Bidders’ Strategic Timing of Acquisition Announcements and the Effects of Payment Method on Target Returns and Competing Bids Sheng-Syan Chena, Robin K. Choub and Yun-Chi Leec aNational Taiwan University bNational Chengchi University cNational Central University I. Introduction In M&A, the method of payment has an information signaling effect Stock returns accrued to target firms are significantly higher in cash payment acquisitions than in stock or mixed payment acquisitions Huang and Walkling, 1987; Servaes, 1991; Schwert, 1996; Heron and Lie, 2002 High initial acquisition premium signals a high bidder valuation Cash payments are used to signal a higher valuation of the target firm so as to preempt any potential competing bidder. Giammarino and Heinkel (1986), Fishman (1988 and 1989), and Hirshleifer and Png (1989) 2 I. Introduction Cash payment transactions are associated with higher acquisition premiums and target shareholders earn higher returns when the acquisition is financed with cash. Cotter and Zenner, 1994; Officer, 2003 and 2006 Cash payment offers may be more effective in deterring competition. Fishman (1989) Mixed evidence: Martin (1996) documents that cash payments preempt competing bids, but Jennings and Mazzeo (1993), Officer (2003), and Bange and Mazzeo (2004) do not find supporting evidence. 3 I. Introduction Existing studies do not distinguish between the announcements of acquisitions made during overnight (nontrading) and daytime (trading) hours. Prior literature suggests that it is important to separate these two different types of acquisition announcements. Patell and Wolfson (1982), Woodruff and Senchack (1988), and Gennotte and Trueman (1996) document that managers strategically choose their announcement timing when disclosing corporate news. Announcement timing may affect the degree of information dissemination preceding the first trade in the post-announcement period. 4 I. Introduction If the news is announced during the market close, the nontrading period allows the information to be fully transmitted to all market participants. Both investors and market makers thus have more time to evaluate the news. The trading procedures during the market opening differ from those employed during normal trading hours. The opening mechanism ensures that much of the nontrading hour information is impounded into the prices (Greene and Watts, 1996; Cao, Ghysels, and Hatheway, 2000; Masulis and Shivakumar, 2002). 5 I. Introduction Market reaction is greater for announcements made during nontrading (Greene and Watts, 1996; Masulis and Shivakumar , 2002). Their findings indicate that nontrading-hours announcements tend to have more information content than trading-hours announcements. We argue that bidder managers’ strategic timing of acquisition announcements is important in assessing the effects of payment method on signaling target valuation and deterring competing bids. 6 II. Hypothesis Development A. The information effect of payment method during different announcement times Bidders may choose to release their acquisition announcements during the market close to achieve the widest possible information dissemination. They would choose to release their cash-financed acquisition with high premiums during nontrading periods. Result in higher stock returns to target shareholders than stock or mixed payment offers. The timing strategy of a bidder is likely to maximize the signaling effect of the method of payment and to create a unanimous market perception of the bidder’s valuation of the target. Market participants have more time to obtain, analyze, and evaluate the new information arising from overnight announcements. II. Hypothesis Development If a bidder decides to make an announcement during normal trading hours, the information