnormal returns).

We study the UK market for several significant reasons. The UK market has a large majority of private-target acquisitions. Doukas and Petmezas (2007) report that 91% of all UK deals involve the acquisition of a privately-held target. We reason that this is a perfect dataset with which to investigate information uncertainty. The nature of the acquisition of a private-target inevitably induces a great deal of information uncertainty through the lack of public information available for such firms. Thus, valuation of the target and of potential synergies is highly subjective. With this in mind, the UK is a most appropriate choice of dataset for examining the effects of private information and information uncertainty. In addition, Faccio and Masulis (2003) document that UK bidders account for 65.3% of all European deals.

8

The issue of equity to an unlisted target's owners would result in the creation of blockholders in the combined entity. 9 By information uncertainty, we mean ambiguity regarding the bidding firm's value (Zhang (2006))

6

Morevoer, the UK is the second most-active merger market in the world aside from the US. Together, these reflect our motivations to study the UK market.

We employ four proxies10 for information uncertainty and these include: the age, size, sigma and trading volume of the bidding firm. To capture whether investors are more likely to possess private information or not, we employ stock price synchronicity as introduced by Roll (1988) and further developed by both Morck et al. (2000) and Chen et al. (2007). Roll (1988) suggests that a low R2 value should be observed in periods of no public news about the firm, indicating that the price movement is triggered by private information. Chen, Goldstein and Jiang (2007) among others11 adopt synchronicity as a measure of stock price informativeness and show that there is a strong positive relationship between the amount of private information within stock prices and the sensitivity of corporate investment to stock prices. Roll (1988) claims that the measure of stock price nonsynchroncity is not correlated with public information and thereby serves as a good approach to capture private information. In Roll's own words, ''the financial press misses a great deal of relevant information generated privately'' (Roll, 1988: 564).

The main findings suggest that under conditions of high information uncertainty and when investors are more likely to possess private information, announcements of takeovers which signal positive news for the bidding firm's intrinsic value (i.e. private stock, private cash and public cash deals) do indeed generate highly positive abnormal returns while takeovers which signal negative news (i.e. public stock) suffer high losses. On the other hand, when uncertainty is lower and investors are likely to possess private information (high synchronicity), zero economical and statistical abnormal returns are obtained irrespective of the type of the deal.

This paper contributes to the corporate and behavioural finance literature in several ways. First, it offers a behavioural approach to explain short-run bidder gains. Previous studies assume that the market is semi-strong efficient and short-run bidder gains captures either potential synergy or revaluation gains. We offer evidence that the findings of Travlos (1987)
10

Age is used as a proxy for information uncertainty by Zhang (2006), Jiang, Lee and Zhang (2004), Barry and Brown (1985), Size by Zhang (2006) and Sigma by Zhang (2006), Jiang, Lee and Zhang (2004).
11

Morck, Yeung and Yu (2000), Durnev, Morck, Yeung and Zarowin (2003), Durnev, Morck and Yeung (2004), Jin and Myers (2006), Fernades and Ferreira (2008), Ferreira, Ferreira and Raposo (2008)) have used stock price nonsynchronicity to examine price informativeness.

7

and Chang (1998) may be driven by high investor sentiment. Second, it shows that investor sentiment is a crucial factor to explain and understand various financial phenomena. When investor sentiment is high, the market overreacts to various types of takeover announcements while low sentiment leads to under reaction. Third, it contributes to the behavioural finance literature by empirically examining the propositions of Daniel et al. (1998). Forth, this study simultaneous examines the effect of uncertainty and private information in the same framework. Finally, it offers further evidence that the market reacts asymmetrically following the announcement of positive and negative signals.

The remainder of this paper is structured as follows. Section 2 describes the data and methodology. Section 3 analyses the empirical findings before Section 4 summarizes the conclusions of the investigation.

2. Sample Selection, Data and Methodology

The sample consists of takeover announcement deals undertaken by UK bidding firms for the period between 01/01/1985 and 31/08/2009. The announcements were collected by Thomson Security Data Corporations (SDC). To be included in our final sample, the deals needed to meet the following criteria:

o

The acquirer is a U.K. firm publicly traded on the London Stock Exchange (LSE) with five days of return data available around the announcement date of the takeover as well as available data for one to three years returns from the DataStream database.

o

The target company is either a listed or unlisted company and can be a domestic or a foreign company.

o o o o o

The acquiring firm purchases at least 50% of the target's shares. The deal value is ?1 million or more. The deal value represents at least 1% of the market value of the acquirer. Multiple deals announced within a 5 day period are excluded. Financial and utility firms, for both bidders and targets, are excluded from the sample (Fuller, Netter and Stegemoeller (2002)).

8

The initial sample from Thompson One Banker was 20,306 deals and after the exclusion of deals according to the above criteria, our final sample totaled 7,019 deals. Of these deals, 4,058 were for takeovers of private targets, 713 were for public firms and 2,248 were initiated to gain control of a subsidiary. We include all private, public and subsidiary firms in our initial sample in order to get a larger sample size and thus allow for us to obtain a more unbiased distribution when stratifying the deals into high versus low information uncertainty according the four proxies used.

Table 1 presents the time-distribution for takeovers from 1985 to 2009 for the overall sample. In addition, the deals are stratified according to the synchronicity measure of private information and the four measures we employ in this work for the classification of information uncertainty, as outlined in the next section.

We observe that at the beginning of the sample period, high synchronicity firms appear to relatively outnumber low synchronicity ones. This indicates that in the late eighties/early nineties, there were less firms whose stock price was more likely to incorporate private information. As the sample period has progressed, the market appears to have become more efficient while investors are more sophisticated today than they previously were. This is supported by the increasing trend of low synchronicity firms over the period 1985 to 2009.

[Insert table 1 about here]

3.1 Measures of Information Uncertainty

In order to capture information uncertainty, we employ four proxies recommended within the literature. The first measure employed is Age. The existing literature suggests that the younger the firm is, the higher the amount of uncertainty there will be regarding the firm's value (Zhang (2006), Jiang, Lee and Zhang (2004), Barry and Brown (1985)). Young firms are associated with a lower amount of information dissemination and thus the age of the firm can be used as a proxy for the level of information uncertainty surrounding its value. We measure age as the difference between the date of incorporation of the firm and the date of the announcement of the acquisition.

9

Size is the second proxy employed in order to capture information uncertainty regarding the bidder's value. Smaller firms are less likely to disclose a lot of information and are less diversified than larger firms. However, small firms also have a lower number of suppliers, investors and customers and therefore the accessibility of information can be more difficult. Hence, small size firms are more likely to be associated with a higher degree of information uncertainty (Zhang (2006)). We measure size as the Market Value (MV) of the bidding firm 20 days before the announcement of the acquisition.

The third proxy employed to capture information uncertainty is the Sigma of the bidder. Bidders with high return volatility are more likely to exhibit uncertainty concerning their true value (Zhang (2006), Jiang, Lee and Zhang (2004)) than those with more stable operations. Sigma is measured as the daily bidder excess returns 200 days before the announcement of the acquisition.

Finally, the Trading Volume is the fourth and last proxy used within this work to measure information uncertainty. Low trading volumes suggest that a lower number of investors are aware or are following the firm. Thus there is likely to be less trading activity associated with the bidders who exhibit a low trading volume. The trading volume of the bidder is measured as the firm's average trading volume twelve months prior to the announcement of the acquisition.

3.2 Measure of Private Information

Recent models emanating from the school of behavioural finance have focused on the role of private information and its subsequent impact on investors' cognitive biases alongside their following investment decisions. One of the roles of financial markets is to facilitate the production and accumulation of information into stock prices. This occurs through the impact of the trading activities of speculators on stock prices. Financial economists support the notion that stock returns incorporate firm-specific and market-wide information. Roll (1988) claims that stock prices move together depending on the amount of firm-specific or marketwide information capitalized in the stock prices. He also explains that stock price movements are influenced by market-wide economic shocks, by industry shocks and by news specific to the firm.

