The probability of informed trading and mergers and acquisitions

2.1 PIN and information asymmetry

The notion of the probability of informed trading stems from the market microstructure literature, where investors are viewed as informed or uninformed. Although uninformed traders are unaware of the specific information possessed by informed traders, they realise that such information influences the trades of informed traders, thereby attaching information content to the composition of trades. Thus, an imbalance of buy or sell orders leads uninformed investors to update their beliefs and eventually cause market prices to converge to values based on the new information. Easley et al.(2002) examine the role of asymmetric information across investors and show that PIN is positively associated with average asset returns. Easley and O’Hara (2004) further suggest that in equilibrium uninformed traders require a higher rate of return when the probability of informed trading is higher, leading to a positive association between PINs and returns. Chen et al.(2007) show that PIN is positively associated with the sensitivity of firm-level investment to stock prices, supporting the view that managers learn from private information that is incorporated into stock trades. Ferreira and Laux (2007) document a positive association between PIN and strong corporate governance, and interpret the results as evidence that enhanced corporate governance leads to private information collection and informed trading by market participants.

Traditionally, the corporate finance literature commonly measures a firm’s degree of information asymmetry according to firm characteristics such as its size, growth opportunities, or tangibility of its assets. PIN differs from these measures and has its unique advantages in several aspects. First and theoretically, the PIN measure is based on a structural market microstructural model, where order imbalance increases among buy and sell orders with informed trading. It is reasonable to believe that market players in close touch with a firm and its business possess better information about the firm and trade based on this superior information. Market microstructure research models the trading behaviour of market players and analyses the information asymmetry about the payoffs of a firm’s securities. In such respect, information asymmetry measures derived from market microstructure research such as PIN have strong theoretical foundation, which accordingly enables researchers to directly capture the degree of information asymmetry using observed trading data, compared to common firm-level measures.

Second, common firm-level measures are often found to inadequately proxy for the degree of information asymmetry between insiders and other market players. For example, Frankel and Li (2004) and Huddart and Ke (2007) show that insider trading activity and profits can hardly be explained by analyst coverage, forecast dispersion and institutional ownership. In addition, these measures tend to have several, sometimes conflictual, interpretations. For instance, analyst forecast dispersion is found to be a superior proxy for differences in opinion rather than information asymmetry (Pasquariello and Vega, 2007; Sadka and Scherbina, 2007), while firm size can represent a firm’s financial attributes, regulatory status or organisational complexity (Dang et al., 2018), which are all substantially distinguished from information asymmetry. On the other hand, the PIN measure is grounded on market microstructure models and is instead designed to capture financial market participants’ time-varying perception of the information advantage held by firm insiders (Bharath et al., 2009). In Section 3.4, we show that PIN has a low correlation with common firm-level measures of information asymmetry, highlighting the empirical difference between PIN and other information asymmetry measures.

Finally, even compared to other microstructure-based measures of information asymmetry such as the components of bid-ask spreads, PIN is arguably a superior proxy. For example, bid-ask spreads tend to capture not only adverse selection costs but also inventory holding costs. Bharath et al.(2009) compare seven different measures of information asymmetry, including the adverse selection component of both quoted and effective bid-ask spreads, PIN, the interaction between stock returns and trading volume (Llorente et al., 2002) and liquidity measures. They find that their composite index of information asymmetry is most highly associated with PIN (coefficient = 0.51), while the correlation coefficients for other information asymmetry measures are all below 0.30.

2.2 The choice of payment method in mergers and acquisitions

The choice of payment method in M&As has been examined extensively. Prior literature documents that the choice of payment method can be explained by differential tax treatment (Gilson et al., 1988; Brown and Ryngaert, 1991), information asymmetry (Myers and Majluf, 1984; Hansen, 1987), capital structure and corporate control motives (Jensen, 1986; Stulz, 1988) and behavioural arguments (Shleifer and Vishny, 2003; Rhodes-Kropf and Viswanathan, 2004).

