The Effects of Horizontal Merger Operating Efficiencies on Rivals, Customers, and Suppliers

The Effects of Horizontal Merger Operating Efficiencies on Rivals, Customers, and Suppliers Abstract We study how operating efficiencies in horizontal mergers affect market reactions of merging firms’ rivals, customers, and suppliers. We measure operating efficiency gains using projections disclosed by merging firms’ insiders. Higher efficiency gains are associated with lower announcement returns to merging firms’ rivals (due to increased equilibrium output of merging firms), higher returns to their customers (due to lower equilibrium price of merging firms’ output), and higher returns to their suppliers (due to the merged firm’s higher equilibrium demand for inputs). Our results suggest that the pass-through of efficiency gains along merging firms’ supply chains is as important as the effects of post-merger changes in market power. 1. Introduction Mergers and acquisitions (M&A) can impact not only the merging firms, but also firms that are economically linked to them: those competing with the merging firms in product markets (rivals) as well as those operating along the merging firms’ supply chains (“corporate customers” and “suppliers”). Industrial organization theory predicts two main channels through which a horizontal merger can impact merging firms’ rivals, customers, and suppliers (related firms hereafter). The first channel is through the change in the structure of the industry in which merging firms operate, which increases their “market power” and affects competition in input and output markets (e.g., Stigler, 1964; Salant, Switzer, and Reynolds, 1983; Perry and Porter, 1985). The second channel is through “operating efficiencies” that merging firms may be able to realize, which affect their demand for inputs and supply of output. Operating efficiencies can take the form of cost savings, such as elimination of duplicate functions and facilities, and/or productivity gains, such as economies of scale (e.g., Kaplan, 2000; Maksimovic and Phillips, 2001; Jovanovic and Rousseau, 2002; Lambrecht, 2004; Rhodes-Kropf and Robinson, 2008). Existing empirical studies that examine the effects of horizontal mergers on related firms tend to concentrate on the market power channel. Three main findings emerge from these studies. First, the values of merging firms’ rivals tend to increase around horizontal merger announcements (e.g., Eckbo, 1983, 1985; Stillman, 1983; Song and Walkling, 2000; Fee and Thomas, 2004; Shahrur, 2005). Second, merger announcement returns to customers and suppliers are generally statistically insignificant (e.g., Fee and Thomas, 2004; Shahrur, 2005). Third, in most cases, measures of market power, such as merging firms’ industry concentration levels and changes, do not display significant relations with changes in valuations of related firms around mergers (e.g., Fee and Thomas, 2004; Shahrur, 2005). In this paper, we focus instead on merger-related operating efficiencies. Industrial organization theory predicts the following effects of operating efficiencies on related firms. First, operating efficiencies give the merged firm a competitive advantage over its rivals, reducing their valuations. Second, operating efficiencies reduce the merged firm’s production costs, leading to lower equilibrium output prices, benefiting the merged firm’s customers. Third, operating efficiencies that take the form of cost savings raise merging firms’ equilibrium optimal output levels and resulting demand for inputs, benefiting the merged firm’s suppliers. On the other hand, operating efficiencies that take the form of productivity gains may reduce the amount of inputs required for producing any given output level, potentially hurting the suppliers. Estimating the effects of operating efficiencies on related firms empirically is not trivial. Previous studies (e.g., Eckbo, 1983, 1985; Fee and Thomas, 2004; Shahrur, 2005) typically proxy for operating efficiencies by bidder’s and target’s combined announcement returns and post-merger changes in operating performance. Operating efficiencies, however, are not the sole driver of short-term and long-term post-merger performance. First and foremost, post-merger changes in industry structure and in the merged firm’s market power are expected to have a positive impact on its value and performance. In addition, investors’ anticipation of merger activity and uncertainty about deal completion may affect merging firms’ combined announcement returns. The combination of various effects of horizontal mergers makes drawing unambiguous inferences about the existence and magnitudes of operating efficiencies from announcement returns difficult (e.g., Kaplan, 2000; Eckbo, 2014). We attempt to overcome these difficulties by using a direct estimate of merger-related operating efficiency gains. We measure these gains using a unique, hand-collected dataset of forecasts disclosed by merging firms’ managers at the time of merger announcements, which is also used in Bernile and Bauguess (2014). To assemble this dataset, for each M&A deal in our sample, we search news stories and press releases from all English-language sources for insiders’ projections of operating efficiency gains. Our empirical results are largely consistent with predicted effects of operating efficiencies. First, operating efficiency gain projections are positively and significantly related to rivals’ reactions to merger announcement. A one-standard-deviation increase in projected efficiency gains is associated with 1.3–2.6% decrease in rivals’ announcement returns. Second, a one-standard-deviation increase in projected gains is associated with 0.3–1.1% increase in customers’ values. Third, a one-standard-deviation increase in efficiency gains is associated with 0.6–2.7% increase in suppliers’ announcement returns. The last result suggests that operating efficiencies tend to take the form of cost savings, increasing the demand for suppliers’ inputs, rather than productivity gains that could reduce the demand for inputs. These relations are more pronounced in subsamples of mergers in which insiders’ projections are expected to be more credible ex ante and/or turn out to be more credible ex post. First, the relations between projected efficiency gains and related firms’ announcement returns tend to be stronger in mergers among more capital-intensive bidders and targets, which have a higher potential to generate cost savings. Second, these relations are more pronounced in mergers in which the market capitalizes a larger portion of projected efficiency gains, i.e., in mergers in which the ratio of combined bidder’s and target’s announcement return to efficiency gain forecast scaled by the merging firms’ combined pre-merger market capitalization is larger. In addition, the effects of projected efficiency gains on related firms tend to be stronger when the implied change in merging firms’ industry structure is lower, i.e., when operating efficiency gains are more likely to be the main driver of related firms’ reactions to a merger announcement. Finally, the relation between projected efficiency gains and rivals’ announcement returns are typically stronger in mergers involving large targets. Collectively, our results are largely consistent with empirical predictions derived from industrial organization theory, and highlight the importance of efficiency gains in M&A not only for the merging firms but also for firms associated with them either horizontally or vertically. Finally, our findings are, for the most part, also consistent with the market power effect of mergers along the supply chain. Post-merger implied change in the Herfindahl index in the merging firms’ industry is negatively and significantly associated with returns to merging firms’ customers and suppliers. Interestingly, in the case of customers, controlling for operating efficiencies is crucial for uncovering the market power effect: without such control, the relation between the implied change in the Herfindahl index and merging firms’ customers’ announcement returns tends to be statistically insignificant. There are important issues to consider when using insiders’ projections as a proxy for expected efficiency gains in mergers. First, managerial projections feature various degrees of detail, and quantifying them is not straightforward. Similar to past studies (e.g., Gilson, Hotchkiss, and Ruback, 2000; Houston, James, and Ryngaert, 2001; Bernile and Bauguess, 2014), we make numerous assumptions when interpreting insiders’ forecasts, regarding forecast horizon, evolution of projected operating efficiency gains over time, as well as discount rate used to estimate the present value of the gains. While our procedure involves many discretionary choices, our results are generally robust to various modifications to it. Second, the disclosure by insiders of managerial projections of operating efficiencies is voluntary. As a consequence, projections are only available in about one-third of M&A deals, although they are typically present in larger deals. Due to the voluntary nature of disclosure and various data limitations that we discuss below, our sample is not large: it includes 480 M&A deals in which we can identify over 5,000 product market rivals, 560 customers, and over 1,500 suppliers.1 Third, since operating efficiency gain projections are voluntary, they are subject to potential self-selection. We control for self-selection by employing the Heckman (1979) two-stage estimation procedure. In the first stage, we estimate a probit model of the decision to disclose efficiency gain forecasts. Our main instrument is the number of earnings guidances provided by both bidder and target in the year preceding the merger announcement. The number of earnings guidance events is unlikely to be directly related to valuation and performance of firms following the merger, i.e., it is likely to satisfy the exclusion restriction. The first-stage estimates confirm that the availability of projections