10

Chen, Goldstein and Jiang (2007) suggest that managers learn from the private information incorporated in stock prices and take advantage of this information within their corporate investment decisions. More specifically, they suggest that private information is incorporated in stock prices through speculators trading activity. An important point to note is that a high level of private information does not imply that stock prices are close to fundamentals. The variation between a stock price and its fundamental value depends on the amount of public information available for that stock as well. The incorporation of private information is a timely procedure and that may imply that stock prices with more private than public information might be further away from fundamentals. Theoretical evidence (Dow and Gorton (1997); Subrahmanyam and Titman (1999)) suggests that managers can extract useful information hidden in stock prices. Stock prices accumulate a lot of information from various trading participants in the market who do not have any other way of communicating with the firm apart from via the trading process. Consequently, stock prices may incorporate information that managers do not have. Different stocks have different levels of private information incorporated within them due to the various costs involved in the acquisition and production of such information (Grossman and Stiglitz (1980)).

This investigation follows Chen, Goldstein and Jiang (2007) to measure stock price synchronicity. The variation of stock returns can be decomposed into the following components - market-wide variation, industry-specific variation and firm-specific variation. This work needs to capture the last component of firm-specific variation which can be measure by the R2 of the following regression:

ri,j,t? ?

i,0

?

?

i,m

rm,t? ?

i,j

rj,t? ?

,t i

where ri,j,t is the return of bidder i in industry j at time t, rm,t is the market return at time t and rj,t is the return of industry j at time t. To construct this regression, weekly returns for a period of 24 weeks (6 months) before the announcement of the acquisition are used.

3.3 Short-Run Event Study Methodology To calculate the acquiring firms' performance and identify the short-run impact of information uncertainty and private information, we employ standard event study
11

methodology (Fuller et al (2002)) to calculate the Cumulative Abnormal Returns (CARs) for a five-day period (-2, +2) around the announcement date, as provided by Datastream. We estimate abnormal returns using the modified market model as follows:
ARi, ?
t

Ri, ?
t

R m, t

Where ARi,t is the excess return of bidder i on day t; Ri,t is the return of bidder i on day t measured as the percentage change in return index including dividends of bidder i; and Rm,t is the market return estimated as the percentage change in FT-All share Index (value weighted) on day t. The CARs are calculated as the sum of the Abnormal Returns (ARi,t) for the five days surrounding the announcement of the bid as per the following equation: CARi?
t? t?

?

(Ri? Rm )

T-statistics are used to test the null hypothesis that the mean CAR is equal to zero for a sample of n firms. We do not report the t-statistic in tables but the p-value instead. The pvalue provides a sense of strength of the evidence against the null hypothesis. The lower the p-value, the stronger the evidence that the mean CAR is statistically significantly different from zero.

3. Empirical Results

3.1 Univariate Analysis Table 2 presents the five-day cumulative abnormal returns (CARs) for the full sample as sorted by the target's listing status (i.e. private or public) and by the method of payment used to finance the deal (i.e. cash, stock or mixed). Bidders for the overall sample generate 1.46% (p value = 0.000) abnormal returns, which is mainly driven by the positive performance of takeovers for private target firms, which enjoy significantly positive gains of 1.69% (0.000). Private deals largely overpopulate the UK M&A market with circa 58% of our deals involving the acquisition of an unlisted target. For acquisitions of targets which are listed firms, bidders suffer marginally significant losses of -0.46% (0.113). While the target listing status impacts the returns generated for bidding firms, the signalling literature indicates that the method of payment used by the bidder to finance the deal plays a significant role in determining the returns to be experienced. With respect to the method of payment used, acquisitions for private targets paid for with stock (3.60%) enjoy 2.47% (0.012) significantly
12

more abnormal returns than those paid for with cash (1.13%) (see Chang (1998), Ang and Cohers (2001), Draper and Paudyal (2006), and Fuller, Netter and Stegemoller (2002)). On the other hand, takeovers for public target firms paid for with equity suffer significant losses (-2.04%) while those paid for with cash generate positive abnormal returns (0.95%). The difference is statistically significant at the 1% significance level (Travlos (1987)).

Table 2 presents the performance of subsidiary firms as well as those privately-held and publicly listed. We observe a similar pattern for bidders which acquire subsidiary firms to those which merge with a privately-held firm. However, the rest of the analysis here focuses mainly on takeovers for private and public targets paid for with cash and stock since the signal from such type of deals is more straightforward for the market to interpret. We choose to include subsidiary and mixed deals in our initial sample in order to have a larger sample size so that we can obtain a more unbiased distribution when dividing the deals into high versus low information uncertainty in line with the four proxies outlined in the previous section.

[Insert table 2 about here]

Tables 3 illustrates the short-term performance of takeovers for private and public targets paid for cash and stock respectively (denoted PrivateCash, PrivateStock, PublicCash and PublicStock respectively) under conditions of information uncertainty as captured by the Age proxy. The younger a firm is, the higher the level of uncertainty there is regarding the firms true value. As outlined earlier, Chang (1998) indicates that PrivateStock acquisitions serve as a positive signal to the market that bidding firm's share price is not overvalued due to the target's acceptance of the bidder's equity. In Panel A, the overall short-term performance of the PrivateStock portfolio is 3.80% (0.000). Under conditions of private information, PrivateStock deals generate even stronger positive abnormal returns of 5.08% (0.005) for the bidder while under low information uncertainty when the private information of investors is likely to be lower, bidders earn lower abnormal returns of 1.93% (0.163). The difference of 3.15% is statistically significant at the 16% level.

Daniel et al. (1998, 2001) suggest that investors tend to overweight their private information and become even more overconfident under conditions of information uncertainty. When private information is included, the overconfidence of the investor is likely to become even
13

more intense so that the differences should be amplified. PrivateStock deals generate even higher abnormal returns of 6.32% (0.062) when there is high information uncertainty and low synchronicity (i.e. higher levels of private information). On the other hand, in the absence of uncertainty and private information, private stock deals obtain marginally positive but insignificant abnormal returns (0.49%). The differential between the two extreme portfolios is heightened at 5.83%, significant at the 13.5% significance level. This indicates that when investors have a level of private information, there is a highly positive reaction following the announcement of events that signal positive news about the bidding firm's intrinsic value. On the other hand, when investors are less likely to have private information and thus have less potential to overestimate its precision, there is no significant market reaction.

[Insert table 3 about here]

The picture for PrivateCash (Panel B) and PublicCash (Panel D) takeovers, which also signal positive news, is similar to the one presented described above for private stock deals. PrivateCash and PublicCash deals are positive but are simultaneously quite indirect signs that the bidding firm is not overvalued. Therefore, the higher the level of private information held by investors, the more positive the market reacts following such deals (2.43% (0.003) for PrivateCash and 3.85% (0.085) for PublicCash) in Panels B and D respectively. Conversely, when the level of private information is low, marginally positive but insignificant abnormal returns are obtained (0.19% (0.517) for PrivateCash and 0.80% (0.166) for PublicCash).

The overall picture is reversed for acquisitions for public target firms paid for with equity (Panel C). PublicStock acquisitions signal negative news to the market regarding the intrinsic value of the bidding firm's value. In the overall sample, PublicStock deals generate significantly negative losses of -2.35% (0.001) for the bidding firm. When we control for uncertainty, the negative performance becomes even more negative under high conditions of information uncertainty with losses of -3.87% (0.005) while it declines to -0.28% (0.786) under lower information uncertainty. When investors are more likely to possess private information, they overweight this information especially under conditions of uncertainty and become even more overconfident. Following the announcement of PublicStock acquisitions which, as discussed, signal negative news (Travlos (1987)), and when investors have high levels of information uncertainty, investors overreact and generate even stronger negative abnormal returns of -5.89% (0.003). On the other hand, under lower information uncertainty
14

and when investors are less likely to possess private information, the propensity for investors to be overconfident is expected to be quite low and therefore no significant market reaction is observed. The differential for public stock takeovers for the two extreme portfolios (High versus Low) is 6.50% (0.005), statistically significant at the 1% level.

The findings described above can be visualized in Figure 1. The first column of each group shows the cumulative abnormal returns for the overall sub-portfolios, which is similar to existing empirical literature. The second column depicts the market reaction to takeover announcements when information uncertainty is high while the third column represents the cumulative abnormal returns for the portfolio in the absence of uncertainty and private information. It is clear that under conditions of uncertainty, there is a market overreaction. There is a highly positive reaction for positive-signalling deals (i.e. PrivateStock, PrivateCash and PublicCash) and a highly negative one for those takeovers signalling negative news (i.e. PublicStock). On the other hand, in the absence of uncertainty and private information, the market reaction is complete, as displayed in the third column for each of the four subportfolios.