Accordingly, a number of firm characteristics such as firm size, asset tangibility, growth opportunities, financial leverage and stock price run-up have been found to explain the choice of payment method. For example, larger firms are likely to have better access to debt markets, making cash financing more feasible (Faccio and Masulis, 2005). In addition, firms with more tangible assets have a lower cost of debt which makes cash financing more attractive (Myers, 1977), whilst managers of firms with valuable investment opportunities prefer to finance investments with stock because it allows them to retain valuable cash resources (Jung et al., 1996; Martin, 1996). Bidders with higher financial leverage are likely to find it difficult to issue further debt which therefore increases the likelihood of stock payment (DeAngelo and Masulis, 1980; Faccio and Masulis, 2005), and bidders with a recent stock price run-up prefer to finance acquisitions with equity (Myers and Majluf, 1984; Hansen, 1987).

2.3 PIN, information asymmetry and the choice of payment method

In this paper, our focus is on the role PIN plays in M&As. Before a deal occurs, both the bidder and target are asymmetrically informed about the true value of their respective firm, as one would expect managers to have superior knowledge of their own firm’s value compared with an outsider. In this case, the choice of payment method is likely to reveal information about the over/undervaluation of the firm and affect the division of synergy gains. A key distinction between a cash deal and a stock deal is that the value of a stock deal depends on the cash flows of the combined firm which in turn is driven by the ‘true’ value of the bidder, the target and any synergy gains. However, the value of a cash deal is independent of these parameters.

In contrast to the relation between acquiring firms’ information asymmetry and payment method, two opposing arguments have been documented in regard to the relation between target firms’ information asymmetry and the method of payment. On the one hand, Hansen (1987) argues that bidders who plan to acquire targets with higher information risk are less likely to pay cash because cash payment increases the risk of overpaying. In consequence these bidders are more likely to use equity financing as it enables the bidder to share the target’s information risk with the target firm shareholders as the value of a stock offer depends on the cash flows of the acquirer, the target and any synergy gains. Empirically, Raman et al.(2013) use the quality of financial reporting as a proxy for information risk and find, consistent with Hansen’s (1987) argument, that bidders tend to share the information risk with target shareholders by using more equity financing to acquire targets with poor financial reporting quality (i.e., high information risk).

On the other hand, Chemmanur et al.(2009) argue and find that the choice of payment method in M&As is driven by the trade-off between the cost of overpayment (higher in a cash offer) and the likelihood of bid success (higher in a cash offer). In the presence of rival bidders who also face such a trade-off, a cash offer by a bidder signals to potential rival bidders that its private valuation of the target is higher, thereby helping to deter competition from other rival bidders and hence increasing the probability of success. Furthermore, the advantage of deterring competition is greater when the level of the target’s information asymmetry faced by rival bidders is higher. As a result, Chemmanur et al.(2009) find that bidders are more likely to use cash financing rather than stock financing when they face a high level of information asymmetry in evaluating the target.

2.4 Acquirer PIN and acquirer announcement returns

Acquisitions may provide firms with potential benefits including economics of scale, vertical integration and synergies. However, prior literature has generally found that acquisitions are at best wealth neutral for acquiring-firm shareholders, and potentially wealth destroying upon deal announcement (Andrade et al., 2001; Moeller et al., 2005). For example, Betton et al.(2008) show that the average cumulative abnormal return (CAR) around the announcement of an M&A deal is close to zero. The worst-case scenario is when large bidders make acquisitions of a public target in an all equity deal.

One traditional explanation for the negative bidder announcement returns is information asymmetry. Myers and Majluf (1984) and Travlos (1987) suggest that by making equity financed bids, managers are likely to provide a signal to the market that their firm’s common stock is overvalued. Consistent with this prediction, Chang (1998) and Fuller et al.(2002) show that equity offers are associated with poorer acquirer returns, while Moeller et al.(2007) document that acquirers with higher information asymmetry experience higher announcement returns in cash-financed deals.