is indeed positively and significantly related to the number of bidder’s and target’s earnings guidances. In addition, the number of guidances is related to merging firms’ style of disclosure, potentially satisfying the relevance restriction. After estimating the first-stage regression, we compute resulting inverse Mills ratios, which we use in the second-stage estimation of the relation between projected operating efficiency gains on the one hand and changes in values of merging firms and related firms on the other hand, to correct for potential self-selection of the disclosure choice. Fourth and relatedly, insiders’ projections of merger-related operating efficiency gains need not be informative because merger decisions and associated forecasts by insiders may be driven by self-interested behavior (e.g., Morck, Shleifer, and Vishny, 1990; Gorton, Kahl, and Rosen, 2009), hubris (e.g., Roll, 1986), and/or market misvaluation (e.g., Shleifer and Vishny, 2003; Rhodes-Kropf and Viswanathan, 2004). To examine whether managerial projections are a valid proxy for expected operating efficiencies, we examine the relation between projected gains on one hand and merging firms’ announcement returns and post-merger changes in operating performance on the other hand. Merging firms’ combined announcement returns are positively and significantly related to operating efficiency projections. A one-standard-deviation increase in the present value of projected gains relative to bidder’s and target’s combined pre-merger market capitalization is associated with 2.5–3.9% increase in combined announcement return. On average, the market capitalizes around 45% of projected efficiency gains. In addition, a one-standard-deviation increase in the value of projected efficiency gains is associated with a 1.5–2.5 percentage point increase in bidder’s ROA over the 3-year period following the merger, with a significant 1.2–1.4 percentage-point decrease in the ratio of cost of goods sold (COGS) to sales, and with a 0.6–1.1 percentage-point reduction in the ratio of sales, general, and administrative expenses (SG&A) to sales. This evidence, which we interpret as suggesting that managerial forecasts of merger-related operating efficiencies are economically relevant, is consistent with findings in Houston, James, and Ryngaert (2001) within a sample of bank mergers and in Bernile and Bauguess (2014) within a sample of mergers that includes vertical and diversifying deals in addition to horizontal ones. Fifth, in principle, insiders’ projections of efficiency gains may be provided with the goal of masking other effects of mergers, such as merging firms’ increased market power. Importantly, if insiders’ forecasts in fact aim to disguise the anticompetitive effects of mergers, then these forecasts should have effects on valuation of related firms that are opposite to the predicted effects of operating efficiencies: higher market power in the merging firms’ industry is likely to benefit rivals and to be detrimental to customers and suppliers. Thus, managers’ incentives to use operating efficiency forecasts to disguise potential anti-competitive effects of mergers work against us finding evidence consistent with the operating efficiencies channel. In addition to directly measuring projected operating efficiencies, we contribute to the literature that analyzes the effects of horizontal mergers on related firms by using Text-based Network Industry Classification (TNIC), proposed by Hoberg and Phillips (2010a, 2010b), to identify firms whose product offerings are similar and which are, therefore, likely to compete in output markets. We use this classification to identify horizontal merger deals as well as merging firms’ rivals. In doing so, we depart from industry definitions used in most previous studies (e.g., Eckbo, 1983, 1985; Fee and Thomas, 2004; Shahrur, 2005; Fan and Goyal, 2006; Ahern, 2012, Ahern and Harford, 2014). This is important because standard industry classifications may not correctly identify interactions among firms in output markets. First, SIC and NAICS classifications are based on production processes, as opposed to products that firms supply. Second, these classifications are static, in the sense that they are rarely adjusted in the face of evolving product markets and/or firms moving across industries. Third, SIC and NAICS classifications impose transitivity, while it is possible that two competing firms may have different rivals. To identify merging firms’ upstream and downstream supply chain relations, we use Cohen and Frazzini’s (2008) data on customer–supplier links, which is based on firms’ disclosures of (large) corporate customers available in Compustat Industry Segment files. This method differs from the more common approach based on industry-level input–output matrices derived from the “make” and “use” tables of the Bureau of Economic Analysis (e.g., Shahrur, 2005; Ahern, 2012). Using customer–supplier links allows us to identify direct effects of operating efficiency gains on firms that are “actually,” rather than “potentially” related to the merging firms. In the next section, we discuss theoretical effects of operating efficiencies. In Section 3, we present the data and provide summary statistics. We examine the effects of operating efficiencies on merging firms in Section 4 and on related firms in Section 5. Section 6 summarizes our findings and concludes. 2. Predicted Effects of Operating Efficiencies on Related Firms In this section, we discuss theoretical effects of operating efficiencies on merging firms’ rivals as well as on their corporate customers and suppliers. The empirical predictions that follow from this discussion form the basis of our subsequent empirical analysis. To corroborate the intuitive arguments articulated in this section, an Online Appendix presents a stylized model that formally derives the empirical predictions. Operating efficiencies raise optimal output levels of the merged firm, ceteris paribus, because they reduce marginal costs of production for a given level of output. The increase in the merged firm’s combined equilibrium output and the associated reduction in equilibrium prices for its products hurts its output market rivals, which face lower residual demand for their products. The magnitude of the negative impact of a merger on rivals increases in the operating efficiency gains stemming from it. Prediction 1. The returns to rivals around M&A announcements are negatively related to operating efficiency gains. The increase in merging firms’ equilibrium output relative to their combined pre-merger output due to operating efficiencies has a positive effect on their customers. In particular, operating efficiencies benefit customers by leading the merged firm to supply greater quantities at lower prices in equilibrium. Therefore, operating efficiencies have a positive effect on customers, and the strength of this effect increases in the magnitude of efficiency gains. Prediction 2. The returns to customers around M&A announcements are positively related to operating efficiency gains. The effect of operating efficiencies on merging firms’ suppliers is more subtle because it depends on the nature of efficiency gains. In particular, two sources are often cited as important drivers of operating efficiencies (e.g., Kaplan, 2000). The first, which we refer to as cost savings, is due to elimination of duplicate functions or reduction in overhead costs. The second type, which we refer to as productivity gains, stems from improvements to the merged firm’s production function relative to pre-merger stand-alone production functions, i.e., a reduction in the quantity of inputs required to produce any given level of output. Cost savings do not affect the quantities of variable inputs required to produce a given level of output. Instead, they may lead to lower costs of production and, as a result, shift the merged firm’s demand for inputs upwards, raising equilibrium input prices and benefiting suppliers. Productivity gains, on the other hand, may lead to a downward shift in the merged firm’s demand for inputs, as more efficient production reduces the amount of inputs required for any given level of output, potentially leading to lower equilibrium input prices. As a result, firms supplying inputs to the merged firm may be hurt by productivity gains. This discussion leads to the following empirical prediction. Prediction 3. The returns to suppliers around M&A announcements depend on the type of operating efficiencies: The relation between the value of operating efficiency gains and suppliers’ announcement returns is positive if operating efficiencies take the form of cost savings. The relation between the value of operating efficiency gains and suppliers’ announcement returns is negative if operating efficiencies take the form of productivity gains. 