For robustness, we conduct this analysis using a further three proxies, namely size, sigma, and trading volume in order to capture information uncertainty. The results are provided in Tables 4, 5 and 6 respectively. The overall picture for the four types of acquisitions (i.e. PrivateStock, PrivateCash, PublicStock and PublicCash) remains highly similar to the evidence indicated by the age proxy.

[Insert tables 4, 5 and 6 about here]

Conclusively, the findings remain consistent across all four proxies. Some of these measures are indeed positively correlated such as size and trading volume. However, other pairs, such as sigma and age or size are negatively correlated (Table 2B). We should note that the differences in the number of observations between the overall portfolios and the High vs. Low ones are due to the fact that medium portfolios are omitted and not included for brevity.12

12

Analytical results which include omitted portfolios are available upon request.

15

5.2. Multivariate Analysis

The M&A literature has documented a number of different factors that affect the performance of bidding firms surrounding the event, such as book-to-market (Rau and Vermaelen, 1998), size (Moeller et al., 2004) relative size (Fuller et al., 2002) and industry diversification (Doukas and Kan, 2004).

The results generated so far within the univariate analysis indicate that when investors hold high levels of private information, PrivateCash and PublicCash deals generate positive and significant abnormal returns while PublicStock deals suffer highly negative abnormal returns. Under conditions where the level of private information held is low; there is no market reaction irrespective of the type of deal. To better examine whether differences in acquirer and deal characteristics explain the abnormal return differentials, we adopt a multivariate regression framework whereby announcement period returns for bidders are regressed against a set of explanatory variables that have been proven in the literature to affect bidders' performance. Moreover, the multivariate framework enables us to overcome issues related to the small number of observations in some portfolios13.

[Insert table 7 about here]

In all regressions we include the following control variables: the bidder's book-to-market value, which is measured by the bidder's net book value of assets divided by its market value one month before the announcement of the deal; the deal's relative size, which is measured as the ratio of the deal value over the bidder's value; a dummy variable for diversifying deals which takes the value of 1 when the acquirer's two-digit SIC code is different from that of the target, and zero otherwise; and a dummy variable that takes the value of 1 if the target is a domestic firm. Finally, other explanatory variables include: the acquirer's lagged excess return for 180 days prior to the bid's announcement; and the market portfolio return (FT-All Share) for the same 180-day period prior to the announcement.

For brevity, we present multivariate analysis only for PrivateStock and PublicStock deals which signal the most positive and negative news respectively (Table 7). The results for
13

There are a relatively low number of acquisitions for public targets paid for with equity in the UK.

16

PrivateCash and PublicCash deals follow similar patterns to the evidence discussed here for PrivateStock acquisitions. As it has been noted, PrivateStock deals serve as the most positive signal while PublicStock acquisitions are the most negative signal regarding the bidding firm's intrinsic value.

Panel A of Table 7 presents the results for the Age proxy. In regressions (1) and (5), we include a dummy variable for High Information Uncertainty (HighIU) that takes the value of 1 if the bidding firm belongs in the top 30% of the youngest firms in our sample. As expected from the univariate analysis, the coefficient is positive for PrivateStock deals (regression 1) and negative for PublicStock deals (regression 5). In both cases, the coefficients are not statistically significant. In Regressions (2) and (6), we include a dummy variable that takes the value of 1 if the deal belongs in the high uncertainty and low synchronicity group. In regression (2), the coefficient carries a positive and insignificant value (0.032) but in regression (6), the coefficient is negative (-0.054) and highly statistically significant. This indicates that under uncertainty, investors overweight their private information and we observe a significant negative relationship between CARs and high levels of private information.

The opposite effect is observed in regressions (3) and (7). The dummy variable HsLiu attempts to capture takeovers when sentiment is low. In regressions (4) and (8), we include a dummy variable that capture all four combinations of high and low information uncertainty along with high and low synchronicity. The results remain similar. The LsHiu (high sentiment) coefficient remains positive and insignificant for PrivateStock deals (regression (4)) while negative and high significant for PublicStock deals (regression (8)). Similar findings are observed for the other three proxies for information uncertainty - size, sigma and trading volume in Panels B, C and D respectively.

From the results gathered, the main conclusions that can be drawn are that there is a positive but insignificant relationship between PrivateStock (a positive signal) and conditions of high levels of private information, while there is also a highly significant negative relationship between PublicStock deals and conditions of high levels of private information. The relationships are reversed in the absence of private information. The multivariate analysis offers supportive evidence to the conclusions drawn from the univariate analysis. Furthermore, these finding are consistent with the existing theory that proposes an
17

asymmetric response to positive and negative news. In particular, Epstein and Schneider (2008), Bernard et al. (1997), La Porta et al. (1997), and Skinner and Sloan (1999) find significant differences in the markets response with regards to receiving a signal of good and bad news. 4. Conclusion This paper examines the market response to takeover announcements. We adopt a behavioural approach to UK mergers and acquisitions under conditions of information uncertainty and private information. More specifically, we examine the short-term bidder gains controlling for information uncertainty regarding the bidder and the effects of the deal, as well as the level of the investor's private information in the surrounding environment of the bidder.

The main findings suggest that under conditions of high information uncertainty and when investors are more likely to possess private information, announcements of takeovers which signal positive news concerning the bidding firm's intrinsic value (i.e. PrivateStock, PrivateCash and PublicCash deals) generate highly positive abnormal returns while takeovers which signal negative news (i.e. PublicStock) suffer high losses. On the other hand, when uncertainty is lower and investors are less likely to possess private information (i.e. high synchronicity), zero economical and statistical abnormal returns are obtained irrespective of the type of the deal undertaken.

This evidence is consistent with the theoretical work of Daniel et al. (1998, 2001) who suggest that investors are overconfident and overreact to public announcements under conditions of uncertainty. Furthermore, they claim that investors, due to a self-attribution bias, become even more overconfident about their own private information following the signal and overreact even more. Consequently, under uncertainty, investors with private information react highly positively following the announcement of good news (i.e. PrivateCash, PrivateStock, PublicCash deals) while they react very negatively following the announcement of bad news (i.e. PublicStock deals). When there is low uncertainty and investors do not possess private information, the market reaction is complete. The multivariate analysis shows that the coefficients of the high uncertainty dummy are mostly negative and significant following the announcement of public acquisitions paid for with stock. This evidence is consistent with Epstein and Schneider (2008) who suggest that
18

investors react asymmetrically to news when they are ambiguous regarding the firm value. Moreover, evidence from Bernard et al. (1997), La Porta et al. (1997) and Skinner and Sloan (1999) also show that there is a significant difference in the markets response with regards to good and bad news.

One of the most interesting findings of our work is that the evidence shows that the negative performance of public stock deals is enlarged under conditions of high levels of private information while we report zero and insignificant returns in the absence of sentiment. In other words, the negative performance of PublicStock deals seems to be highly affected when the signal is enhanced under uncertainty. In acquisitions for public targets paid for with equity, two financial acts take place. The first concerns the announcement of the takeover itself while the second concerns the offer (issuance) of equity to the target firm. Our findings show that in the absence of uncertainty, when the signalling effect is weak, takeovers for public target firms paid for with equity do not necessarily suffer losses for the bidder as the literature has historically suggested. New evidence from Golubov, Petmezas and Travlos (2011) supports this school of thought. Comparing a sample of PublicStock acquisitions with a sample of SEOs, they show that when the signalling effect from the issuance of equity is removed, equity-finance takeovers are no more value-destructive investment projects.

Overall, this paper offers a behavioural explanation for the market reaction following takeover announcement. The short-run market reaction to M&As announcements reflects either potential synergy or revaluation gains. Our evidence suggests that there is a market overreaction driven by investor biases. Investors' biases increase especially with uncertainty and will also depend on the signal conveyed by each type of takeover. Investors will react either highly positively or negatively with private information following a positive or negative signal respectively. In the absence of uncertainty, the market reaction is complete.

19

References Ang, J. and Kohers, N. (2001), 'The take-over market for privately held companies: the US experience', Cambridge Journal of Economics, 25, 723-748.

Barry, C. B. and Brown, S. J. (1985), 'Differential Information and Security Market Equilibrium', Journal of Financial and Quantitative Analysis, 20, 407-422.