2.5 Target firm announcement returns and PIN

On the flip side, prior literature provides unambiguous evidence that target firm shareholders are generally the winners in M&A transactions. Andrade et al.(2001) and Betton et al.(2008) report that the average announcement abnormal returns for target firms are significantly positive at around 15 percent, and are approximately 20 percent for all-cash deals. The well-documented ‘winner’s curse’ is one of the main reasons for the positive abnormal announcements returns for target firms. Bazerman and Samuelson (1983) suggest that the bidder who most overestimates the value of the target firm will be the winning bidder, and the failure to discount bids in response to greater uncertainty increases the magnitude of the winner’s curse. McNichols and Stubben (2015) find empirical evidence to show that targets with higher information risk (a lower quality of accounting information) are more difficult to value and are more likely to result in overvaluation by acquirers.

3.1 The sample

We obtain our sample of US mergers and acquisitions from the Thomson Financial SDC M&A Database. Our sample meets the following criteria: (i) observations are from the period 1994–2011; (ii) deals are completed; (iii) the method of payment is known; (iv) the transaction is greater than US$1 million and at least 1 percent of the acquirer’s market value of equity; (v) the acquirer and target firms are not from the financial or utility industry and have financial statement information and stock return data available on Compustat and CRSP respectively; (vi) the PIN measure for the acquirer and target is available. Our final sample includes 1,724 mergers and acquisitions made by 1,235 acquirers. The definitions of all variables are provided in Table 1.

Table 2 reports the distribution of the mergers and acquisitions by year and industry. Panel A shows that the number of M&A deals dropped significantly in 2001–2002 and 2009, at the end of the tech-boom and the global financial crisis periods respectively. Panel B shows that a majority of the acquirers are from the manufacturing, services, and transportation and communications industries.

3.2 Probability of informed trading

The yearly PIN measure provided by Stephen Brown is computed using the Venter and de Jongh (2006) model and is an extended version of the popular EKO market microstructure model (Easley et al., 1997). PIN is a firm-specific estimate of the probability that a trade originates from a privately informed investor and therefore it directly captures the amount of PINs among investors in the secondary market. The extension by Venter and de Jongh (2006) improves the model’s ability to fit data in the real world by allowing for a strong positive correlation between buy and sell orders, which in turn accommodates the arrival of public information in the market without necessarily attributing it to trading on private information. Overall, PINs computed using the Venter and de Jongh (2006) model better capture the asymmetric information component of PIN as opposed to the liquidity component of PIN as argued by Duarte and Young (2009). Nevertheless, we control for firm liquidity using the Amihud (2002) illiquidity measure in all of our regressions to ensure that our results are not driven by liquidity captured in the PIN component.

3.3 Summary statistics

Table 3 provides summary statistics. Yearly PINs and firm characteristics are all measured at the fiscal year-end prior to the acquisition announcement. Panel A reports that the average acquirer and target PIN are 0.14 and 0.17 respectively, and on average deals are funded by 68 percent cash (Cashper). Panel B shows that the average announcement-period acquirer CARs are ~0 percent while the average CARs for the targets are much higher at around 14 percent, which is consistent with the findings in prior literature (Dong et al., 2006; Ahern, 2012). Panels C and D show that the average acquirer is larger and has stock that is more liquid than the average target firm. Lastly Panel E shows that within the sample, 23 percent of M&As are tender offers and 18 percent are cross-industry deals.

3.4 Correlations between PIN, firm characteristics and information asymmetry

The key contribution of our paper is in examining the role of PIN in M&As, as there is already a battery of papers that document the association between commonly used information asymmetry measures and M&A outcomes. To confirm that PINs differ systematically from the measures of information asymmetry commonly used in the literature, we present a correlation matrix between PINs, firm characteristics and several measures of information asymmetry for both the acquirer and the target firms. Specifically, we use tangibility of a firm’s assets, the dispersion of analyst forecasts (Moeller et al., 2007; Chemmanur et al., 2009), analyst forecast errors (Chemmanur et al., 2009), earnings quality (McNichols and Stubben, 2015), idiosyncratic volatility and the number of large shareholders (Moeller et al., 2007) as explicit measures of information asymmetry. Table 4 shows that firms with higher PINs are smaller and have less liquid stocks and a lower valuation. More importantly, consistent with our expectations, the results show that PIN has a very low correlation with common measures of information asymmetry (<0.09), suggesting that PIN differs systematically from the information asymmetry measures commonly used in the M&A literature.