3. Data and Summary Statistics 3.1 Data Sources We use the following data sources: Our sample of M&A is from Thomson Reuters’ Securities Data Company (SDC). We manually retrieve insiders’ forecasts of merger-related operating efficiency gains from a keyword search of news sources in the Dow Jones Factiva database. To identify mergers between firms that compete in the same product market (horizontal mergers hereafter), as well as to define the set of their product market rivals, we rely on Hoberg and Phillips’ (2010a, 2010b) TNIC, which is based on a measure of textual similarity between firms’ product descriptions in 10-K filings for each pair of Compustat firms. We use firm-level customer–supplier linkages, established by Cohen and Frazzini (2008), who rely on Compustat Industry Segment data, to identify actual supply chain relations of merging firms. We use the NBER Patent Citations Data Project to measure technological similarities between bidders and targets. We use the Bureau of Economic Analysis data to identify bidders and targets operating in the same geographical areas. We measure merger announcement returns and post-merger changes in merging firms’ cost and profitability ratios using data from the Center for Research in Securities Prices (CRSP) and Compustat Annual Industrial Files, respectively. To control for potential self-selection in disclosing efficiency gain projections, we instrument for insiders’ propensity to supply operating efficiency forecasts using acquirer’s and target’s pre-merger propensity to provide earnings guidances. Our alternative instruments are based on analyst coverage of bidder’s and target’s equity, and the proportion of firms’ equity held by institutional investors. The data on analyst coverage and earnings guidance are from Thomson Reuters’ I/B/E/S database, while the data on institutional holdings are from Thomson Reuters’ 13-F database. We use Bebchuk, Cohen, and Ferrell’s (2009) entrenchment index as a measure of corporate governance. We use Harford’s (2005) definitions of industry deregulation events. Data availability imposes restrictions on the sample period we examine. In particular, Hoberg–Phillips TNIC classification is available starting from 1996. Cohen–Frazzini customer–supplier relations are available until 2005. In addition, the data on insiders’ projections of operating efficiencies were hand-collected for years 1990–2005. As a result of these limitations, our final sample covers the 10-year period between 1996 and 2005. 3.2 Quantifying Forecasts of Operating Efficiency Gains Our measure of operating efficiency gains is based on merging firms’ insiders’ forecasts upon merger announcements. For each deal in our sample, we search news stories and press releases from all English-language sources on Factiva for management’s public forecasts of merger-related operating efficiency gains.2 We impose the following criteria when retrieving related news stories and press releases: Factiva keyword search window: (“variations of target name” and “variations of bidder name”) and (merger* or acquisition* or tender*) and (earn* or profi* or syne* or enha* or add or addi* or accre* or contrib* or save or cost* savi* or reve* or incr* or decr* or redu* or cut or cutt* or dilut* or neutr* or impro* or econ* of scal* or expec* or anticip*). Factiva date range window: from (announcement date-7) to (completion/withdrawal date). The strength of the effects of operating efficiencies on related firms is expected to increase in the magnitude (present value) of projected efficiency gains. Estimating the present value of projected gains is not straightforward, since managerial forecasts do not follow a particular template, as evident from typical examples of the wording of these forecasts found in the Appendix. In some cases, insiders provide timetables specifying the magnitude of annual operating efficiency gains expected to be realized during intermediate years before the benefits of the merger fully materialize. In the majority of cases, however, managerial projections are not as detailed. Some forecasts are limited to the first post-completion year. Others refer to annual gains expected between 2–4 years post-completion, but provide no guidance regarding all or some of the intermediate years. In other cases, forecasts quantify annual gains with no guidance about their horizon. Finally, sometimes insiders disclose the cumulative amount of operating efficiency gains expected to be realized over 3, 5, or 10 years after the deal’s completion. To estimate the present value of managerial forecasts of operating efficiencies, we make the following assumptions, which are similar to those made in past studies (e.g., Kaplan and Ruback, 1995; Gilson, Hotchkiss, and Ruback, 2000; Houston, James, and Ryngaert, 2001; Bernile and Bauguess, 2014). First, when annual forecasts are provided, we assume that the last projected year is the steady-state level of operating efficiencies. Second, when no details about the timing are provided, we assume that the steady-state level of operating efficiencies is reached in year 4 after merger completion. Third, when cumulative or annual forecasts have no timing or are missing intermediate years, we assume that expected operating efficiencies grow at a rate of 100% per year until the steady-state is reached.3 Fourth, we assume a flat 36% tax rate. Fifth, we discount estimated annual operating efficiencies using the weighted average of merging firms’ costs of unlevered equity, using as weights the firms’ pseudo-market value of assets—i.e., market value of common equity plus book value of long-term debt, and liquidation value of preferred equity, measured 60 trading days prior to when the target company is “put in play”.4 Finally, we use a 10-year horizon for operating efficiencies. We show in the Online Appendix that the results are robust to discounting annual operating efficiencies using a fixed 10% rate and to assuming that operating efficiencies would materialize over alternative horizons, such as 5 years or indefinitely. 3.3 Summary Statistics Panel A of Table I presents the distribution of mergers, horizontal mergers, and mergers with available projections of operating efficiency gains over time. Table I. Summary statistics This table consists of five panels. Panel A presents the distribution of mergers, horizontal mergers, and mergers with available efficiency gain projections over time. Panel B presents summary statistics of estimated value of projected operating efficiency gains when available. Panel C contains summary statistics of deal and industry characteristics. Panels D and E present summary statistics of bidder and target characteristics respectively. Panle A: This panel reports the annual number of M&As, horizontal M&As, and horizontal M&As with available operating efficiency projections. The sample of mergers is from Thomson Reuters’ SDC database. M&As are defined as acquisitions of assets, acquisitions of majority interest, or mergers in which a bidder holds less than 50% of target’s common stock on the offer date and the deals has come to a resolution (i.e., completed or withdrawn), where an offer date occurred between January 1996 and December 2005. A horizontal M&A is a merger between two firms in the same Hoberg and Phillips (2010a, 2010b) TNIC industry. M&As with projections are those in which our manual search for managerial operating efficiency forecasts in Factiva yielded results that enabled us to quantify the value of projected efficiency gains based on the algorithm discussed in Section 3.2. In computation of value-weighted proportion of horizontal M&As with projections, the weights are based on book values of target firms’ assets as of the fiscal year end prior to the year of merger announcement. Panel A. Mergers, horizontal mergers, and mergers with projections over time Year Number of M&As Number of horizontal M&As % horizontal M&As Number of horizontal M&A % horizontal M&As % horizontal M&As w/projections w/projections (EW) w/projections (VW) 1996 236 81 33.00 22 31.34 66.91 1997 304 205 67.16 56 29.44 65.55 1998 319 207 64.48 70 35.83 75.47 1999 343 225 67.96 63 28.57 68.15 2000 268 170 65.22 50 28.00 56.11 2001 226 158 71.78 49 35.17 75.33 2002 130 90 70.31 22 23.33 59.94 2003 149 114 76.19 43 38.39 74.65 2004 148 115 79.43 62 54.46 82.92 2005 140 113 80.62 43 36.54 72.40 Total 2,263 1,478 65.31 480 32.48 69.72 Panel A. Mergers, horizontal mergers, and mergers with projections over time Year Number of M&As Number of horizontal M&As % horizontal M&As Number of horizontal M&A % horizontal M&As % horizontal M&As w/projections w/projections (EW) w/projections (VW) 1996 236 81 33.00 22 31.34 66.91 1997 304 205 67.16 56 29.44 65.55 1998 319 207 64.48 70 35.83 75.47 1999 343 225 67.96 63 28.57 68.15 2000 268 170 65.22 50 28.00 56.11 2001 226 158 71.78 49 35.17 75.33 2002 130 90 70.31 22 23.33 59.94 2003 149 114 76.19 43 38.39 74.65 2004 148 115 79.43 62 54.46 82.92 2005 140 113 80.62 43 36.54 72.40 Total 2,263 1,478 65.31 480 32.48 69.72 Table I. Summary statistics This table consists of five panels. Panel A presents the distribution of mergers, horizontal mergers, and mergers with available efficiency gain projections over time. Panel B presents summary statistics of estimated value of projected operating efficiency gains when available. Panel C contains summary statistics of deal and industry characteristics. Panels D and E present summary statistics of bidder and target characteristics respectively. Panle A: This panel reports the annual number of M&As, horizontal M&As, and horizontal M&As with available operating efficiency projections. The sample of mergers is from Thomson Reuters’ SDC database. M&As are defined as acquisitions of assets, acquisitions of majority interest, or mergers in which a bidder holds less than 50% of target’s common stock on the offer date and the deals has come to a resolution (i.e., completed or withdrawn), where an offer date occurred between January 1996 and December 2005. A horizontal M&A is a merger between two firms in the same Hoberg and Phillips (2010a, 2010b) TNIC industry. M&As with projections are those in which our manual search for managerial operating efficiency forecasts in Factiva yielded results that enabled us to quantify the value of projected efficiency gains based on the algorithm discussed in Section 3.2. In computation of value-weighted proportion of horizontal M&As with projections, the weights are based on book values of target firms’ assets as of the fiscal year end prior to the year of merger announcement. Panel A. Mergers, horizontal mergers, and mergers with projections over time Year Number of M&As Number of horizontal M&As % horizontal M&As Number of horizontal M&A % horizontal M&As % horizontal M&As w/projections w/projections (EW) w/projections (VW) 1996 236 81 33.00 22 31.34 66.91 1997 304 205 67.16 56 29.44 65.55 1998 319 207 64.48 70 35.83 75.47 1999 343 225 67.96 63 28.57 68.15 2000 268 170 65.22 50 28.00 56.11 2001 226 158 71.78 49 35.17 75.33 2002 130 90 70.31 22 23.33 59.94 2003 149 114 76.19 43 38.39 74.65 2004 148 115 79.43 62 54.46 82.92 2005 140 113 80.62 43 36.54 72.40 Total 2,263 1,478 65.31 480 32.48 69.72 Panel