Benos, A. (1998), 'Aggressive of Survival of Overconfident Traders', Journal of Financial Markets, 1, 353-383.

Bernard, Victor, Jacob Thomas, and James Wahlen, 1997, Accounting based stock price anomalies: Separating market nefficiencies from risk, Contemporary Accounting Research 14, 89-136.

Bouwman, C.H.S., Fuller, K. and Nain, A.S. (2009) 'Market Valuation and Acquisition Quality: Empirical Evidence', The Review of Financial Studies, 22, 633-679.

Chan, K. and Hameed, A. (2006), 'Stock price synchronicity and analyst coverage in emerging markets', Journal of Financial Economics, 80, 115-147.

Chang, S. Y. (1998), 'Takeovers of privately held targets, methods of payment, and bidder returns', Journal of Finance, 53, 773-784.

Chen, Q., Goldstein, I. and Jiang, W. (2007), 'Price informativeness and investment sensitivity to stock price', Review of Financial Studies, 20, 619-650.

Daniel, K., Hirshleifer, D. and Subrahmanyam, A. (1998), 'Investor psychology and security market under- and overreactions', Journal of Finance, 53, 1839-1885.

Daniel, K. D., Hirshleifer, D. and Subrahmanyam, A. (2001), 'Overconfidence, arbitrage, and equilibrium asset pricing', Journal of Finance, 56, 921-965.

20

DeBondt, W. F. M. and Thaler, R. (1985), 'Does the Stock-Market Overreact', Journal of Finance, 40, 793-805.

Dow, J. and Gorton, G. (1997), 'Stock market efficiency and economic efficiency: Is there a connection?' Journal of Finance, 52, 1087-1129.

Draper, P. and Paudyal, K. (2006), 'Acquisitions: Private versus public', European Financial Management, 12, 57-80.

Draper, P. and Paudyal, K. (2008), 'Information asymmetry and bidders' gains', Journal of Business Finance and Accounting, 35, 376-405.

Durnev, A., Li, K., Morck, R. and Yeung, B. (2004), 'Capital markets and capital allocation: Implications for economies in transition', Economics of Transition, 12, 593-634.

Durnev, A., Morck, R. and Yeung, B. (2004), 'Value-enhancing capital budgeting and firmspecific stock return variation', Journal of Finance, 59, 65-105.

Durnev, A., Morck, R., Yeung, B. and Zarowin, P. (2003), 'Does greater firm-specific return variation mean more or less informed stock pricing?' Journal of Accounting Research, 41, 797-836.

Einborn, H. (1980), 'Overconfidence in Judgement', New Directions for Methodology of Social and Behavioural Science', 4, 1-16.

Epstein, L. G. and Schneider, M. (2008), 'Ambiguity, information quality, and asset pricing', Journal of Finance, 63, 197-228.

Faccio, M. and Masulis, R.W. (2005) 'The choice of payment method in European mergers and acquisitions', The Journal of Finance, 60, 1345-1388.

21

Fernandes, N. and Ferreira, M. (2008) 'Does international cross-listing improve the information environment', Journal of Financial Economics, 88, 216-244.

Ferreira, D., Ferreira, M. and Raposo, C. (2008), 'Board Structure and Price Informativeness', ECGI - Finance Working Paper No. 160/2007.

Foerster, S.R. and Karolyi, G.A. (1999) 'The Effects of Market Segmentation and Investor Recognition on Asset Prices: Evidence from Foreign Stocks Listing in the United States', The Journal of Finance, 54, 981-1013.

Fuller, K., Netter, J. and Stegemoller, M. (2002), 'What do returns to acquiring firms tell us? Evidence from firms that make many acquisitions', Journal of Finance, 57, 1763-1793.

Golubov, A., Petmezas, D. & Travlos, N.G. (2011). 'When It Pays to Pay Your Investment Banker: New Evidence on the Role of Financial Advisors in M&As', Journal of Finance, Forthcoming.

Griffin, D. and Tversky, A. (1992), 'The Weighing of Evidence and the Determinants of Confidence', Cognitive Psychology, 24, 411-435.

Grossman, S. J. and Stiglitz, J. E. (1980), 'Stockholder Unanimity in Making Production and Financial Decisions', Quarterly Journal of Economics, 94, 543-566.

Gugler, K., Mueller, D.C. and Yurtoglu, B.B. (2006) 'The Determinants of Merger Waves', University of Vienna: Working Paper.

Harford, J. (2005) 'What drives merger waves?', Journal of Financial Economics, 77, 529560.

Hirshleifer, D. (2001), 'Investor psychology and asset pricing', The Journal of Finance, 56, 1533-1597.

22

Jensen, M. and Ruback, R.S. (1983) 'The Market for Corporate Control: The Scientific Evidence', Journal of Financial Economics, 11, 5-50.

Jiang, G., Lee, C. And Zhang, Y. (2005), 'Information uncertainty and expected returns', Review of Accounting Studies, 10, 185-221.

Jin, L., andMyers, S. (2006), 'R2 around the world: New theory and new tests', Journal of Financial Economics, 79, 257-292.

La Porta, R., Lakonishok, J., Shleifer, A. and Vishny, R. (1997), 'Good news for value stocks: Further evidence on market efficiency', Journal of Finance, 52, 859-874. Limmack, R.J. (1991) 'Corporate Mergers and Shareholder Wealth Effects: 1977-1986', Accounting and Business Research, 21, 239-251.

Merton, R.C. (1987) 'A Simple Model of Capital Market Equilibrium with Incomplete Information', Journal of Finance, 42, 48-510.

Mitchell, M.L. and Mulherin, J.H. (1996) 'The Impact of Industry Shocks on Takeover and Restructuring Activity', Journal of Financial Economics, 41, 193-229.

Moeller, S. B., Schlingemann, F. P. and Stulz, R. M. (2004), 'Firm size and the gains from acquisitions', Journal of Financial Economics, 73, 201-228.

Morck, R., Yeung, B. and Yu, W. (2000), 'The information content of stock markets: why do emerging markets have synchronous stock price movements?', Journal of Financial Economics, 58, 215-260.

Myers, S. C. and Majluf, N. S. (1984), 'Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not Have', Journal of Financial Economics, 13, 187-221.

Odean, T. (1999), 'Do investors trade too much?', American Economic Review, 89, 12791298.

23

Owen, S. (2006) 'The history and mystery of merger waves: A UK and US perspective', The University of New South Wales: Working Paper Number 2006-02.

Rhodes-Kropf, M. and Viswanathan, S. (2004) 'Market Valuation and Merger Waves', The Journal of Finance, 59, 2685-2718.

Roll, R. 'R2,'Journal of Finance, 43, 541-66.

Rosen, R.J. (2006) 'Merger Momentum and Investor Sentiment: The Stock Market Reaction to Merger Announcements', Journal of Business, 79, 987-1017. Shleifer, A. and Vishny, R. W. (2003), 'Stock market driven acquisitions', Journal of Financial Economics, 70, 295-311.

Skinner, D. J., Sloan, R. (1999), 'Earnings surprises, growth expectations and stock returns or Don't let an earnings torpedo sink your portfolio', Unpublished working paper, University of Michigan Business School.

Subrahmanyam, A. and Titman, S. (1999), 'The going-public decision and the development of financial markets', Journal of Finance, 54, 1045-1082.

Travlos, N. G. (1987), 'Corporate Takeover Bids, Methods of Payment, and Bidding Firms Stock Returns', Journal of Finance, 42, 943-963.

Zhang, F. (2006), 'Information Uncertainty and Stock Returns', The Journal of Finance, 61, 105-137.