4.1 Results on PIN and method of payment

Table 5 reports the regression results of the relationship between acquirer PIN, target PIN, and the method of payment in M&As, controlling for various acquirer and target firm characteristics and deal characteristics. To capture the incremental impact of PIN on M&A outcomes, we control for several measures of information asymmetry for both the acquirer and the target. Specifically, we control for the difference between acquirer’s and target’s dispersion of analyst forecasts (Moeller et al., 2007; Chemmanur et al., 2009), analyst forecast errors (Chemmanur et al., 2009), earnings quality measured by abnormal accruals (McNichols and Stubben, 2015), idiosyncratic volatility and the number of large shareholders (Moeller et al., 2007). To ensure that stock illiquidity is not driving our results, we also include the Amihud (2002) illiquidity measure as a control variable. In addition, due to the presence of serial acquirers in our sample, the residuals in our regressions can be correlated and therefore may overstate the t-statistics (Peterson, 2009). To control for this potential issue, we cluster standard errors by acquiring firm in all our regressions. We also control for time and industry fixed effects in all of our regression models.

We start with a two-boundary Tobit model that regresses the percentage of cash financing on acquirer PIN and target PIN controlling for various acquirer and target firm characteristics. The results in column (1) show that acquirers with higher PINs use a higher percentage of cash in financing their M&A deals, and also acquirers tend to use more equity financing when acquiring targets with higher PINs. Both of these relationships are statistically significant at the 1 percent level.

The coefficients of firm-level control variables exhibit the expected signs. Larger firms tend to use more cash due to their better access to debt financing (Faccio and Masulis, 2005); firms with more tangible assets have a lower cost of debt, which makes cash financing more attractive (Myers, 1977); firms with more investment opportunities prefer to use equity financing because it allows them to have more discretion over their internal capital (Martin, 1996); highly leverage firms are more constrained in their ability to issue debt and as a consequence are more likely use stock financing (Faccio and Masulis, 2005); bidders prefer to finance with stock when their equity experiences a recent stock price run-up and has a higher market valuation (Myers and Majluf, 1984; Hansen, 1987); and highly leveraged targets have higher level of information risk and hence bidders are more likely to use equity financing to share the target’s high information risk (Petacchi, 2015). Furthermore, we see that all of the variables capturing the difference in information asymmetry between the acquirer and target are positively related (at the 1 percent level) to the percentage of cash financing, which is consistent with the view that acquirers with higher information asymmetry are more likely to use cash financing due to their high cost of equity.

Next, we include deal-level controls in column (2). The results show that the coefficient estimates of acquirer PIN and target PIN continue to be significant at the 1 percent level. The coefficients of our deal-level control variables also exhibit the expected signs. Acquirers tend to use cash as the method of payment in tender offers in order to avoid long delays as tender offers involving stock as the financing method must be made in accordance with the Securities Act of 1933 (Gilson, 1986; Fishman, 1989); bidders acquiring large targets are more likely to use equity financing to share the information risk as information asymmetry is likely to rise as target assets rise in value relative to those of a bidder (Hansen, 1987); and bidders are more likely to use equity financing to acquire targets in a different industry to share the target’s information risk due to their lack of knowledge in the target industry.

For robustness, we follow Faccio and Masulis (2005) and use an ordered Probit model to examine the association between acquirer and target PINs and the method of payment in M&As. The dependent variable takes the value of 1 for pure stock deals, 2 for mixed stock and cash and 3 for all-cash deals. The results are reported in columns (3) and (4). Consistent with the results in columns (1) and (2), we show that acquirers with higher PINs are more likely to use cash financing, and acquirers are more likely to use equity financing when acquiring targets with higher PINs. Overall, the results in Table 5 confirm the first and second hypotheses that acquirers with a higher PIN are more likely to use cash financing as cash is less sensitive to private information than stock, and bidders use more equity financing to acquire targets with a higher PIN to share the target’s information risk with the target’s shareholders as suggested by Hansen (1987).