asymmetry that exists between informed traders and other market participants cannot be reduced (Francis, Pagach, and Stephan, 1992). Chaotic traders tend to immediately react to daytime disclosures, thereby creating excessive noise during trading hours. The information signaling effect of the payment method will thus be weakened, resulting in greater uncertainty with regard to the market perception of the bidder’s valuation of the target and his/her determination to acquire it. 8 II. Hypothesis Development B. Preemptive bidding during different announcement times Cash payment offers tend to lower the likelihood of competition than other offers when the announcements are made during nontrading periods. Competing bids are negatively related to the probability of bid success (Officer, 2003; Bange and Mazzeo, 2004; Hsieh and Walkling, 2005), Due to higher premiums A greater likelihood of completing proposed acquisitions for those overnight announcements that involve all cash payment. In contrast, the positive effects of the payment method, in terms of deterring potential bidders and completing proposed transactions, are less likely when the announcements are made during normal trading hours. 9 III. Data and Methodology A. Data and sample description SDC Worldwide Mergers and Acquisitions database from 1995 to 2004. NYSE, Amex, and Nasdaq. The specific timing of the public announcement is identified by searching the Dow Jones News Retrieval Service (DJNRS) database. Intraday trade and quote data from the Trade and Quote (TAQ) database Examine the real-time market reaction to acquisition events in the two-hour trading period immediately surrounding the announcements. 10 III. Data and Methodology 1,230 acquisition announcements. The NYSE, Amex, and Nasdaq are open for trading from 9:30 a.m. to 4:00 p.m. EST. If an acquisition is announced outside this trading period or during holiday periods, this is defined as a nontrading-hours announcement An announcement made during 9:30 a.m. to 4:00 p.m. is referred to as a trading-hours announcement. Table 1 reports the sample distributions and summary statistics for the 1,230 initial acquisition announcements. 11 Table 1 Sample Distributions and Summary Statistics Panel A: Sample Distribution by Announcement Timing and Payment Method Payment Method Announcement Timing Cash Stock Mixed Nontrading Hours 329 366 299 (26.75) (29.76) (24.31) Percentage of Sample (%) Trading Hours Percentage of Sample (%) Total Percentage of Sample (%) Total 994 (80.81) 86 88 62 236 (6.99) (7.15) (5.04) (19.19) 415 454 361 1,230 (33.74) (36.91) (29.35) (100.00) Panel B: Descriptive Statistics of Other Bidder, Target, and Deal Characteristics Subsample Full Sample 0.14 Nontrading-Hours Announcements 0.15 Trading-Hours Announcements 0.11 Difference 0.04* Termination Fee 0.62 0.63 0.56 0.07* Poison 0.01 0.01 0.01 -0.00 Hostile 0.06 0.05 0.08 -0.03 Toehold 0.05 0.05 0.08 -0.03 Financial Industry 0.28 0.29 0.25 0.05 Relatedness 0.65 0.64 0.67 -0.03 Relative Size 0.32 0.31 0.40 -0.10** [0.14] [0.12] [0.18] [-0.06]*** 4.92 4.40 6.96 [1.90] [1.82] [2.22] Variables Tender Offer Market-to-Book ratio -2.56 [-0.40]*** Panel C: Sample Distribution by Year Year N 1995 8 1996 12 1997 17 1998 47 1999 162 2000 301 2001 230 2002 134 2003 162 2004 157 III. Data and Methodology B. Methodology B.1. Measurement of acquisition premiums The acquisition premium employed in this study is defined as follows: Acquisition Prem ium Bid Price Last Preacquisition Target Closing Price Last Preacquisition Target Closing Price 13 III. Data and Methodology B. Methodology B.2. Measurement for the stock price impact of acquisition announcements Follow Busse and Green (2002) Calculate the percentage price change from the prevailing mid-quote data. The real-time share price response to the target firm is measured by the cumulative stock returns starting from one trading hour before to one trading hour after the announcement. For nontrading-hours announcements, the cumulative returns over a twohour trading window is adjusted for changes in the overall market. 