A. Mergers, horizontal mergers, and mergers with projections over time Year Number of M&As Number of horizontal M&As % horizontal M&As Number of horizontal M&A % horizontal M&As % horizontal M&As w/projections w/projections (EW) w/projections (VW) 1996 236 81 33.00 22 31.34 66.91 1997 304 205 67.16 56 29.44 65.55 1998 319 207 64.48 70 35.83 75.47 1999 343 225 67.96 63 28.57 68.15 2000 268 170 65.22 50 28.00 56.11 2001 226 158 71.78 49 35.17 75.33 2002 130 90 70.31 22 23.33 59.94 2003 149 114 76.19 43 38.39 74.65 2004 148 115 79.43 62 54.46 82.92 2005 140 113 80.62 43 36.54 72.40 Total 2,263 1,478 65.31 480 32.48 69.72 Panel B: This panel reports summary statistics for the estimated value of managerial projections of operating efficiencies scaled by combined pre-offer market capitalization of the merging firms’ equity. The sample consists of horizontal mergers with available projections of efficiency gains. The computation of the estimated value of efficiency gains is described in Section 3.2. We compute the present value for three horizons: 10 years, perpetuity, and 5 years. We discount annual estimates of efficiency gains using either bidder’s and target’s estimated combined WACC, or 10%, resulting in six estimates of scaled efficiency gain projections. Panel B. Projected operating efficiency gains Estimates value of projected operating efficiency gains Mean (%) Standard deviation (%) Q1 (%) Median (%) Q3 (%) Number of observation 10 years, combined WACC 6.94 11.44 1.58 3.60 7.65 480 Perpetual, combined WACC 12.74 16.81 2.82 6.65 15.78 480 5 years, combined WACC 3.71 6.48 0.79 1.92 3.88 480 10 years, 10% discount rate 10.29 17.36 2.28 5.41 10.53 480 Perpetual, 10% discount rate 16.55 20.08 4.10 9.50 18.94 480 5 years, 10% discount rate 5.56 9.95 1.15 2.74 5.79 480 Panel B. Projected operating efficiency gains Estimates value of projected operating efficiency gains Mean (%) Standard deviation (%) Q1 (%) Median (%) Q3 (%) Number of observation 10 years, combined WACC 6.94 11.44 1.58 3.60 7.65 480 Perpetual, combined WACC 12.74 16.81 2.82 6.65 15.78 480 5 years, combined WACC 3.71 6.48 0.79 1.92 3.88 480 10 years, 10% discount rate 10.29 17.36 2.28 5.41 10.53 480 Perpetual, 10% discount rate 16.55 20.08 4.10 9.50 18.94 480 5 years, 10% discount rate 5.56 9.95 1.15 2.74 5.79 480 Panel B: This panel reports summary statistics for the estimated value of managerial projections of operating efficiencies scaled by combined pre-offer market capitalization of the merging firms’ equity. The sample consists of horizontal mergers with available projections of efficiency gains. The computation of the estimated value of efficiency gains is described in Section 3.2. We compute the present value for three horizons: 10 years, perpetuity, and 5 years. We discount annual estimates of efficiency gains using either bidder’s and target’s estimated combined WACC, or 10%, resulting in six estimates of scaled efficiency gain projections. Panel B. Projected operating efficiency gains Estimates value of projected operating efficiency gains Mean (%) Standard deviation (%) Q1 (%) Median (%) Q3 (%) Number of observation 10 years, combined WACC 6.94 11.44 1.58 3.60 7.65 480 Perpetual, combined WACC 12.74 16.81 2.82 6.65 15.78 480 5 years, combined WACC 3.71 6.48 0.79 1.92 3.88 480 10 years, 10% discount rate 10.29 17.36 2.28 5.41 10.53 480 Perpetual, 10% discount rate 16.55 20.08 4.10 9.50 18.94 480 5 years, 10% discount rate 5.56 9.95 1.15 2.74 5.79 480 Panel B. Projected operating efficiency gains Estimates value of projected operating efficiency gains Mean (%) Standard deviation (%) Q1 (%) Median (%) Q3 (%) Number of observation 10 years, combined WACC 6.94 11.44 1.58 3.60 7.65 480 Perpetual, combined WACC 12.74 16.81 2.82 6.65 15.78 480 5 years, combined WACC 3.71 6.48 0.79 1.92 3.88 480 10 years, 10% discount rate 10.29 17.36 2.28 5.41 10.53 480 Perpetual, 10% discount rate 16.55 20.08 4.10 9.50 18.94 480 5 years, 10% discount rate 5.56 9.95 1.15 2.74 5.79 480 Panle C: This panel presents summary statistics for characteristics of M&A deals and industries in which deals take place. The sample consists of horizontal mergers with available projections of efficiency gains. HHI before merger is the Herfindahl index of sales in the bidder’s and target’s industry, defined as a set of firms belonging to either bidder’s or target’s TNIC industry. Implied change in HHI is defined as the difference between hypothetical HHI, obtained while attributing both the bidder’s and target’s pre-merger sales to the bidder, and pre-merger HHI. Number of rivals is the number of non-merging firms in the bidder’s and/or target’s TNIC industry. Product similarity is textual similarity of bidder’s and target’s product descriptions, as in Hoberg and Phillips (2010a, 2010b). See footnote 5 for description of product similarity. Technology similarity is the overlap in proportions of citations to bidder’s and target’s patents filed in each technology subcategory, as defined in NBER patent data project. See footnote 6 for the calculation of technology similarity. Same BEA is an indicator that equals one if bidder’s and target’s headquarters are located in the same geographical area, as defined by the Bureau of Economic Analysis. Deregulation window is an indicator variable equaling one if an industry went through a deregulation event, as defined in Harford (2005) in the year of M&A announcement or in previous 3 years. Number of competing bidders is the number of competing offers for the same target. % Cash (% Equity) is the percentage of consideration taking the form of cash (equity). Combined CAR [−1, 1] ([−10, 10], [−20, 20]) is cumulative returns during 3 (21, 41) trading days around M&A announcement. Panel C. Deal and industry characteristics Mean Standard deviation Q1 Median Q3 Number of observation HHI before merger 33.45% 21.68% 16.20% 25.89% 46.76% 480 Implied change in HHI 6.96% 12.26% 0.11% 0.96% 7.51% 480 Number of rivals 12.46% 7.44% 5.00% 15.00% 2.00% 480 Product similarity 12.59% 9.41% 6.42% 10.77% 16.21% 480 Technology similarity 7.66% 22.46% 0.00% 0.00% 0.00% 480 Same BEA 0.52 0.50 0.00 1.00 1.00 480 Deregulation window 0.098 0.298 0.000 0.000 0.000 480 Number of competing bidders 1.11 0.38 1.00 1.00 1.00 480 % Cash 21.51 32.85 0.00 0.00 38.03 480 % Equity 70.12 36.14 49.83 84.93 100.00 480 Completed deal 0.873 0.333 1.000 1.000 1.000 480 Combined CAR [−1, 1] 2.05% 7.91% −2.26% 0.93% 5.63% 480 Combined CAR [−10, 10] 2.39% 11.59% −4.38% 1.29% 8.77% 480 Combined CAR [−20, 20] 2.11% 14.82% −5.76% 1.22% 9.70% 480 Bidder change in ROS [−1, 1] 0.014 0.114 −0.010 0.007 0.034 399 Bidder change in ROS [−1, 3] 0.012 0.181 −0.024 0.004 0.038 376 Bidder change in COGS-to-sales [−1, 1] −0.012 0.080 −0.034 −0.006 0.012 399 Bidder change in COGS-to-sales [−1, 3] −0.010 0.094 −0.037 −0.004 0.023 376 Bidder change in SGNA-to-sales [−1, 1] 0.001 0.065 −0.012 0.001 0.017 399 Bidder change in SGNA-to-sales [−1, 3] −0.001 0.079 −0.013 0.003 0.020 376 Panel C. Deal and industry characteristics Mean Standard deviation Q1 Median Q3 Number of observation HHI before merger 33.45% 21.68% 16.20% 25.89% 46.76% 480 Implied change in HHI 6.96% 12.26% 0.11% 0.96% 7.51% 480 Number of rivals 12.46% 7.44% 5.00% 15.00% 2.00% 480 Product similarity 12.59% 9.41% 6.42% 10.77% 16.21% 480 Technology similarity 7.66% 22.46% 0.00% 0.00% 0.00% 480 Same BEA 0.52 0.50 0.00 1.00 1.00 480 Deregulation window 0.098 0.298 0.000 0.000 0.000 480 Number of competing bidders 1.11 0.38 1.00 1.00 1.00 480 % Cash 21.51 32.85 0.00 0.00 38.03 480 % Equity 70.12 36.14 49.83 84.93 100.00 480 Completed deal 0.873 0.333 1.000 1.000 1.000 480 Combined CAR [−1, 1] 2.05% 7.91% −2.26% 0.93% 5.63% 480 Combined CAR [−10, 10] 2.39% 11.59% −4.38% 1.29% 8.77% 480 Combined CAR [−20, 20] 2.11% 14.82% −5.76% 1.22% 9.70% 480 Bidder change in ROS [−1, 1] 0.014 0.114 −0.010 0.007 0.034 399 Bidder change in ROS [−1, 3] 0.012 0.181 −0.024 0.004 0.038 376 Bidder change in COGS-to-sales [−1, 1] −0.012 0.080 −0.034 −0.006 0.012 399 Bidder change in COGS-to-sales [−1, 3] −0.010 0.094 −0.037 −0.004 0.023 376 Bidder change in SGNA-to-sales [−1, 1] 0.001 0.065 −0.012 0.001 0.017 399 Bidder change in SGNA-to-sales [−1, 3] −0.001 0.079 −0.013 0.003 0.020 376 Panle C: This panel presents summary statistics for characteristics of M&A deals and industries in which deals take place. The sample consists of horizontal mergers with available projections of efficiency gains. HHI before merger is the Herfindahl index of sales in the bidder’s and target’s industry, defined as a set of firms belonging to either bidder’s or target’s TNIC industry. Implied change in HHI is defined as the difference between hypothetical HHI, obtained while attributing both the bidder’s and target’s pre-merger sales to the bidder, and pre-merger HHI. Number of rivals is the number of non-merging firms in the bidder’s and/or target’s TNIC industry. Product similarity is textual similarity of bidder’s and target’s product descriptions, as in Hoberg and Phillips (2010a, 2010b). See footnote 5 for description of product similarity. Technology similarity is the overlap in proportions of citations to bidder’s and target’s patents filed in each technology subcategory, as defined in NBER patent data project. See footnote 6 for the calculation of technology similarity. Same BEA is an indicator that equals one if bidder’s and target’s headquarters are located in the same geographical area, as defined by the Bureau of Economic Analysis. Deregulation window is an indicator variable equaling one if an industry went through a deregulation event, as defined in Harford (2005) in the year of M&A announcement or in previous 3 years. Number of competing bidders is the number of competing offers for the same target. % Cash (% Equity) is the percentage of consideration taking the form of cash (equity). Combined CAR [−1, 1] ([−10, 10], [−20, 20]) is cumulative returns during 3 (21, 41) trading days around M&A announcement. Panel C. Deal and industry characteristics Mean Standard deviation Q1 Median Q3 Number