4.2 Results on PIN and acquirer announcement returns

Next, we examine the relation between PIN and acquirer announcement returns. To calculate acquirer abnormal returns around the deal announcement, we follow Masulis et al.(2007) by first estimating the market model for each acquirer over a 200-day period ending 11 days before the announcement date (−210, −11) with the CRSP value-weighted return used as the reference market return. We then use the estimated parameters to calculate the cumulative abnormal returns (CARs) over a five-day (−2, 2) event window centred on the announcement date.

The results in Table 5 suggest that the choice of method of payment in takeovers is non-random, and are associated with bidder and target firm characteristics. In addition, acquisition decisions, like many other firm decisions, are also non-random and influenced by managerial expectations and firm characteristics. To control for possible self-selection bias, we follow Huang et al. (2014) and use a two-stage Heckman model (Heckman, 2014). In particular, in the first stage, we employ a probit regression of the likelihood of takeovers for the full sample, where firm size, Tobin’s Q, leverage, asset tangibility, leverage, stock price run-up, sales growth and price-to-earnings ratio are used as the explanatory variables (Huang et al., 2014). For cash, mixed or stock deals, we model the likelihood of cash (or stock) deals using the specification in column 4 of Table 5 as the first-stage probit model. In the second stage, we include the inverse Mills ratio as an explanatory variable in the analyses of announcement returns. Table 6 presents the results of the second stage regression on the relationship between PINs and acquirer CARs around the deal announcement date.

The results in column 1 of Table 6 show that acquirers with higher PINs are more likely to experience higher CARs around the deal announcement date. When dividing the sample into cash-only deals, mixed deals, and stock-only deals, we find that the positive relationship between acquirer PINs and announcement-period CARs is only evident in cash deals (at the 1 percent level), consistent with H3a that acquirers with a higher PIN are likely to obtain a higher abnormal announcement returns for cash deals. However, we fail to find supportive evidence for H3b that bidders with higher PINs obtain lower announcement returns when paying by stock, as the coefficient on bidder PINs is negative but insignificant for stock deals. This suggests that, different from other common measures of information asymmetry, the association between PINs and announcement returns is not existent when stock is used as a medium of exchange in takeovers.

We also include target PINs and examine its relation with acquirer announcement returns. Consistent with the view that higher PINs of the targets are likely to lead to reduced precision in the estimate of target firm value and an increased likelihood of overvalued offer prices, we find a negative and significant association between target PINs and acquirer returns, especially for cash deals. This supports H4b that bidders obtain lower announcement returns when acquiring targets with high PINs in cash deals.

Overall, our results of a positive relation between acquirers’ PINs and announcement returns for cash offers are consistent with the finding of Moeller et al.(2007) that cash offers made by bidders with high information asymmetry are considered a signal of undervaluation of the bidder’s equity, thereby leading to higher announcement-period abnormal returns.

4.3 Results on target PIN and target announcement returns

Lastly, we examine the relation between PINs and target firm announcement returns. Table 7 presents the regression results on the relation between acquirer PINs, target PINs and target CARs around the deal announcement date. We use the same procedure to calculate target firms’ 5-day CAR around the deal announcement as employed for acquiring firms, and the Heckman two-stage model to address possible selection bias. The results in column (1) show that, on average, targets with higher PINs have higher announcement-period CARs. When separating deals into cash, mixed and stock deals, we find that the positive and significant association between target PINs and CARs is only evident in the sample of cash deals (at the 5 percent level). The results are consistent with H4a that target firms with higher PINs obtain higher announcement returns when receiving cash payment. This finding also supports the view that bidders who have discovered the high level of information asymmetry in the target firm are likely to use more equity financing when acquiring targets to share the information risk with the target’s shareholders.

5.1 Offer premium and PIN

As the actual offer prices of deals are often available, prior research also analyses takeover premiums in addition to the announcement-period abnormal returns of the target firms. Similar to the argument for the targets’ announcement-period CARs, we expect target PINs to be positively related to takeover premiums and to be concentrated in cash deals. Table 8 reports the regression results on the relation between PIN and takeover premiums. Following prior studies (Lin et al., 2011; Golubov et al., 2012), the takeover premium is defined as the 4-week bid premium in percentage reported by SDC (field PREM4WK). The 4-week bid premium is calculated as the offer price over target’s closing stock price 4 weeks prior to the announcement date.