24

Figure 1
This figure illustrates the five days Cumulative Abnormal Returns for the PrivateCash, PrivateStock, PublicCash and PublicStock portfolios. The first bar of each group presents the overall performance of the portfolio. The second (grey) bar show the performance under high investor sentiment and the third (stripped) bar shows the performance under low investor sentiment. High Sentiment is described as the combination of High information Uncertainty and High probability that investors possess private information. Low Sentiment is described as the combination of Low information Uncertainty and Low probability that investors possess private information. In this graph, Uncertainty is captured by the proxy of Age.
8.00% 6.00% 4.00% 2.00% 0.00% PrivateCash -2.00% -4.00% -6.00% -8.00% PrivateStock PublicCash PublicStock

All High Sentiment Low Sentiment

25

26

Table 1. Summary Statistics by acquisitions by Year The table presents the number of acquisitions by year and the percentage of total number of acquisitions by synchronicity and information uncertainty proxies. The summary statistics are provided on the basis of a sample of 6043 acquisitions from 1985 to 2009 undertaken by 1883 unique bidders. Acquirers are publicly 2traded firms listed on the London Stock Exchange (LSE). Synchronicity is measured as the R of the following regression: ri,j,t= ?i,0+?i,m rm,t + ?i,j rj,t +?i,t where ri,j,t is the return of bidder i in industry j a2t time t, rm,t is the market return at time t and rj,t is the return of industry j at time t. The lowest 33% R firms are classified as low synchronicity, the highest 33% R2 firms as high synchronicity and the rest as medium. Information uncertainty is approached with the proxy of Age. The 33% youngest acquirers are classified as high uncertainty, the 33% oldest as low uncertainty and the medium 33% as of medium uncertainty. Age is measured as the difference between the incorporation date of the firm until the announcement date of the deal. Size is also used as a proxy. The 33% smallest acquirers are classified as high uncertainty, the 33% largest as low uncertainty and the medium 33% as of medium uncertainty. Size is measured as the market capitalization (MV) of the bidding firm 20 days before the announcement date of the deal. For the Sigma proxy, the 33% highest sigma acquirers are classified as high uncertainty, the 33% lowest sigma as low uncertainty and the medium 33% as of medium uncertainty. Sigma is measured by the standard deviation of daily excess returns 200 days before the announcement date of the deal. Finally, descriptive statistics for the Trading Volume proxy which is split as the 33% less active acquirers are classified as high uncertainty, the 33% most active as low uncertainty and the medium 33% as of medium uncertainty. Trading Volume is measured as the average of the monthly trading volume of the acquirer before the announcement date of the deal.

27

Overall Sample
High h Synchr onicity 4 12 38 75 89 62 51 83 114 142 137 73 128 117 106 157 159 107 81 77 115 122 191 114 34 2015 33 .34 Low Synchron icity 2 11 21 56 84 61 60 56 78 88 67 155 157 124 188 203 104 97 74 118 160 172 136 59 34 2015 33 .34

IU by Age
High IU 2 01 39 62 73 50 39 47 61 80 81 87 150 160 181 223 175 130 87 121 177 172 187 91 27 2309 38 .21 Low IU 0 15 19 49 19 16 17 23 36 40 14 99 120 187 162 141 102 79 73 70 100 101 102 71 18 2309 38 .21

IU by Size
High IU 1 211 23 33 171 103 79 93 110 170 203 40 64 145 156 141 143 127 90 118 155 162 164 96 33 2340 38 .72 Low IU 6 0 0 1 1 1 6 20 33 47 82 205 223 163 178 234 138 111 80 96 134 120 180 112 22 2340 38 .72

IU by IU by Trading High SigmLow a HigVolume
IU IU 1 5 01987 48 15 1988 125 35 1989 164 49 1990 97 22 1991 89 22 1992 119 25 1993 159 44 1994 196 56 1995 193 55 1996 239 54 93 133 258 352 226 169 153 80 95 109 105 138 75 2233 36 .95 61 1997 75 103 25 38 23 18 15 107 180 117 156 22 0 2233 36 .95 110 136 124 116 129 85 106 141 157 168 97 34 1741 28 .81 104 125 152 105 91 81 101 126 124 149 105 19 1741 28 .81 IU 0 28 81 96 46 39 50 33 83 44 49 303 70 Low IU 0 7 19 27 17 22 32 63 51 73

Year 1985 1986

All 9 20 11 50 51 39 35 38 50 50 50 109 123

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Total Total (%)

382 459 531 404 325 246 306 405 412 464 269 79 6043 100

28

Table 2. Cumulative Abnormal Returns (CARs) for the Entire Sample This table presents the Cumulative Abnormal Returns (CARs) during five days (-2, +2) surrounding the announcement for the entire sample. Abnormal returns are calculated using a modified marketadjusted model: ARit = Rit - Rmt where Rit is the return on firm i at time t and Rmt is the value-weighted Market Index Return (FT-All Share). All acquirers are publicly traded firms listed on the London Stock Exchange (LSE). The number of bids for each category is reported below the mean return. Significance levels at 1%, 5% and 10% are represented by 'a', 'b' and 'c', respectively. The Dif (1)-(2) represents the differences in mean CARs for the five days (-2, +2) around the acquisition announcement of cash versus stock acquisitions. P-values are reported in brackets.

All N p-value Private N p-value Public N p-value Subsidiary N p-value

All 1 .4 6 %a 7019 (0.000) 1 .6 9 %a 4058 (0.000) -0.46% 713 (0.113) 1 .6 5 %a 2248 (0.000)

Cash (1) 1 .3 0 %a 3199 (0.000) 1 .1 3 %a 1416 (0.000) 0 .9 5 %b 297 (0.012) 1 .5 4 %a 1486 (0.000)

Stock(2) 1 .7 0 %a 544 (0.002) 3 .6 0 %a 248 (0.000) -2.04%a 208 (0.001) 5 .2 0 %a 88 (0.000)

Mixed (3) 1 .5 7 %a 3276 (0.000) 1 .8 2 %a 2394 (0.000) -0.89%c 208 (0.099) 1 .4 4 %a 674 (0.000)

Dif (1)-(2) -0.40% (0.482) -2.47%b (0.012) 2 .9 9 %a (0.000) -3.66%a (0.007)

Table 2B: Correlation Matrix of Proxies Age 1.000 0.362 -0.185 0.363 Size 1.000 -0.191 0.791 Sigma VO

Age Size Sigma VO

1.000 -0.109

1.000

29

Table 3. Cumulative Abnormal Returns (CARs) of High and Low Uncertainty and High and Low Synchronicity Acquirers by Age of the Acquiring Firm This table presents the Cumulative Abnormal Returns (CARs) during five days (-2, +2) surrounding the announcement of high and low information uncertainty acquirers by the age2of the acquirer and high and low synchronicity acquirers. Synchronicity is measured as the R of the following regression: ri,j,t= ?i,0+?i,m rm,t + ?i,j rj,t +?i,t where ri,j,t is the return of bidder i in industry j at time t, rm,t is the market return at time t and rj,t is the return of industry j at time t. Abnormal returns are calculated using a modified market-adjusted model: ARit = Rit - Rmt where Rit is the return on firm i at time t and Rmt is the value-weighted Market Index Return (FT-All Share). All acquirers are publicly traded firms listed on the London Stock Exchange (LSE). The 33% youngest acquirers are classified as high uncertainty, the 33% oldest as low uncertainty and the medium 33% as of medium uncertainty. Age is measured as the difference bet2ween the incorporation date of the firm until the announcement date of the deal. The lowest 33% R firms are classified as low synchronicity, the highest 33% R2 firms as high synchronicity and the rest as medium. Panel A illustrates the gains to acquirers for private target paid for with stock, Panel B for acquisitions for private target paid for with cash, Panel C for acquisitions for public target paid for with stock and Panel D for acquisitions for public target paid for with cash. Cash deals are deals financed with 100% cash and stock deals are deals financed 100% with stock. The Dif [(1)-(2)] at the last row of each panel represents the differences in mean CARs for the five days (-2, +2) around the acquisition announcement of low versus high synchronicity bidders. The Dif (3)-(4)] at the last column of each panel represents the differences in mean CARs for the five days (-2, +2) around the acquisition announcement of high versus low uncertainty bidders. The diagonal differential in each panel represent the difference in mean CARs for the five days (-2, +2) around the acquisition announcement between low synchronicity-high uncertainty versus high synchronicity-low uncertainty bidders. Significance levels at 1%, 5% and 10% are represented by 'a', 'b' and 'c', respectively. P-values are reported in brackets.