14 III. Data and Methodology C. Evidence on different information processes by announcement in days of the week and by exchanges As previously pointed out, a unanimous market perception of the bidder’s valuation of the target is more likely to achieve, if the time of nontrading periods is longer. One way of testing this hypothesis is to examine the effects of announcements on weekends versus on weekdays. The signaling effect should be stronger for announcements made after market close on weekends than on weekdays. Moreover, it is interesting to see whether the effects of nontrading periods are affected by trading mechanisms. 15 Table 2 Target Firm Stock Price Reaction to Acquisition Announcements by Day of the Week and by Exchanges Panel A: Target Firm Stock Price Reaction to Nontrading-Hours Acquisition Announcements for Subsamples Grouped by Payment Method and Weekend vs. Weekday Announcements Weekend Weekday Difference A.1. Cash Payment Mean (%) 27.61 21.63 5.98* Median (%) 22.39 15.43 6.96** N 85 244 A.2. Stock Payment Mean (%) 11.45 12.04 -0.59 Median (%) 7.94 7.71 0.22 N 105 261 A.3. Mixed Payment Mean (%) 16.27 14.07 2.20 Median (%) 13.41 11.20 2.22 N 94 205 Panel B: Target Firm Stock Price Reaction to Acquisition Announcements for Subsamples Grouped by Exchanges and Announcement Timing Nontrading-Hours Trading-Hours Difference Announcements Announcements B.1. NYSE Mean (%) 13.34 4.44 8.90*** Median (%) 9.53 0.56 8.97*** N 216 67 B.2. AMEX Mean (%) 11.29 2.67 8.62*** Median (%) 8.61 0.50 8.12** N 67 16 B.3. Nasdaq Mean (%) 17.92 9.01 8.91*** Median (%) 12.87 3.27 9.59*** N 711 153 16 Panel A: Nontrading-hours cash offers made during weekend have significantly higher mean and median cumulative returns than those made during weekdays. Panel B: Nontrading-hours announcements are found to have significantly higher mean and median cumulative returns than trading-hours announcements across three exchanges. IV. Evidence on the Information Effect of the Payment Method for Different Announcement Times Table 3 Cross-Sectional Regression Analyses of Acquisition Premiums Nontrading-Hours Announcements Variable Model 1 Model 2 Model 1 Model 2 Intercept 0.2767 0.2721 0.2486 0.2358 (22.95)*** (8.91)*** (11.46)*** (4.64)*** 0.0865 0.0635 0.0007 0.0015 (3.52)*** (2.35)** Cash Payment Tender Offer (1.13) 0.0261 Poison 0.0307 (1.02) (0.72) -0.0836 -0.0146 (-1.11) Hostile (-0.07) 0.0059 -0.0077 (0.14) Toehold (-0.09) -0.1336 -0.0553 (-2.81)*** Financial Industry (-0.77) -0.0383 -0.0598 (-1.57) Relatedness (-1.44) 0.0372 Relative Size Market-to-Book Ratio 0.0426 (1.57) (1.12) -0.0360 -0.0741 (-1.99)** (-2.33)** -0.0010 0.0007 (-2.64)*** N 994 2 Adjusted R F-Value (0.04) 0.0666 (1.28) A. Acquisition premiums For overnight announcements, the cash payment dummy are positive and statistically significant. For daytime announcements, the acquisition premiums are insignificantly related to the cash payment dummy. (0.02) 0.0500 Termination Fee Trading-Hours Announcements 866 (0.98) 236 219 0.0132 0.0273 0.0043 0.0313 14.32*** 3.43*** 0.00 1.70* Table 4 Cross-Sectional Regression Analyses of Target Firm Stock Price Reaction to Acquisition Announcements Variable Intercept Cash Payment Nontrading-Hours Announcements Model 1 Model 2 0.1317 -0.0033 (18.61)*** (-0.14) 0.1000 (5.75)*** 0.0763 (4.90)*** Trading-Hours Announcements Model 1 Model 2 0.0628 0.0148 (5.82)*** (0.58) 0.0276 (1.54) 0.0313 (1.62) Acquisition Premium 0.3851 (6.22)*** 0.1714 (5.22)*** Tender Offer 0.0246 (1.07) 0.0286 (1.02) Termination Fee 0.0403 (2.79)*** 0.0428 (2.13)** Poison -0.0180 (-0.45) 0.0023 (0.03) Hostile 0.0071 (0.34) 0.0284 (0.74) Toehold -0.0234 (-0.80) 0.0151 (0.45) Financial Industry -0.0096 (-0.65) 0.0067 (0.34) Relatedness 0.0153 (1.08) -0.0331 (-1.83)* Relative Size -0.0177 (-1.85)* -0.0149 (-0.97) Market-to-Book Ratio -0.0002 (-1.28) -0.0001 (-0.35) N Adjusted R2 F-value 994 0.0417 44.20*** 866 0.3591 45.06*** 236 0.0058 2.38 219 0.1415 4.27*** B. Target firm stock price reaction The coefficients for the cash payment dummy are significantly positive at the 1% level for overnight announcements. They are statistically insignificant for daytime announcements. Table 5 Cross-Sectional Regression Analyses of Target Firm Stock Price Reaction to Acquisition Announcements Net of Acquisition Premiums Nontrading-Hours Announcements Trading-Hours Announcements Variable Model 1 Model 2 Model 1 Model 2 Intercept -0.1450 -0.1735 -0.1858 -0.1806 (-14.12)*** (-7.98)*** (-10.21)*** (-3.72)*** 0.0298 0.0497 0.0269 0.0301 (1.67)* (2.52)** Cash Payment Tender Offer (0.74) 0.0049 -0.0266 (0.19) Termination Fee (-0.47) 0.0237 0.0174 (1.26) Poison (0.43) 0.0265 0.0145 (0.29) Hostile (0.08) -0.0044 0.0348 (-0.11) Toehold (0.45) 0.0463 0.0609 (1.10) Financial Industry (0.89) 0.0133 Relatedness 0.0563 (0.69) (1.42) -0.0018 -0.0684 (-0.10) Relative Size (-1.89)* 0.0042 0.0465 (0.37) Market-to-Book Ratio (1.54) 0.0003 -0.0007 (0.74) N 994 2 (0.77) 866 Adjusted R 0.0018 0.0010 F-value 2.79* 1.08 (-1.03) 236 -0.0016 0.62 219 0.0154 1.34 C. Target firm stock price reaction net of acquisition premiums The effect of cash payment offers is correlated with the acquisition premiums. Overnight cash payment offers still have a marginal effect on the target’s stock price reaction after considering the effect of acquisition premiums on target announcement returns. IV. Evidence on the Information Effect of the Payment Method for Different Announcement Times The results from Tables 3 to 5 support our hypothesis. For overnight acquisition announcements, cash payment offers are associated with significantly higher acquisition premiums and target returns. For daytime announcements, no significant differences in the acquisition premiums and target returns are found between cash acquisitions and those using other payment methods. Even after considering the effect from acquisition premiums, cash payment offers cause greater market reaction than other payment offers for nontrading-hours announcements, but not for trading-hours announcements. 20 V. Evidence on Deterring Competition by Cash Payments for Different Announcement Times A. The relationship between the payment method and competition Table 6 Frequency of Single and Multiple Bidders Total Single Bidders Payment Method N N % Nontrading-Hours Announcements Cash 329 321 97.57 Stock 366 348 95.08 Mixed 299 279 93.31 Subtotal 994 948 95.37 Trading-Hours Announcements Cash 86 73 84.88 Stock 88 81 92.05 Mixed 62 53 85.48 Subtotal 236 207 87.71 Whole Sample 1,230 1,155 93.90 Multiple Bidders N % 8 18 20 46 2.43 4.92 6.69 4.63 13 7 9 29 75 15.12 7.95 14.52 12.29 6.10 Nontrading-hours: Cash(2.43%) v.s. Stock (4.92%), Diff=-2.49%, 10% significant Cash(2.43%) v.s. Mixed (6.69%), Diff=-4.26%, 5% significant Trading-hours: All differences are insignificant. The frequency of multiple competitors is significantly lower for cash payment offers for nontrading-hours announcements, this relationship does not hold for trading-hours announcements. 