of observation HHI before merger 33.45% 21.68% 16.20% 25.89% 46.76% 480 Implied change in HHI 6.96% 12.26% 0.11% 0.96% 7.51% 480 Number of rivals 12.46% 7.44% 5.00% 15.00% 2.00% 480 Product similarity 12.59% 9.41% 6.42% 10.77% 16.21% 480 Technology similarity 7.66% 22.46% 0.00% 0.00% 0.00% 480 Same BEA 0.52 0.50 0.00 1.00 1.00 480 Deregulation window 0.098 0.298 0.000 0.000 0.000 480 Number of competing bidders 1.11 0.38 1.00 1.00 1.00 480 % Cash 21.51 32.85 0.00 0.00 38.03 480 % Equity 70.12 36.14 49.83 84.93 100.00 480 Completed deal 0.873 0.333 1.000 1.000 1.000 480 Combined CAR [−1, 1] 2.05% 7.91% −2.26% 0.93% 5.63% 480 Combined CAR [−10, 10] 2.39% 11.59% −4.38% 1.29% 8.77% 480 Combined CAR [−20, 20] 2.11% 14.82% −5.76% 1.22% 9.70% 480 Bidder change in ROS [−1, 1] 0.014 0.114 −0.010 0.007 0.034 399 Bidder change in ROS [−1, 3] 0.012 0.181 −0.024 0.004 0.038 376 Bidder change in COGS-to-sales [−1, 1] −0.012 0.080 −0.034 −0.006 0.012 399 Bidder change in COGS-to-sales [−1, 3] −0.010 0.094 −0.037 −0.004 0.023 376 Bidder change in SGNA-to-sales [−1, 1] 0.001 0.065 −0.012 0.001 0.017 399 Bidder change in SGNA-to-sales [−1, 3] −0.001 0.079 −0.013 0.003 0.020 376 Panel C. Deal and industry characteristics Mean Standard deviation Q1 Median Q3 Number of observation HHI before merger 33.45% 21.68% 16.20% 25.89% 46.76% 480 Implied change in HHI 6.96% 12.26% 0.11% 0.96% 7.51% 480 Number of rivals 12.46% 7.44% 5.00% 15.00% 2.00% 480 Product similarity 12.59% 9.41% 6.42% 10.77% 16.21% 480 Technology similarity 7.66% 22.46% 0.00% 0.00% 0.00% 480 Same BEA 0.52 0.50 0.00 1.00 1.00 480 Deregulation window 0.098 0.298 0.000 0.000 0.000 480 Number of competing bidders 1.11 0.38 1.00 1.00 1.00 480 % Cash 21.51 32.85 0.00 0.00 38.03 480 % Equity 70.12 36.14 49.83 84.93 100.00 480 Completed deal 0.873 0.333 1.000 1.000 1.000 480 Combined CAR [−1, 1] 2.05% 7.91% −2.26% 0.93% 5.63% 480 Combined CAR [−10, 10] 2.39% 11.59% −4.38% 1.29% 8.77% 480 Combined CAR [−20, 20] 2.11% 14.82% −5.76% 1.22% 9.70% 480 Bidder change in ROS [−1, 1] 0.014 0.114 −0.010 0.007 0.034 399 Bidder change in ROS [−1, 3] 0.012 0.181 −0.024 0.004 0.038 376 Bidder change in COGS-to-sales [−1, 1] −0.012 0.080 −0.034 −0.006 0.012 399 Bidder change in COGS-to-sales [−1, 3] −0.010 0.094 −0.037 −0.004 0.023 376 Bidder change in SGNA-to-sales [−1, 1] 0.001 0.065 −0.012 0.001 0.017 399 Bidder change in SGNA-to-sales [−1, 3] −0.001 0.079 −0.013 0.003 0.020 376 Panel D: This panel presents characteristics of bidders in horizontal M&As with available projections of operating efficiency gains. MV of equity is the product of share price times the number of shares outstanding in the last trading day prior to M&A announcement. Sales are revenues. Assets are book assets. Tobin’s Q is the ratio of pseudo-market value to book value of assets. Pseudo-market value is the sum of common equity market value and book value of common equity and long-term debt. Leverage is the ratio of long-term debt to pseudo-market value of assets. Profitability is the ratio of EBITDA to sales. Capital stock is the ratio of capital stock to assets. Capital stock is computed using perpetual inventory method, which is described in footnote 7. Stock price runup is equity return over 11-month period ending 1 month prior to M&A announcement. Daily return volatility is the standard deviation of daily returns during the 11-month period ending 1 month prior to M&A announcement. Number of analysts is the number of equity research analysts following the firm in the quarter prior to M&A announcement, as reported in Thomson Reuters I/B/E/S. % Institutional investors is the proportion of firm’s equity held by 13-F filers at the end of the quarter prior to M&A announcement, as reported in Thomson Reuters 13-F database. Number of earnings guidances is the number of earnings guidance events in the year prior to merger announcement, as reported in I/B/E/S. E-index is the entrenchment index from Bebchuk, Cohen, and Ferrell (2009). All accounting variables are from the fiscal year preceding the year of the M&A announcement. Panel D. Bidder characteristics Mean Standard deviation Q1 Median Q3 Number of observation MV of equity 11,416 24,755 795 3,132 9,786 480 Sales 5,390 11,433 387 1,522 5,670 480 Assets 21,429 68,183 906 3,629 15,043 480 Tobin’s Q 1.56 2.44 0.29 0.94 1.81 480 Leverage 0.26 0.21 0.09 0.19 0.39 480 Profitability 0.07 1.03 0.03 0.06 0.13 480 Capital stock 0.303 0.448 0.131 0.233 0.601 480 Stock price runup 23.49% 47.73% −3.53% 19.57% 42.76% 480 Daily return volatility 2.43% 1.34% 1.55% 2.06% 2.86% 480 Number of analysts 5.55 5.40 0.00 5.00 9.00 480 % Institutional investors 52.98% 25.13% 30.68% 54.23% 74.05% 480 Number of earnings guidances 1.47 2.13 0.00 1.00 2.00 480 E-index 2.59 1.37 2.00 3.00 4.00 126 Panel D. Bidder characteristics Mean Standard deviation Q1 Median Q3 Number of observation MV of equity 11,416 24,755 795 3,132 9,786 480 Sales 5,390 11,433 387 1,522 5,670 480 Assets 21,429 68,183 906 3,629 15,043 480 Tobin’s Q 1.56 2.44 0.29 0.94 1.81 480 Leverage 0.26 0.21 0.09 0.19 0.39 480 Profitability 0.07 1.03 0.03 0.06 0.13 480 Capital stock 0.303 0.448 0.131 0.233 0.601 480 Stock price runup 23.49% 47.73% −3.53% 19.57% 42.76% 480 Daily return volatility 2.43% 1.34% 1.55% 2.06% 2.86% 480 Number of analysts 5.55 5.40 0.00 5.00 9.00 480 % Institutional investors 52.98% 25.13% 30.68% 54.23% 74.05% 480 Number of earnings guidances 1.47 2.13 0.00 1.00 2.00 480 E-index 2.59 1.37 2.00 3.00 4.00 126 Panel D: This panel presents characteristics of bidders in horizontal M&As with available projections of operating efficiency gains. MV of equity is the product of share price times the number of shares outstanding in the last trading day prior to M&A announcement. Sales are revenues. Assets are book assets. Tobin’s Q is the ratio of pseudo-market value to book value of assets. Pseudo-market value is the sum of common equity market value and book value of common equity and long-term debt. Leverage is the ratio of long-term debt to pseudo-market value of assets. Profitability is the ratio of EBITDA to sales. Capital stock is the ratio of capital stock to assets. Capital stock is computed using perpetual inventory method, which is described in footnote 7. Stock price runup is equity return over 11-month period ending 1 month prior to M&A announcement. Daily return volatility is the standard deviation of daily returns during the 11-month period ending 1 month prior to M&A announcement. Number of analysts is the number of equity research analysts following the firm in the quarter prior to M&A announcement, as reported in Thomson Reuters I/B/E/S. % Institutional investors is the proportion of firm’s equity held by 13-F filers at the end of the quarter prior to M&A announcement, as reported in Thomson Reuters 13-F database. Number of earnings guidances is the number of earnings guidance events in the year prior to merger announcement, as reported in I/B/E/S. E-index is the entrenchment index from Bebchuk, Cohen, and Ferrell (2009). All accounting variables are from the fiscal year preceding the year of the M&A announcement. Panel D. Bidder characteristics Mean Standard deviation Q1 Median Q3 Number of observation MV of equity 11,416 24,755 795 3,132 9,786 480 Sales 5,390 11,433 387 1,522 5,670 480 Assets 21,429 68,183 906 3,629 15,043 480 Tobin’s Q 1.56 2.44 0.29 0.94 1.81 480 Leverage 0.26 0.21 0.09 0.19 0.39 480 Profitability 0.07 1.03 0.03 0.06 0.13 480 Capital stock 0.303 0.448 0.131 0.233 0.601 480 Stock price runup 23.49% 47.73% −3.53% 19.57% 42.76% 480 Daily return volatility 2.43% 1.34% 1.55% 2.06% 2.86% 480 Number of analysts 5.55 5.40 0.00 5.00 9.00 480 % Institutional investors 52.98% 25.13% 30.68% 54.23% 74.05% 480 Number of earnings guidances 1.47 2.13 0.00 1.00 2.00 480 E-index 2.59 1.37 2.00 3.00 4.00 126 Panel D. Bidder characteristics Mean Standard deviation Q1 Median Q3 Number of observation MV of equity 11,416 24,755 795 3,132 9,786 480 Sales 5,390 11,433 387 1,522 5,670 480 Assets 21,429 68,183 906 3,629 15,043 480 Tobin’s Q 1.56 2.44 0.29 0.94 1.81 480 Leverage 0.26 0.21 0.09 0.19 0.39 480 Profitability 0.07 1.03 0.03 0.06 0.13 480 Capital stock 0.303 0.448 0.131 0.233 0.601 480 Stock price runup 23.49% 47.73% −3.53% 19.57% 42.76% 480 Daily return volatility 2.43% 1.34% 1.55% 2.06% 2.86% 480 Number of analysts 5.55 5.40 0.00 5.00 9.00 480 % Institutional investors 52.98% 25.13% 30.68% 54.23% 74.05% 480 Number of earnings guidances 1.47 2.13 0.00 1.00 2.00 480 E-index 2.59 1.37 2.00 3.00 4.00 126 Panel E: The panel presents characteristics of targets in horizontal M&As with available projections of operating efficiency gains. See Panel D for variable definitions. Panel E. Target characteristics Mean Standard deviation Q1 Median Q3 Number of observation MV of equity 2,714 4,992 189 651 2,368 480 Sales 2,453 5,730 106 398 2,039 480 Assets 7,268 26,084 355 1,084 4,282 480 Tobin’s Q 1.35 2.25 0.25 0.82 1.48 480 Leverage 0.29 0.27 0.11 0.24 0.46 480 Profitability 0.02 1.49 −0.01 0.07 0.14 480 Capital stock 0.281 0.440 0.110 0.198 0.491 480 Stock price runup 17.00% 54.01% −14.28% 11.58% 38.18% 480 Daily return volatility 2.85% 1.69% 1.73% 2.30% 3.39% 480 Number of analysts 3.99 4.58 0.00 3.00 7.00 480 % Institutional investors 49.16% 27.20% 27.50% 50.89% 70.19% 480 Number of earnings guidances 0.95 1.54 0.00 0.00 1.00 480 E-index 2.77 1.15 2.00 3.00 3.00 89 Panel E. Target characteristics Mean Standard deviation Q1 Median Q3 Number of observation MV of equity 2,714 4,992 189 651 2,368 480 Sales 2,453 5,730 106 398 2,039 480 Assets 7,268 26,084 355 1,084 4,282 480 Tobin’s Q 1.35 2.25 0.25 0.82 1.48 480 Leverage 0.29 0.27 0.11 0.24 0.46 480 Profitability 0.02 1.49 −0.01 0.07 0.14 480 Capital stock 0.281 0.440 0.110 0.198 0.491 480 Stock price runup 17.00% 54.01% −14.28% 11.58% 38.18% 480 Daily return volatility 2.85% 1.69% 1.73% 2.30% 3.39% 480 Number of analysts 3.99 4.58 0.00 3.00 7.00 480 % Institutional investors 49.16% 27.20% 27.50% 50.89% 70.19% 480 Number of earnings guidances 0.95 1.54 0.00 0.00 1.00 480 