30

Panel A: Private Targets paid for with Stock

All
Mean p-value N Mean p-value N 3.80% 0.000 226

HighIA
5.08% 0.005 106

LowIA
1.93% 0.163 37

Dif(H-L)
3.15% 0.160

All
3.80% 0.000 226

HiuLs
6.32% 0.062 36

LiuHs
0.49% 0.807 16

Dif(HiuLs-LiuHs)
5.83% 0.134

Panel B: Private Targets paid for with Cash

All
Mean p-value N Mean p-value N 1.22% 0.000 1201

HighIA
1.64% 0.000 351

LowIA
0.64% 0.003 471

Dif(H-L)
1.00% 0.037

All
1.22% 0.000 1201

HiuLs
2.43% 0.003 131

LiuHs
0.19% 0.517 222

Dif(HiuLs-LiuHs)
2.24% 0.009

Panel C: Public Targets paid for with Stock

All
Mean p-value N Mean p-value N -2.35% 0.001 187

HighIA
-3.87% 0.005 75

LowIA
-0.28% 0.786 52

Dif(H-L)
-3.58% 0.035

All
-2.35% 0.001 187

HiuLs
-5.89% 0.003 24

LiuHs
0.61% 0.618 21

Dif(HiuLs-LiuHs)
-6.50% 0.005

Panel D: Public Targets paid for with Cash

All
Mean p-value N Mean p-value N 1.14% 0.007 253

HighIA
2.09% 0.039 56

LowIA
0.94% 0.065 111

Dif(H-L)
1.15% 0.305

All
1.14% 0.007 253

HiuLs
3.85% 0.085 16

LiuHs
0.80% 0.166 75

Dif(HiuLs-LiuHs)
3.05% 0.176

31

Table 4. Cumulative Abnormal Returns (CARs) of High and Low Uncertainty and High and Low Synchronicity Acquirers by Size of the Acquiring Firm This table presents the Cumulative Abnormal Returns (CARs) during five days (-2, +2) surrounding the announcement of high and low information uncertainty acquirers by the age2of the acquirer and high and low synchronicity acquirers. Synchronicity is measured as the R of the following regression: ri,j,t= ?i,0+?i,m rm,t + ?i,j rj,t +?i,t where ri,j,t is the return of bidder i in industry j at time t, rm,t is the market return at time t and rj,t is the return of industry j at time t. Abnormal returns are calculated using a modified market-adjusted model: ARit = Rit - Rmt where Rit is the return on firm i at time t and Rmt is the value-weighted Market Index Return (FT-All Share). All acquirers are publicly traded firms listed on the London Stock Exchange (LSE). The 33% smallest acquirers are classified as high uncertainty, the 33% largest as low uncertainty and the medium 33% as of medium uncertainty. Size is measured as the market capitalization (MV) of the bidding firm 20 days before the announcement dat 2e of the deal. The lowest 33% R2 firms are classified as low synchronicity, the highest 33% R firms as high synchronicity and the rest as medium. Panel A illustrates the gains to acquirers for private target paid for with stock, Panel B for acquisitions for private target paid for with cash, Panel C for acquisitions for public target paid for with stock and Panel D for acquisitions for public target paid for with cash. Cash deals are deals financed with 100% cash and stock deals are deals financed 100% with stock. The Dif [(1)-(2)] at the last row of each panel represents the differences in mean CARs for the five days (-2, +2) around the acquisition announcement of low versus high synchronicity bidders. The Dif (3)-(4)] at the last column of each panel represents the differences in mean CARs for the five days (-2, +2) around the acquisition announcement of high versus low uncertainty bidders. The diagonal differential in each panel represent the difference in mean CARs for the five days (-2, +2) around the acquisition announcement between low synchronicity-high uncertainty versus high synchronicity-low uncertainty bidders. Significance levels at 1%, 5% and 10% are represented by 'a', 'b' and 'c', respectively. Pvalues are reported in brackets.

32

Panel A: Private Targets paid for with Stock

All
Mean p-value N Mean p-value N 3.80% 0.000 226

HighIU
5.51% 0.002 126

LowIU
1.44% 0.265 45

Dif(H-L)
4.06% 0.057

All
3.80% 0.000 226

HiuLs
6.57% 0.011 59

LiuHs
0.76% 0.686 29

Dif(HiuLs-LiuHs)
5.81% 0.065

Panel B: Private Targets paid for with Cash

All
Mean p-value N Mean p-value N 1.22% 0.000 1201

HighIU
2.66% 0.000 313

LowIU
0.46% 0.056 480

Dif(H-L)
2.20% 0.000

All
1.22% 0.000 1201

HiuLs
3.02% 0.000 156

LiuHs
0.50% 0.100 293

Dif(HiuLs-LiuHs)
2.52% 0.005

Panel C: Public Targets paid for with Stock

All
Mean p-value N Mean p-value N -2.35% 0.001 187

HighIU
-3.82% 0.005 71

LowIU
-0.74% 0.569 52

Dif(H-L)
-3.08% 0.099

All
-2.35% 0.001 187

HiuLs
-4.10% 0.035 29

LiuHs
-1.08% 0.482 32

Dif(HiuLs-LiuHs)
-3.03% 0.210

Panel D: Public Targets paid for with Cash

All
Mean p-value N Mean p-value N 1.14% 0.007 253

HighIU
3.25% 0.024 31

LowIU
0.73% 0.150 173

Dif(H-L)
2.52% 0.091

All
1.14% 0.007 253

HiuLs
4.23% 0.044 13

LiuHs
1.12% 0.038 118

Dif(HiuLs-LiuHs)
3.11% 0.135

33

Table 5. Cumulative Abnormal Returns (CARs) of High and Low Uncertainty and High and Low Synchronicity Acquirers by Sigma of the Acquiring Firm This table presents the Cumulative Abnormal Returns (CARs) during five days (-2, +2) surrounding the announcement of high and low information uncertainty acquirers by the age2of the acquirer and high and low synchronicity acquirers. Synchronicity is measured as the R of the following regression: ri,j,t= ?i,0+?i,m rm,t + ?i,j rj,t +?i,t where ri,j,t is the return of bidder i in industry j at time t, rm,t is the market return at time t and rj,t is the return of industry j at time t. Abnormal returns are calculated using a modified market-adjusted model: ARit = Rit - Rmt where Rit is the return on firm i at time t and Rmt is the value-weighted Market Index Return (FT-All Share). All acquirers are publicly traded firms listed on the London Stock Exchange (LSE). The 33% highest sigma acquirers are classified as high uncertainty, the 33% lowest sigma as low uncertainty and the medium 33% as of medium uncertainty. Sigma is measured by the standard deviation of daily excess returns 200 days before the announcement date of the deal. The lowest 33% R2 firms are classified as low synchronicity, the highest 33% R2 firms as high synchronicity and the rest as medium. Panel A illustrates the gains to acquirers for private target paid for with stock, Panel B for acquisitions for private target paid for with cash, Panel C for acquisitions for public target paid for with stock and Panel D for acquisitions for public target paid for with cash. Cash deals are deals financed with 100% cash and stock deals are deals financed 100% with stock. The Dif [(1)-(2)] at the last row of each panel represents the differences in mean CARs for the five days (-2, +2) around the acquisition announcement of low versus high synchronicity bidders. The Dif (3)-(4)] at the last column of each panel represents the differences in mean CARs for the five days (-2, +2) around the acquisition announcement of high versus low uncertainty bidders. The diagonal differential in each panel represent the difference in mean CARs for the five days (-2, +2) around the acquisition announcement between low synchronicity-high uncertainty versus high synchronicity-low uncertainty bidders. Significance levels at 1%, 5% and 10% are represented by 'a', 'b' and 'c', respectively. Pvalues are reported in brackets.