21 Table 7 Logistic Regression Analyses of the Presence of Competing Bids Nontrading-Hours Announcements Trading-Hours Announcements Variable Model 1 Model 2 Model 1 Model 2 Intercept -2.8034 -2.5446 -2.1252 -3.7757 (-16.78)*** (-5.42)*** (-8.03)*** (-3.79)*** -0.8886 -1.1559 0.3999 -0.1704 (-2.25)** (-2.42)** Cash Payment Tender Offer (1.00) 0.4951 Termination Fee 0.5696 (0.98) (0.70) -0.7254 -0.8078 (-1.89)* Poison (-0.28) (-1.43) -14.0302 -0.7678 (-0.03) (-0.41) Hostile Toehold Financial Industry 2.5069 3.6232 (5.36)*** (3.69)*** -13.0174 -12.6365 (-0.04) (-0.03) -1.1862 -0.0299 (-2.55)** Relatedness (-0.05) 0.6135 2.3548 (1.55) Relative Size (2.65)*** 0.1973 -0.3370 (1.65)* Market-to-Book Ratio (-0.50) -0.1166 0.0022 (-1.50) N Pseudo R 994 2 0.0160 866 0.2174 (0.29) 236 0.0056 219 0.2895 For overnight announcements, the coefficients on the cash payment dummy are negative and statistically significant at the 5% level. For daytime announcements, the cash payment dummy has no significant influence on competing bids. Table 8 Logistic Regression Analyses of the Presence of Competing Bids for Subsamples Grouped by Acquisition Premium Size and Payment Method High Premiums Low Premiums Variable Cash Others Cash Others Intercept -3.1369 -3.2111 -2.6432 -2.0235 (-2.44)** (-3.59)*** -1.3411 -0.8581 Overnight (-1.65)* Tender Offer Termination Fee Poison Hostile Toehold Financial Industry 0.0347 (2.12)** -1.5415 -0.1790 (-1.55) (-0.29) -11.0799 -15.2206 (-0.02) (-0.01) Market-to-Book Ratio Pseudo R -1.9571 (-1.21) -1.0961 (-0.88) -2.1701 (-0.77) 1.0120 (1.39) -0.8525 (-1.86)* -12.2604 (-0.01) 2.3289 (1.74)* (3.87)*** (2.56)** (3.61)*** -13.1961 -12.6020 -11.8323 -13.2902 (-0.05) (-0.02) (-0.05) (-0.03) -1.7727 2.1868 1.3729 -0.3391 (-0.55) 0.8594 (1.38) 0.5177 (1.13) (1.47) -0.1373 -0.0027 0.3864 (-0.21) 353 0.2326 -0.0068 (-0.01) 2.5983 (1.46) -0.3850 (-0.26) -0.4185 (-0.98) 160 0.4964 (-1.62) 4.7239 194 2 (-2.13)** -0.7869 3.0407 (-0.69) N -2.1576 (-3.00)*** 2.0796 (1.83)* Relative Size 1.5026 (0.03) (-1.38) Relatedness (-1.51) (-1.39) -0.8840 (-1.60) 0.7944 (1.49) 0.0669 (0.33) -0.0030 (-0.21) 378 0.1748 Robustness checks: Larger acquisition premiums deter competition, irrespective announcement timing. For cash payment offers, overnight announcements are associated with significantly lower frequency of competition in both high and low premium subsamples. V. Evidence on Deterring Competition by Cash Payments for Different Announcement Times B. The relationship between the method of payment and the success of the bid Table 9 Frequency of Successful and Unsuccessful Acquisition Bids Total Successful Bids Payment Method N N % Nontrading-Hours Announcements Cash 329 289 87.84 Stock 366 302 82.51 Mixed 299 246 82.27 Subtotal 994 837 84.21 Trading-Hours Announcements Cash 86 66 76.74 Stock 88 67 76.14 Mixed 62 51 82.26 Subtotal 236 184 77.97 Whole Sample 1,230 1,021 83.01 Unsuccessful Bids N % 40 64 53 157 12.16 17.49 17.73 15.79 20 21 11 52 209 23.26 23.86 17.74 22.03 16.99 Nontrading-hours: Cash(87.84%) v.s. Stock (82.51%), Diff=5.33%, 5% significant Cash(87.84%) v.s. Mixed (82.27%), Diff=5.57%, 10% significant Trading-hours: All differences are insignificant. For overnight acquisition announcements, deals involving all cash payment are more likely to be successful than deals involving other payment methods. However, this relationship does not hold for daytime acquisition announcements. 24 Table 10 Logistic Regression Analyses of the Likelihood of Acquisition Bid Success Nontrading-Hours Announcements Trading-Hours Announcements Variable Model 1 Model 2 Model 1 Model 2 Intercept 1.5441 0.5552 1.3049 0.3120 (15.17)*** (2.24)** (6.55)*** 0.4331 0.7872 (2.20)** (2.85)*** Cash Payment Tender Offer -0.1110 (-0.34) 0.5549 Poison Financial Industry 1.4433 (6.88)*** (3.29)*** -12.6252 (-0.01) -3.0514 -2.4591 (-6.33)*** (-2.88)*** 1.8577 1.7039 (2.31)** (1.97)** 0.7164 -0.0644 (2.60)*** Relatedness (-0.15) 0.1960 Relative Size Market-to-Book Ratio 0.5628 (0.83) (1.39) -0.2546 -0.5566 (-2.15)** (-1.85)* 0.0124 0.0251 (0.72) N 994 2 Pseudo R 