E-index 2.77 1.15 2.00 3.00 3.00 89 Panel E: The panel presents characteristics of targets in horizontal M&As with available projections of operating efficiency gains. See Panel D for variable definitions. Panel E. Target characteristics Mean Standard deviation Q1 Median Q3 Number of observation MV of equity 2,714 4,992 189 651 2,368 480 Sales 2,453 5,730 106 398 2,039 480 Assets 7,268 26,084 355 1,084 4,282 480 Tobin’s Q 1.35 2.25 0.25 0.82 1.48 480 Leverage 0.29 0.27 0.11 0.24 0.46 480 Profitability 0.02 1.49 −0.01 0.07 0.14 480 Capital stock 0.281 0.440 0.110 0.198 0.491 480 Stock price runup 17.00% 54.01% −14.28% 11.58% 38.18% 480 Daily return volatility 2.85% 1.69% 1.73% 2.30% 3.39% 480 Number of analysts 3.99 4.58 0.00 3.00 7.00 480 % Institutional investors 49.16% 27.20% 27.50% 50.89% 70.19% 480 Number of earnings guidances 0.95 1.54 0.00 0.00 1.00 480 E-index 2.77 1.15 2.00 3.00 3.00 89 Panel E. Target characteristics Mean Standard deviation Q1 Median Q3 Number of observation MV of equity 2,714 4,992 189 651 2,368 480 Sales 2,453 5,730 106 398 2,039 480 Assets 7,268 26,084 355 1,084 4,282 480 Tobin’s Q 1.35 2.25 0.25 0.82 1.48 480 Leverage 0.29 0.27 0.11 0.24 0.46 480 Profitability 0.02 1.49 −0.01 0.07 0.14 480 Capital stock 0.281 0.440 0.110 0.198 0.491 480 Stock price runup 17.00% 54.01% −14.28% 11.58% 38.18% 480 Daily return volatility 2.85% 1.69% 1.73% 2.30% 3.39% 480 Number of analysts 3.99 4.58 0.00 3.00 7.00 480 % Institutional investors 49.16% 27.20% 27.50% 50.89% 70.19% 480 Number of earnings guidances 0.95 1.54 0.00 0.00 1.00 480 E-index 2.77 1.15 2.00 3.00 3.00 89 We restrict the initial sample of M&As to acquisitions of assets, acquisitions of majority interest, or mergers in which a bidder holds less than 50% of target’s common stock on the offer date and the deal has come to a resolution (i.e., completed or withdrawn). Our initial sample contains 2,263 M&A announcements. A horizontal merger is defined as a merger between two firms that belong to the same TNIC in the Hoberg–Phillips data library, i.e., firms that have sufficiently similar product descriptions in their 10-K filings.5 Around 65% of mergers in our sample (1,478) are classified as horizontal. In about one-third of those (480) we are able to identify and quantify managerial projections of operating efficiency gains. Importantly, efficiency gain forecasts are available in mergers involving targets that amount to 70% of aggregate target firms’ assets. Therefore, while the majority of deals have no insiders’ projections of operating efficiency gains, those that involve relatively large targets and, thus, are likely to have the most significant impact on related firms, are typically included in our sample. Panel B reports summary statistics of the estimated value of projected operating efficiency gains scaled by bidder’s and target’s combined pre-merger market capitalization, for three horizons of efficiency gain realizations—10-year, perpetual, and 5-year, and using two discount rate estimates—bidder and target’s combined WACC or 10%. The mean (median) value of projected operating efficiency gains over 10 years using bidder’s and target’s combined WACC as the discount rate—the baseline measure that we use in most empirical tests—as a fraction of bidder’s and target’s combined pre-merger market capitalization, is 6.9% (3.6%). These values are somewhat higher—10.3% (5.4%)—when we discount annual efficiency projections using 10% discount rate. Mean (median) operating efficiency gains are larger—12.7% (6.7%) for an infinite horizon, and smaller—3.7% (1.9%) for a 5-year horizon. There is substantial variation in projected gains, as indicated by the spread between the first and third quartiles (1.6% and 7.7%, respectively, for the baseline measure). In unreported analysis, we also examine the distribution of projected efficiency gains as a proportion of target’s pre-merger value. The mean (median) efficiency gain projection over 10-year horizon discounted using bidder’s and target’s combined WACC and scaled by target’s value is 35% (19%), while the corresponding mean (median) values for perpetual and 5-year horizons are 65% and 19% (35% and 10%), respectively. Panel C presents summary statistics of deal characteristics for the 480 mergers with operating efficiency projections. Pre-merger concentration in the merging firms’ industry, which is defined as the set of all firms belonging to either bidder’s or target’s TNIC industry, computed in the year prior to the deal announcement, is quite high, as evident from the mean (median) Herfindahl index (HHI hereafter) of 33% (26%). The mean implied change in HHI, defined as the difference between hypothetical HHI, obtained while attributing both the bidder’s and target’s pre-merger sales to the bidder on the one hand, and actual pre-merger HHI on the other hand, is sizable (7%), while the median is much smaller (1%). The mean (median) number of non-merging firms operating in bidder’s and/or target’s TNIC industry is 12 (15). The degree of product similarity between bidders and targets, measured as textual similarity of their product descriptions, is quite high—the mean is 13% and the median is 11%, whereas the unconditional mean pairwise similarity among all pairs of Compustat firms is 2%. We follow Jaffe (1986) and Bena and Li (2014) and measure the degree of technological similarity between a bidder and target as the overlap in proportions of citations to the two firms’ patents filed in each technological subcategory, as defined in the NBER Patent Citations Data project.6 The mean technological similarity between bidder and target is 8%, while the median is zero, since in more than half of deals either bidder or target have not filed any patents. Bidders and targets are headquartered in the same Bureau of Economic Analysis geographical region in 52% of deals. Every tenth deal occurs following deregulation, defined based on industry-level deregulation events in Harford (2005). The mean number of competing bidders is 1.11, and in the majority of deals there is only one bidder. Most deals involve exchange of equity—the mean (median) proportion of offer price paid in equity is 70% (85%). Eighty-seven percent of announced deals are completed. When measuring merging firms’ cumulative announcement returns (CARs), we use three return windows. The first one includes three trading days around an acquisition announcement ([−1, 1]). The other two windows are broader and include 21 and 41 trading days around the announcement ([−10, 10] and [−20, 20], respectively). Bidder’s and target’s daily abnormal returns are computed using the market-model-based expected return as a benchmark. The reported combined CARs are weighted average bidder’s and target’s CARs, with the weights corresponding to pre-merger market capitalizations. Consistent with the vast merger literature, the mean (median) combined (i.e., market-cap-weighted average bidder’s and target’s) announcement return is around 2% (1%). For completed mergers, we measure post-merger change in profitability as the difference between the merged firm’s return on sales (ROS), defined as EBITDA-to-sales ratio in the year after the deal completion, and the bidder’s EBITDA-to-sales ratio in the year prior to the merger announcement. An alternative measure is the difference between the merged firm’s average ROS in the 3 years after the deal completion and the bidder’s ROS in the year prior to the merger announcement. We measure post-merger changes in COGS-to-sales ratio and in SG&A-to-sales ratio similarly. Consistent with the positive mean and median combined announcement returns, mergers increase profitability and reduce COGS-to-sales ratio by over one percentage point on average. However, mergers do not have a meaningful effect on mean SG&A-to-sales ratio. Panels D and E of Table I present summary statistics of bidders’ and targets’ characteristics, respectively. Bidders tend to be much larger than targets in terms of market capitalization, sales, and book assets—the means are 11.4, 5.4, and 21.4 billion for bidders, respectively, while the corresponding means for targets are 2.7, 2.5, and 7.3 billion. Bidders tend to have larger Tobin’s Q than targets (1.56 and 1.35 on average, respectively). Bidders have lower leverage than targets (0.26 and 0.29 on average, respectively) and higher profitability than targets (0.07 and 0.02 on average, respectively). Bidders and targets have similar mean capital-stock-to-assets ratios, computed using a variant of perpetual inventory method (e.g., Hall, Jaffe, and Trajtenberg, 2005; Hirshleifer, Hsu, and Li, 2013)—0.30 and 0.28, respectively.7 Bidders have higher mean (median) stock price runup, defined as raw stock return over 11 months period ending 1 month prior to merger announcement, than targets—23% (20%) for the former compared with 17% (12%) for the latter. Bidder shares tend to be less volatile on average than those of targets (mean daily return volatilities are 2.4% and 2.8%, respectively). Bidders are followed by more analysts than targets (5.5 vs. 4 on average), are somewhat more likely to be held by institutional investors (53% vs. 49% on average), and issue earnings guidances more often (1.47 in the year prior to the merger announcement for bidders on average, compared with 0.95 for targets). There is no significant difference between bidders’ and targets’ corporate governance, as measured by Bebchuk, Cohen, and Ferrell (2009) entrenchment index. 4. Projected and Realized Operating Efficiency Gains In this section, we first examine some of the factors related to the propensity of merging firms’ insiders to provide operating efficiency gain forecasts and then analyze the relation between projected and realized efficiency gains, while controlling for possible self-selection in the decision to disclose projections. 