34

Panel A: Private Targets paid for with Stock

All
Mean p-value N Mean p-value N 4.04% 0.000 219

HighIU
4.30% 0.006 131

LowIU
3.04% 0.134 51

Dif(H-L)
1.26% 0.620

All
4.04% 0.000 219

HiuLs
6.57% 0.046 45

LiuHs
0.03% 0.982 19

Dif(HiuLs-LiuHs)
6.54% 0.063

Panel B: Private Targets paid for with Cash

All
Mean p-value N Mean p-value N 1.17% 0.000 1168

HighIU
1.77% 0.000 360

LowIU
0.74% 0.000 379

Dif(H-L)
1.03% 0.054

All
1.17% 0.000 1168

HiuLs
2.75% 0.014 116

LiuHs
0.58% 0.023 158

Dif(HiuLs-LiuHs)
2.17% 0.058

Panel C: Public Targets paid for with Stock

All
Mean p-value N Mean p-value N -2.43% 0.001 181

HighIU
-4.69% 0.001 80

LowIU
-0.34% 0.712 48

Dif(H-L)
-4.35% 0.007

All
-2.43% 0.001 181

HiuLs
-5.52% 0.008 26

LiuHs
0.15% 0.941 13

Dif(HiuLs-LiuHs)
-5.67% 0.051

Panel D: Public Targets paid for with Cash

All
Mean p-value N Mean p-value N 1.15% 0.008 250

HighIU
2.15% 0.041 80

LowIU
0.48% 0.305 90

Dif(H-L)
1.67% 0.145

All
1.15% 0.008 250

HiuLs
5.14% 0.045 17

LiuHs
0.46% 0.300 58

Dif(HiuLs-LiuHs)
4.69% 0.068

35

Table 6. Cumulative Abnormal Returns (CARs) of High and Low Uncertainty and High and Low Synchronicity Acquirers by Trading Volume of the Acquiring Firm This table presents the Cumulative Abnormal Returns (CARs) during five days (-2, +2) surrounding the announcement of high and low information uncertainty acquirers by the age2of the acquirer and high and low synchronicity acquirers. Synchronicity is measured as the R of the following regression: ri,j,t= ?i,0+?i,m rm,t + ?i,j rj,t +?i,t where ri,j,t is the return of bidder i in industry j at time t, rm,t is the market return at time t and rj,t is the return of industry j at time t. Abnormal returns are calculated using a modified market-adjusted model: ARit = Rit - Rmt where Rit is the return on firm i at time t and Rmt is the value-weighted Market Index Return (FT-All Share). All acquirers are publicly traded firms listed on the London Stock Exchange (LSE). The 33% less active acquirers are classified as high uncertainty, the 33% most active as low uncertainty and the medium 33% as of medium uncertainty. Trading Volume is measured as the average of the monthly trading volume of the acquirer before the announcement date of the deal. The lowest 33% R2 firms are classified as low synchronicity, the highest 33% R2 firms as high synchronicity and the rest as medium. Panel A illustrates the gains to acquirers for private target paid for with stock, Panel B for acquisitions for private target paid for with cash, Panel C for acquisitions for public target paid for with stock and Panel D for acquisitions for public target paid for with cash. Cash deals are deals financed with 100% cash and stock deals are deals financed 100% with stock. The Dif [(1)-(2)] at the last row of each panel represents the differences in mean CARs for the five days (-2, +2) around the acquisition announcement of low versus high synchronicity bidders. The Dif (3)-(4)] at the last column of each panel represents the differences in mean CARs for the five days (-2, +2) around the acquisition announcement of high versus low uncertainty bidders. The diagonal differential in each panel represent the difference in mean CARs for the five days (-2, +2) around the acquisition announcement between low synchronicity-high uncertainty versus high synchronicity-low uncertainty bidders. Significance levels at 1%, 5% and 10% are represented by 'a', 'b' and 'c', respectively. Pvalues are reported in brackets.

36

Panel A: Private Targets paid for with Stock

All
Mean p-value N Mean p-value N 4.02% 0.002 162

HighIU
3.77% 0.049 87

LowIU
3.34% 0.060 35

Dif(H-L)
0.43% 0.868

All
4.02% 0.002 162

HiuLs
4.23% 0.103 38

LiuHs
1.27% 0.639 18

Dif(HiuLs-LiuHs)
2.96% 0.423

Panel B: Private Targets paid for with Cash

All
Mean p-value N Mean p-value N 1.27% 0.000 977

HighIU
2.43% 0.000 284

LowIU
0.57% 0.028 369

Dif(H-L)
1.87% 0.001

All
1.27% 0.000 977

HiuLs
2.95% 0.000 137

LiuHs
0.44% 0.187 240

Dif(HiuLs-LiuHs)
2.52% 0.003

Panel C: Public Targets paid for with Stock

All
Mean p-value N Mean p-value N -2.79% 0.002 130

HighIU
-6.85% 0.001 40

LowIU
0.01% 0.994 46

Dif(H-L)
-6.86% 0.005

All
-2.79% 0.002 130

HiuLs
-6.48% 0.012 18

LiuHs
-1.33% 0.451 27

Dif(HiuLs-LiuHs)
-5.14% 0.085

Panel D: Public Targets paid for with Cash

All
Mean p-value N Mean p-value N 1.12% 0.017 217

HighIU
3.14% 0.038 36

LowIU
0.76% 0.184 129

Dif(H-L)
2.38% 0.135

All
1.12% 0.017 217

HiuLs
2.29% 0.338 14

LiuHs
1.12% 0.084 90

Dif(HiuLs-LiuHs)
1.17% 0.632

37

Table 7. Regressions of CARs on Information Uncertainty, Synchronicity and Deal Features This table presents regression estimates of the acquirer's five-day cumulative abnormal return controlling for information uncertainty and synchronicity of the bidder's stock price. In Panel A, the 33% youngest acquirers are classified as high uncertainty, the 33% oldest as low uncertainty and the medium 33% as of medium uncertainty. Age is measured as the difference between the incorporation date of the firm until the announcement date of the deal. In Panel B, the 33% smallest acquirers are classified as high uncertainty, the 33% largest as low uncertainty and the medium 33% as of medium uncertainty. Size is measured as the market capitalization (MV) of the bidding firm 20 days before the announcement date of the deal. In Panel C, the 33% highest sigma acquirers are classified as high uncertainty, the 33% lowest sigma as low uncertainty and the medium 33% as of medium uncertainty. Sigma is measured by the standard deviation of daily excess returns 200 days before the announcement date of the deal. In Panel D, the 33% less active acquirers are classified as high uncertainty, the 33% most active as low uncertainty and the medium 33% as of medium uncertainty. Trading Volume is measured as the average of the monthly trading volume of the acquirer before the announcement date of the deal. Synchronicity is measured as the R2 of the following regression: ri,j,t= ?i,0 +?i,m rm,t + ?i,j rj,t +?i,t where ri,j,t is the return of bidder i in industry j a2t time t, rm,t is the market return at time t and rj,t is the return of industry j at time t. The lowest 33% R firms are classified as low synchronicity, the highest 33% R2 firms as high synchronicity and the rest as medium. HighIU dummy takes the value of 1 of the bid was announced by a high information uncertainty bidder according to the four proxies, and zero otherwise. The HsHiu, HsLiu, LsHiu, LsLiu takes the value of 1 is the deal belong to the high (low) information uncertainty (synchronicity) group respectively. Diversifying deals is a dummy that takes the value of 1 when the acquirer's two-digit SIC code is different from that of the target and 0 otherwise. Bidder's market-to-book is measured by the bidder's market value a month before the announcement of the deal divided by its net book value of assets; a deal's relative size is the ratio between target and bidder size. Domestic deals dummy takes the value of 1 for acquisitions of UK firms and zero otherwise. Finally, other explanatory variables include: the acquirer's lagged excess return for 180 days prior to the bid's announcement; and the market portfolio return (FT-All Share) for the same 180-day period prior to the announcement. Pvalues are reported in square brackets under the coefficients. Significance levels at 1%, 5% and 10% are represented by 'a', 'b' and 'c', respectively.