Similar to the results for announcement-period CARs in Table 7, the results in column 1 show that bidders tend to pay a higher takeover premium when acquiring targets with higher PINs. However, acquirer PINs are not significantly associated with takeover premium. When considering cash, mixed and stock deals separately, we find that a positive relationship between targets’ PIN and takeover premium only occurs in cash deals (significant at the 10 percent level). Interestingly, the association between target PIN and takeover premiums in equity deals is negative, suggesting that the sharing of information risk with target shareholders results in reduced offer prices. Overall, the results in Table 8 confirm that targets with higher PINs tend to receive a higher takeover premium in cash deals.

5.2 Alternative measure of PIN

Many of the seminal empirical studies (Easley et al., 2002; Easley and O’Hara, 2004) investigating the role of PIN on asset pricing and corporate policies use the EKO PIN measure developed by Easley et al.(2002). For our results to be comparable to those in these previous studies, we also use the EKO PIN provided by Soeren Hvidkjaer to examine the relationship between PIN and acquisition outcomes. This alternative PIN measure covers the period from 1983 to 2001. In addition, we acknowledge that more than half of our sample firms are from the manufacturing industry. While industry fixed effect is included in our analysis, we also use an industry-adjusted PIN measure to minimise the effect of potential outliers. Table 9 presents the results that use the industry-adjusted PIN measure and the EKO PIN measure. We find that our main results continue to hold.

5.3 PIN and method of payment: the role of corporate governance and Reg FD

Agency theory predicts that, due to conflict of interest, managers do not always act in the best interest of shareholders in corporate decisions such as mergers and acquisitions. Thus, corporate governance mechanisms can play an important role in preventing managers from engaging in value-destroying acquisitions. For example, Masulis et al.(2007) document that acquiring firms with more antitakeover provisions obtain lower announcement returns, suggesting managers who are immune to the disciplining of the takeover market tend to engage in value-destroying takeovers. However, it is important to note that ‘corporate governance does not have a single and unique impact on takeovers’ (Aktas et al., 2016: p. 248). Accordingly, it is unclear whether the association between PINs and acquisition outcomes would be more pronounced or weaker among firms with strong governance environments.

On the one hand, the relation between PINs and acquisition outcomes can be weaker if a better governance environment and monitoring reduce information asymmetry and insider trading. On the other hand, Ferreira and Laux (2007) show that firms with fewer antitakeover provisions (more external disciplining) demonstrate a higher degree of private information flow, indicating a more pronounced relation between PINs and takeover outcomes. To examine the effect of governance, we follow Masulis et al.(2007) to include the G-Index based on a firm’s antitakeover provisions in the regression models and its interaction with PIN. The results reported in column (1) of Table 10 show that, compared to acquirers with fewer antitakeover provisions, bidders with less external disciplining (i.e., more provisions) are less inclined to use cash financing when its own PIN is high, but tend to use more cash financing when acquiring targets with a high PIN to share the target’s information risk (Hansen, 1987). The finding is consistent with Masulis et al.(2007) that managerial tendency to engage in value-destroying takeovers increases with less disciplining from a firm’s governance mechanisms.

Regulation Fair Disclosure (FD) imposed by the Securities and Exchange Commission (SEC) in 2000 aimed to prohibit disclosure of material private information to selected market participants. However, prior research documents mixed evidence on the effectiveness of Regulation FD in reducing a firm’s degree of information asymmetry. Sidhu et al.(2008) show that Nasdaq’s stocks’ degree of information asymmetry increases by about 40 percent after the adoption of Regulation FD. On the other hand, Eleswarapu et al.(2004) document a significant decline in information asymmetry following Regulation FD. Given the mixed evidence on the effectiveness of Regulation FD, it is unclear whether and how the implementation of Regulation FD affects the relation between PINs and acquisition outcomes. We assess the impact of Regulation FD by interacting the indicator of Regulation FD with PIN. The results of our analysis in Table 10 shows that, after the implementation of Regulation FD in 2000, bidders with high PINs are less likely to use cash financing for mergers and acquisitions, but continue to rely on more equity financing when acquiring targets with high PINs.