0.0059 (0.96) 1.6320 (1.50) Toehold 866 0.2247 (-0.33) 1.4348 Hostile 0.4364 -0.2291 (1.30) Termination Fee (0.61) (0.94) 236 0.0005 219 0.2216 For overnight announcements, the coefficients on the cash payment dummy are significantly positive. For daytime announcements, the cash payment dummy does not have any significant impact on the probability of bid success. V. Evidence on Deterring Competition by Cash Payments for Different Announcement Times Tables 6 to 8 show that overnight announcements of cash payment offers lower the likelihood of competition. Tables 9 and 10 show that cash payment acquisition offers announced overnight are more likely to result in successful completion. Because overnight cash payment offers are more likely to deter potential competing, the likelihood of bid success is higher for these deals. 26 VI. Two-stage estimation of cash payments and acquisition premiums for different announcement times Payment methods and premiums are likely to be simultaneously determined. Thus, we need to show that managers strategically announce cash payment offers during nontrading periods, which result in higher premiums. To address the potential endogeneity problem between cash payment offers and acquisition premiums, we follow Officer (2003) and estimate the relation between these two decisions using a two-stage regression approach. 27 Table 11 Two-stage Regression Analysis of Cash Payments and Acquisition Premiums by Announcement Timing Variable Intercept Cash or Premium Instrument Second-stage Regressions Nontrading-Hours Trading-Hours Announcements Announcements Dep. Var. = Dep. Var. = Dep. Var. = Dep. Var. = Acquisition Cash Acquisition Cash Premium Payment Premium Payment (OLS) (Logistic) (OLS) (Logistic) 0.3007 -0.7443 0.1956 3.3131 (11.46)*** (-0.49) (3.22)*** (1.55) 0.0395 (2.43)** Tender Offer Termination Fee 2.2922 (0.48) 0.0411 (1.17) 1.8086 (4.18)*** -9.7756 (-1.22) 2.2874 (3.03)*** 0.0523 (1.91)* -0.9819 (-4.96)*** 0.0911 (1.43) -1.1856 (-2.73)*** Poison -0.1092 (-0.90) 1.0754 (1.04) 0.0025 (0.01) -0.5646 (-0.36) Hostile 0.0133 (0.25) -0.4169 (-0.98) -0.0130 (-0.16) 0.0596 (0.08) Toehold -0.1626 (-2.72)*** 1.2789 (1.77)* -0.0610 (-0.85) -0.3971 (-0.49) Financial Industry -0.0216 (-0.77) -0.4634 (-1.83)* -0.0361 (-0.75) -1.1646 (-1.96)** Relatedness 0.0340 (1.45) Relative Size 0.0193 (0.65) Market-to-Book Ratio N Adjusted R2 Pseudo R2 0.0595 (1.46) --1.4004 (-3.88)*** -0.0524 (-1.36) -1.2547 (-1.87)* -0.0009 (-1.51) -0.0009 (-0.11) 0.0056 (1.32) -0.1113 (-1.94)* 866 0.0216 - 866 0.1710 219 0.0359 - 219 0.1965 Overnight cash payments have a significantly positive influence on acquisition premiums, while premiums do not significantly affect the probability of cash offers. Cash payments offers result in higher premiums for target shareholders only when announcements are made during nontrading periods. Managers strategically announce cash payment offers during the overnight periods as an instrument to signal the high valuation of the target to the market. VII. Conclusion For overnight announcements, the acquisition premiums and target stock returns are significantly higher when the acquisitions are financed with cash. Furthermore, overnight cash payment offers are more likely to result in a significantly lower frequency of multiple competitors and a significantly higher frequency of bid. However, we do not find such relationship for daytime acquisition announcements. Acquisition announcements signaling the bidder’s high valuation of the target, through a cash offer with large premiums, is more likely to be made during nontrading hours. The acquisition announcement timing of bidders is an important consideration in assessing the effects of the method of payment on signaling target valuation and preempting competing bids. 29