38

Panel A: Age CARs HighIU HsHiu HsLiu (Low Sentiment) LsHiu (High Sentiment) (0.403) LsLiu M/B Relative Size Domestic deals Diversifying FTALLASH(-180,-3) Ri -Rm(-180,-3) Intercept N Adj. R2 % -0.001 (0.138) -0.001 (0.226) -0.006 (0.773) -0.001 (0.976) 0 .101 (0.141) 0 .000 (0.992) 0 .034 (0.104) 222 4.58% -0.001 (0.131) -0.002 (0.379) -0.007 (0.741) 0 .001 (0.967) 0 .092 (0.199) 0 .001 (0.913) 0 .0 4 4b (0.016) 217 4.23% -0.001 (0.108) -0.002 (0.352) -0.009 (0.684) -0.001 (0.981) 0 .085 (0.242) 0 .000 (0.968) 0 .0 5 6a (0.006) 217 4.33% (1 ) 0 .031 (0.162) PrivateStock (2 ) (3 ) (4 ) (5 ) -0.024 (0.146) PublicStock (6 ) (7 ) (8 ) (1 ) 0 .0 5 1b (0.015) PrivateStock (2 ) (3 )

Panel B: Size (4 ) (5 ) -0.02 (0.165) PublicStock (6 ) (7 ) (8 )

-0.013 (0.756) -0.048c -0.046c (0.053) (0.073)
b b

0 .006 (0.869) 0 .0 2 9c 0 .025

0 .058 (0.331) -0.047b -0.032 (0.039) (0.170) -0.049 (0.231) 0 .035 (0.224) 0 .063 (0.165) -0.001c (0.094) -0.002 (0.327) -0.014 (0.534) -0.002 (0.913) 0.08 (0.266) 0 .001 (0.887) 0 .0 6 3a (0.003) 217 4.70% -0.001c (0.056) -0.002 (0.249) -0.017 (0.432) -0.004 (0.869) 0 .075 (0.298) 0 .001 (0.826) 0 .0 5 1b (0.012) 217 6.29% -0.003b (0.022) -0.004 (0.306) -0.038 (0.113) 0 .003 (0.850) 0 .1 4 6a (0.008) -0.002 (0.479) 0 .023 (0.305) 178 12.87% 0 .020

-0.022 (0.660) 0 .014

(0.057) (0.106)
c0 .0 3 2

(0.235) (0.409) 0 .037 -0.035 (0.098) -0.032 (0.137) 0 .022 (0.567) -0.003b (0.035) -0.005 (0.266) -0.039 (0.101) 0 .000 (0.991) 0 .1 5 6a (0.006) -0.003b (0.020) -0.005 (0.235) -0.041c (0.093) 0 .003 (0.861) 0 .1 4 8a (0.004)

0 .026 -0.054 (0.026) 0 .011 (0.705)

(0.513) -0.012 (0.630) -0.001 (0.116) -0.002 (0.348) -0.009 (0.691) 0 .001 (0.952) 0 .089 (0.216) 0 .000 (0.928) 0 .0 5 3a (0.010) 217 4.80% -0.002c (0.078) -0.005 (0.232) -0.039 (0.114) -0.001 (0.948) 0 .1 5 7a (0.006) -0.001 (0.840) 0 .025 (0.285) 178 13.34%

(0.015)

-0.002c (0.086) -0.005 (0.257) -0.040c (0.089) -0.005 (0.735) 0 .1 7 2a (0.002) 0 .000 (0.923) 0 .026 (0.271) 177 15.41%

-0.002b (0.029) -0.005 (0.268) -0.038 (0.112) 0 .000 (0.998) 0 .1 5 4a (0.007) -0.002 (0.754) 0 .013 (0.574) 177 13.04%

-0.002 (0.106) -0.005 (0.281) -0.040c (0.098) -0.005 (0.715) 0 .1 7 3a (0.001) 0 .000 (0.997) 0.02 (0.389) 177 16.15%

-0.001 (0.200) -0.002 (0.134) -0.018 (0.398) -0.004 (0.852) 0 .085 (0.222) 0 .001 (0.852) 0 .031 (0.122) 222 5.97%

-0.001 (0.148) -0.002 (0.364) -0.01 (0.649) -0.001 (0.948) 0 .081 (0.260) 0 .000 (0.926) 0 .0 4 4b (0.024) 217 4.64%

-0.003b (0.014) -0.005 (0.236) -0.044c (0.070) 0 .002 (0.890) 0 .1 5 7a (0.005) -0.002 (0.353) 0 .028 (0.243) 177 13.89%

-0.002 -0.002 (0.670) (0.593) 0 .014 0 .023 (0.545) (0.327) 177 177 12.71% 14.81

39

Table 7-continued

Panel C: Sigma CARs HighIU HsHiu HsLiu (Low Sentiment) LsHiu (High Sentiment) LsLiu M/B Relative Size Domestic deals Diversifying FTALLASH(-180,-3) Ri -Rm(-180,-3) Intercept N Adj. R2 % -0.001 (0.101) -0.001 (0.250) -0.005 (0.819) -0.004 (0.858) 0 .106 (0.168) 0 .000 (0.998) 0 .036 (0.137) 222 4.10% -0.001c (0.080) -0.002 (0.369) -0.006 (0.782) -0.005 (0.823) 0 .104 (0.173) 0 .001 (0.919) 0 .0 4 3b (0.029) 217 4.91% -0.001 (0.103) -0.002 (0.360) -0.009 (0.686) -0.003 (0.880) 0 .086 (0.234) 0 .000 (0.977) 0 .0 5 9a (0.005) 217 4.60% 0 .044 (0.240) -0.053a (1 ) 0 .023 (0.342) PrivateStock (2 ) (3 ) (4 ) (5 ) -0.032b (0.044) PublicStock (6 ) (7 ) (8 ) (1 ) 0 .013 (0.659)

Panel D: Trading Volume PrivateStock (2 ) (3 ) (4 ) (5 ) -0.067a (0.001) PublicStock (6 ) (7 ) (8 )

0 .003 (0.931) -0.045b 0 .015

-0.004 (0.906) 0 .008

-0.047 (0.306) -0.049 -0.053 (0.113) (0.117) -0.004 (0.905) 0 .010 (0.843) -0.001c (0.074) -0.002 (0.295) -0.017 (0.520) -0.013 (0.622) 0 .119 (0.179) 0 .000 (0.932) 0 .0 7 5a (0.002) 158 6.27% -0.001c (0.080) -0.002 (0.421) -0.016 (0.586) -0.014 (0.626) 0 .123 (0.175) 0 .000 (0.928) 0 .0 7 8a (0.002) 158 6.79% -0.005a (0.001) -0.005 (0.228) -0.051b (0.046) -0.003 (0.853) 0 .1 7 2a (0.010) -0.002 (0.772) 0 .0 5 4b (0.038) 125 23.67% 0 .033

-0.091 (0.203) 0 .021

(0.002) (0.015) 0 .039 (0.314) -0.011 (0.703) -0.001c (0.074) -0.002 (0.360) -0.007 (0.770) -0.006 (0.777) 0 .108 (0.162) 0 .000 (0.965) 0 .0 5 0b (0.023) 217 5.57% -0.002b (0.045) -0.005 (0.282) -0.047c (0.055) 0 .003 (0.855) 0 .1 1 7b (0.040) -0.001 (0.951) 0 .038 (0.112) 177 14.36%

-0.042c (0.061)

(0.526) (0.718) -0.042c (0.058) -0.006 (0.786) -0.003b (0.027) -0.005 (0.257) -0.040c (0.093) -0.001 (0.970) 0 .1 5 1a (0.008) -0.002 (0.680) 0 .018 (0.424) 176 12.29% -0.003c (0.067) -0.006 (0.240) -0.045c (0.056) -0.004 (0.773) 0 .1 5 2a (0.008) -0.001 (0.871) 0 .032 (0.179) 176 14.60% -0.001c (0.080) -0.001 (0.189) -0.02 (0.491) -0.013 (0.618) 0 .112 (0.181) 0 .000 (0.990) 0 .0 6 3a (0.008) 162 5.38%

0 .008 (0.813)

-0.064b (0.011)

(0.109) (0.300) -0.060b (0.021) 0 .048 (0.340) -0.005a (0.001) -0.005 (0.361) -0.056b (0.042) 0 .000 (0.992) 0 .1 8 0b (0.018) -0.005a (0.000) -0.005 (0.315) -0.058b (0.034) -0.005 (0.794) 0 .1 6 3b (0.023)

-0.003c (0.063) -0.006 (0.241) -0.045c (0.057) -0.004 (0.774) 0 .1 5 6a (0.005) -0.001 (0.857) 0 .031 (0.185) 176 14.49%

-0.001c (0.084) -0.002 (0.330) -0.015 (0.590) -0.013 (0.645) 0 .115 (0.202) 0 .000 (0.981) 0 .0 6 5a (0.004) 158 5.44%

-0.005a (0.000) -0.005 (0.318) -0.0603b (0.019) 0 .000 (0.990) 0 .1 8 9a (0.009) -0.002 (0.271) 0 .0 5 2c (0.051) 125 19.49%

-0.001 -0.004 (0.802) (0.281) 0 .029 0 .0 4 7c (0.283) (0.068) 125 125 16.54% 24.12

40

41



doc_553572350.docx