4.1 Determinants of the Propensity to Provide Efficiency Gain Projections Within the sample of 1,478 horizontal mergers, we estimate probit regressions in which the dependent variable is an indicator equalling 1 for 480 deals with available operating efficiency gain projections and equaling zero otherwise. The set of independent variables includes various deal characteristics, as well as bidders’ and targets’ characteristics that may be related to insiders’ decision to provide forecasts. Important for the analysis of the relation between efficiency gain projections and realized gains, the set of independent variables also includes instruments that are likely to be related to the availability of projections but are less likely to be related to merger announcement returns and post-merger changes in profitability and cost ratios. Our main instruments are bidder’s and target’s number of earnings guidances during the year ending 1 month prior to merger announcement. The frequency of earnings guidance events is related to firms’ disclosure styles and is unlikely to be directly related to post-merger valuation and performance, i.e., it is likely to satisfy the exclusion restriction. In one of the specifications, we also use additional instruments for the availability of operating efficiency gain projections: the number of analysts following the bidder and target and the percentage of bidder’s and target’s common shares held by institutional investors (13-F filers). While these instruments are less likely to satisfy the exclusion restriction, as they may be related to post-merger valuation and performance through, e.g., improved monitoring, they may still be helpful in controlling for self-selection when estimating the relations between insiders’ operating efficiency gain projections on the one hand and outcomes for merging firms and related firms on the other hand. The reason is that estimation of a selection model does not necessarily require satisfaction of exclusion restrictions for identification purposes, since the model can be identified by non-linearity (e.g., Heckman and Navarro-Lozano, 2004; Li and Prabhala, 2005). Additional variables that may be related to the availability of efficiency gain projections include: pre-merger industry structure, proxied by the number of bidder’s and target’s product market rivals, defined as all non-merging firms in either bidder’s or target’s TNIC industry in the pre-merger year; post-merger implied change in bidder’s and target’s HHI; various facets of complementarity between bidder and target—the degree of product similarity, the degree of technological similarity, and the geographical overlap indicator; deal characteristics, including the number of competing bidders, the percentage of equity and cash considerations in the M&A offer, and the numbers of bidder’s and target’s identifiable customers and suppliers; and, finally, bidder’s and target’s characteristics—the logarithm of pre-merger market capitalization, Tobin’s Q, leverage, profitability, capital stock, pre-merger stock price runup, pre-merger return volatility, and entrenchment index (if present). The estimates of probit regressions of insiders’ propensity to provide efficiency gain projections are presented in Table II. Table II. Propensity to disclose operating efficiency projections This table presents estimates of probit regressions estimated in a sample of all horizontal M&As, as defined in Panel A of Table I, in which the independent variable equals one if a merger has efficiency gains projections available and equals zero otherwise. See Panels B–E of Table I for variable definitions. Standard errors are reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. R-squared is McFadden R-squared. Dependent variable Availability of projections Availability of projections Intercept −1.957*** −1.576*** (−5.26) (−3.93) Bidder # earnings guidances 0.047** 0.039* (2.05) (1.66) Target # earnings guidances 0.086*** 0.065** (2.70) (1.99) Bidder # analysts 0.002 (0.15) Target # analysts 0.061*** (3.91) Bidder % inst. investors 0.047 (0.24) Target % inst. investors 0.155 (0.74) Number of rivals −0.342* −0.297* (−1.92) (−1.71) Implied change in HHI 0.802** 0.777** (2.09) (2.01) Product similarity −0.763* −0.762* (−1.86) (−1.83) Technology similarity −0.275 −0.273 (−1.41) (−1.38) Same BEA 0.124 0.118 (1.48) (1.37) Deregulation 0.052 0.052 (0.34) (0.32) Number of competing bidders −0.076 −0.088 (−0.63) (−0.71) % Cash 0.037 −0.061 (0.14) (−0.22) % Equity 0.634** 0.529** (2.47) (2.02) Number of customers 0.072 0.087 (0.57) (0.61) Number of suppliers −0.131 0.022 (−0.44) (0.16) Log(bidder MV equity) 0.255*** 0.164*** (7.66) (3.96) Log(target MV equity) 0.466*** 0.379*** (10.69) (7.59) Bidder Tobin’s Q −0.047** −0.042* (−2.13) (−1.92) Target Tobin’s Q −0.029 −0.034 (−1.32) (−1.52) Bidder leverage 0.280 0.358 (1.14) (1.44) Target leverage 0.477** 0.491** (2.18) (2.22) Bidder profitability −0.075 −0.047 (−0.94) (−0.57) Target profitability 0.040 0.038 (0.86) (0.80) Bidder capital stock −0.275 −0.352 (−0.77) (−0.98) Target capital stock 0.286 0.053 (0.74) (0.13) Bidder runup −0.211** −0.231** (−2.42) (−2.54) Target runup −0.028 0.006 (−0.33) (0.07) Bidder return volatility −3.988 −3.092 (−0.88) (−0.68) Target return volatility −0.178 −0.743 (−0.05) (−0.21) Bidder E-index presence 0.010 0.066 (0.05) (0.33) Target E-index presence −0.588** −0.550* (−2.04) (−1.90) Bidder E-index×presence 0.101 0.093 (1.60) (1.45) Target E-index×presence 0.239** 0.245** (2.37) (2.42) Number of observation 1,478 1,478 R-squared 0.243 0.251 Dependent variable Availability of projections Availability of projections Intercept −1.957*** −1.576*** (−5.26) (−3.93) Bidder # earnings guidances 0.047** 0.039* (2.05) (1.66) Target # earnings guidances 0.086*** 0.065** (2.70) (1.99) Bidder # analysts 0.002 (0.15) Target # analysts 0.061*** (3.91) Bidder % inst. investors 0.047 (0.24) Target % inst. investors 0.155 (0.74) Number of rivals −0.342* −0.297* (−1.92) (−1.71) Implied change in HHI 0.802** 0.777** (2.09) (2.01) Product similarity −0.763* −0.762* (−1.86) (−1.83) Technology similarity −0.275 −0.273 (−1.41) (−1.38) Same BEA 0.124 0.118 (1.48) (1.37) Deregulation 0.052 0.052 (0.34) (0.32) Number of competing bidders −0.076 −0.088 (−0.63) (−0.71) % Cash 0.037 −0.061 (0.14) (−0.22) % Equity 0.634** 0.529** (2.47) (2.02) Number of customers 0.072 0.087 (0.57) (0.61) Number of suppliers −0.131 0.022 (−0.44) (0.16) Log(bidder MV equity) 0.255*** 0.164*** (7.66) (3.96) Log(target MV equity) 0.466*** 0.379*** (10.69) (7.59) Bidder Tobin’s Q −0.047** −0.042* (−2.13) (−1.92) Target Tobin’s Q −0.029 −0.034 (−1.32) (−1.52) Bidder leverage 0.280 0.358 (1.14) (1.44) Target leverage 0.477** 0.491** (2.18) (2.22) Bidder profitability −0.075 −0.047 (−0.94) (−0.57) Target profitability 0.040 0.038 (0.86) (0.80) Bidder capital stock −0.275 −0.352 (−0.77) (−0.98) Target capital stock 0.286 0.053 (0.74) (0.13) Bidder runup −0.211** −0.231** (−2.42) (−2.54) Target runup −0.028 0.006 (−0.33) (0.07) Bidder return volatility −3.988 −3.092 (−0.88) (−0.68) Target return volatility −0.178 −0.743 (−0.05) (−0.21) Bidder E-index presence 0.010 0.066 (0.05) (0.33) Target E-index presence −0.588** −0.550* (−2.04) (−1.90) Bidder E-index×presence 0.101 0.093 (1.60) (1.45) Target E-index×presence 0.239** 0.245** (2.37) (2.42) Number of observation 1,478 1,478 R-squared 0.243 0.251 Table II. Propensity to disclose operating efficiency projections This table presents estimates of probit regressions estimated in a sample of all horizontal M&As, as defined in Panel A of Table I, in which the independent variable equals one if a merger has efficiency gains projections available and equals zero otherwise. See Panels B–E of Table I for variable definitions. Standard errors are reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. R-squared is McFadden R-squared. Dependent variable Availability of projections Availability of projections Intercept −1.957*** −1.576*** (−5.26) (−3.93) Bidder # earnings guidances 0.047** 0.039* (2.05) (1.66) Target # earnings guidances 0.086*** 0.065** (2.70) (1.99) Bidder # analysts 0.002 (0.15) Target # analysts 0.061*** (3.91) Bidder % inst. investors 0.047 (0.24) Target % inst. investors 0.155 (0.74) Number of rivals −0.342* −0.297* (−1.92) (−1.71) Implied change in HHI 0.802** 0.777** (2.09) (2.01) Product similarity −0.763* −0.762* (−1.86) (−1.83) Technology similarity −0.275 −0.273 (−1.41) (−1.38) Same BEA 0.124 0.118 (1.48) (1.37) Deregulation 0.052 0.052 (0.34) (0.32) Number of competing bidders −0.076 −0.088 (−0.63) (−0.71) % Cash 0.037 −0.061 (0.14) (−0.22) % Equity 0.634** 0.529** (2.47) (2.02) Number of customers 0.072 0.087 (0.57) (0.61) Number of suppliers −0.131 0.022 (−0.44) (0.16) Log(bidder MV equity) 0.255*** 0.164*** (7.66) (3.96) Log(target MV equity) 0.466*** 0.379*** (10.69) (7.59) Bidder Tobin’s Q −0.047** −0.042* (−2.13) (−1.92) Target Tobin’s Q −0.029 −0.034 (−1.32) (−1.52) Bidder leverage 0.280 0.358 (1.14) (1.44) Target leverage 0.477** 0.491** (2.18) (2.22) Bidder profitability −0.075 −0.047 (−0.94) (−0.57) Target profitability 0.040 0.038 (0.86) (0.80) Bidder capital stock −0.275 −0.352 (−0.77) (−0.98) Target capital stock 0.286 0.053 (0.74) (0.13) Bidder runup −0.211** −0.231** (−2.42) (−2.54) Target runup −0.028 0.006 (−0.33) (0.07) Bidder return volatility −3.988 −3.092 (−0.88) (−0.68) Target return volatility −0.178 −0.743 (−0.05) (−0.21) Bidder E-index presence 0.010 0.066 (0.05) (0.33) Target E-index presence −0.588** −0.550* (−2.04) (−1.90) Bidder E-index×presence 0.101 0.093 (1.60) (1.45) Target E-index×presence 0.239** 0.245** (2.37) (2.42) Number of observation 1,478 1,478 R-squared 0.243 0.251 Dependent variable Availability of projections Availability of projections Intercept −1.957*** −1.576*** (−5.26) (−3.93) Bidder # earnings guidances 0.047** 0.039* (2.05) (1.66) Target # earnings guidances 0.086*** 0.065** (2.70) (1.99) Bidder # analysts 0.002 (0.15) Target # analysts 0.061*** (3.91) Bidder % inst. investors 0.047 (0.24) Target % inst. investors 0.155 (0.74) Number of rivals −0.342* −0.297* (−1.92) (−1.71) Implied change in HHI 0.802** 0.777** (2.09) (2.01) Product similarity −0.763* −0.762* (−1.86) (−1.83) Technology similarity −0.275 −0.273 (−1.41) (−1.38) Same BEA 0.124 0.118 (1.48) (1.37) Deregulation 0.052 0.052 (0.34) (0.32) Number of competing bidders −0.076 −0.088 (−0.63) (−0.71) % Cash 0.037 −0.061 (0.14) (−0.22) % Equity 0.634** 0.529** (2.47) (2.02) Number of customers 0.072 0.087 (0.57) (0.61) Number of suppliers −0.131 0.022 (−0.44) (0.16) Log(bidder MV equity) 0.255*** 0.164*** (7.66) (3.96) Log(target MV equity) 0.466*** 0.379*** (10.69) (7.59) Bidder Tobin’s Q −0.047** −0.042* (−2.13) (−1.92) Target Tobin’s Q −0.029 −0.034 (−1.32) (−1.52) Bidder leverage 0.280 0.358 (1.14) (1.44) Target leverage 0.477** 0.491** (2.18) (2.22) Bidder profitability −0.075 −0.047 (−0.94) (−0.57) Target profitability 0.040 0.038 (0.86) (0.80) Bidder capital stock −0.275 −0.352 (−0.77) (−0.98) Target capital stock 0.286 0.053 (0.74) (0.13) Bidder runup −0.211** −0.231** (−2.42) (−2.54) Target runup −0.028 0.006 (−0.33) (0.07) Bidder return volatility −3.988 −3.092 (−0.88) (−0.68) Target return volatility −0.178 −0.743 (−0.05) (−0.21) Bidder E-index presence 0.010 0.066 (0.05) (0.33) Target E-index presence −0.588** −0.550* (−2.04) (−1.90) Bidder E-index×presence 0.101 0.093 (1.60) (1.45) Target E-index×presence 0.239** 0.245** (2.37) (2.42) Number of observation 1,478 1,478 R-squared 0.243 0.251 The first column includes only the main instruments for operating efficiency gain projections provision—the number of bidder’s and target’s earnings guidance events—along with other industry, deal, bidder, and target characteristics, while the second column includes in addition analyst-coverage-based and institutional-holdings-based instruments. Consistent with the merging firms’ disclosure styles affecting the availability of efficiency gain forecasts, bidder’s and target’s number of earnings guidance events are both positively associated with projections’ availability. The relation is significant at 5% and 10% levels for the number of bidder’s earnings guidance events in both specifications, and at 1% and 5% levels for the number of target’s earnings guidance events. These relations are also economically meaningful: when all independent variables take their mean values, a one-standard-deviation increase in the number of bidder’s (target’s) earnings guidance events—2.13 (1.54)—is associated with a 3.8 (5) percentage point increase in the likelihood of insiders providing merger efficiency gain projections, which corresponds to 12% (15%) of the mean proportion of mergers with projections. These results suggest that the instrument satisfies the relevance restriction. Among the other instruments in the second column, only the target’s number of analysts is statistically and economically significant. The availability of insiders’ projections is related to pre-merger structure of the bidder’s and target’s industry and to post-merger anticipated change in it. It is negatively associated with the number of bidder’s and target’s product market rivals (significant at 10% level), and is positively associated with post-merger implied change in HHI in the bidder’s and target’s industry (significant at 5% level). Both these findings suggest that insiders are more likely to provide projections when the industry is more concentrated and when the expected change in industry concentration following the merger is larger, i.e., in cases in which horizontal mergers may result in increased market power. The propensity to provide projections is decreasing in product similarity between the bidder and target, suggesting that managers attempt to justify mergers in which operating efficiencies may be more difficult to obtain; is increasing in the proportion of equity in offering price, consistent with bidder managers attempting to mitigate possible concerns regarding bidder’s equity being overvalued, but is decreasing in bidder’s Tobin’s Q and in its stock price runup, casting doubt on this interpretation; is increasing in both the bidder’s and target’s market capitalizations, consistent with efficiency gain projections being more frequent in larger deals; and is increasing in target’s leverage and decreasing in target’s entrenchment index (where available), suggesting that insiders are more likely to disclose efficiency gain forecasts in bids for highly indebted and poorly governed targets. 4.2 Are Operating Efficiency Gain Forecasts Related to Realized Merger Gains? Our empirical analysis rests on two important premises. The first one is that insiders’ projections of operating efficiency gains are economically meaningful—i.e., informative about expected value and performance implications of M&As. The second one is that managerial projections in fact reflect anticipated operating efficiencies, as opposed to disguising other potential effects of mergers, such as changes in market power. To examine the validity of the first premise, we analyze the relation between efficiency gain forecasts on the one hand and merging firms’ announcement returns and post-merger operating performance on the other hand. We begin by estimating regressions in which the dependent variable is merging firms’ value-weighted mean cumulative abnormal return over the [−1, 1], [−10, 10], or [−20, 20] announcement windows. The main independent variable is the value of projected efficiency gains over 10-year horizon, estimated using bidder’s and target’s combined WACC as the discount rate, and scaled by the merging firms’ combined market capitalization prior to merger announcement. We also control for other factors that may be related to post-merger changes in merging firms’ values: their industry structure, their product and technology similarity, and geographical overlap, as well as various deal, bidder, and target characteristics. Lastly, we account for potential endogeneity of insiders’ decision to provide efficiency gain forecasts by applying the Heckman (1979) correction for self-selection, i.e., by including the inverse Mills ratios from the first-stage probit regression reported in Table II.8 The estimates of OLS regressions with year-fixed effects are reported in Panel A of Table III. Table III. Projected operating efficiency gains and merging firms’ valuation and performance This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and target’s combined CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20] in Panel A and bidder’s change in ROS [−1, 1], ROS [−1, 3], COGS-to-sales [−1, 1], COGS-to-sales [−1, 3], SG&A-to-sales [−1, 1], and SG&A-to-sales [−1, 3] in Panel B. The sample contains all mergers with available efficiency projections in Panel A and all completed mergers in Panel B. The main dependent variable is the value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I for the definitions of the rest of the variables. The inverse Mills ratio is computed from estimates of regression in Column 1 of Table II. The regressions are estimated with year-fixed effects. Standard errors are reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. Panel A. Efficiency gain projections and announcement returns Dependent variable Combined CAR [−1, 1] Combined CAR [−10, 10] Combined CAR [−20, 20] Intercept 0.146** 0.234** 0.164 (2.38) (2.51) (1.38) PV of projected efficiency gains 0.217*** 0.279*** 0.344*** (3.28) (3.32) (4.99) Implied change in HHI −0.005 0.039 0.080 (−0.16) (0.85) (1.36) Product similarity 0.022 0.051 0.042 (0.56) (0.84) (0.54) Technology similarity 0.016 0.001 −0.036 (0.94) (0.05) (−1.06) Same BEA −0.003 −0.007 −0.008 (−0.46) (−0.60) (−0.54) Deregulation −0.015 −0.021 0.002 (−1.17) (−1.10) (0.08) Number of competing bidders −0.010 −0.012 −0.024 (−1.08) (−0.85) (−1.26) % cash 0.047* 0.057 0.041 (1.89) (1.50) (0.84) % equity −0.025 −0.029 −0.039 (−1.01) (−0.79) (−0.83) Log(bidder MV equity) −0.011** −0.016** −0.010 (−2.01) (−1.96) (−0.98) Log(bidder MV of equity/target MV equity) 0.005 −0.005 −0.003 (0.51) (−0.34) (−0.19) Bidder Tobin’s Q −0.001 −0.006** −0.007* (−0.76) (−2.08) (−1.78) Target Tobin’s Q −0.001 −0.003 0.000 (−0.37) (−0.84) (−0.01) Bidder leverage −0.016 −0.058* −0.056 (−0.78) (−1.81) (−1.38) Target leverage 0.008 −0.017 −0.011 (0.39) (−0.54) (−0.27) Bidder profitability −0.002 0.037 0.056* (−0.10) (1.60) (1.86) Target profitability −0.001 −0.030** −0.017 (−0.14) (−1.98) (−0.89) Bidder capital stock 0.013 0.008 0.037 (0.40) (0.16) (0.60) Target capital stock 0.019 0.068 0.002 (0.56) (1.32) (0.04) Bidder runup 0.026*** 0.017 −0.020 (2.72) (1.18) (−1.09) Target runup −0.036*** −0.033*** −0.044*** (−4.37) (−2.64) (−2.77) Bidder return volatility −0.546 −0.548 −0.005 (−1.08) (−0.72) (−0.00) Target return volatility −0.527 −0.743 −1.136 (−1.34) (−1.25) (−1.49) Bidder E-index presence 0.032* 0.021 0.021 (1.91) (0.82) (0.65) Target E-index presence −0.000 0.012 0.043 (−0.01) (0.33) (0.92) Bidder E-index×presence −0.004 −0.008 −0.002 (−0.80) (−0.98) (−0.24) Target E-index×presence −0.004 −0.010 −0.018 (−0.52) (−0.85) (−1.22) Inverse Mills ratio 0.090** 0.067* 0.015 (3.20) (1.89) (0.78) Number of observation 480 480 480 Adjusted R-squared 0.210 0.160 0.160 Panel A. Efficiency gain projections and announcement returns Dependent variable Combined CAR [−1, 1] Combined CAR [−10, 10] Combined CAR [−20, 20] Intercept 0.146** 0.234** 0.164 (2.38) (2.51) (1.38) PV of projected efficiency gains 0.217*** 0.279*** 0.344*** (3.28) (3.32) (4.99) Implied change in HHI −0.005 0.039 0.080 (−0.16) (0.85) (1.36) Product similarity 0.022 0.051 0.042 (0.56) (0.84) (0.54) Technology similarity 0.016 0.001 −0.036 (0.94) (0.05) (−1.06) Same BEA −0.003 −0.007 −0.008 (−0.46) (−0.60) (−0.54) Deregulation −0.015 −0.021 0.002 (−1.17) (−1.10) (0.08) Number of competing bidders −0.010 −0.012 −0.024 (−1.08) (−0.85) (−1.26) % cash 0.047* 0.057 0.041 (1.89) (1.50) (0.84) % equity −0.025 −0.029 −0.039 (−1.01) (−0.79) (−0.83) Log(bidder MV equity) −0.011** −0.016** −0.010 (−2.01) (−1.96) (−0.98) Log(bidder MV of equity/target MV equity) 0.005 −0.005 −0.003 (0.51) (−0.34) (−0.19) Bidder Tobin’s Q −0.001 −0.006** −0.007* (−0.76) (−2.08) (−1.78) Target Tobin’s Q −0.001 −0.003 0.000 (−0.37) (−0.84) (−0.01) Bidder leverage −0.016 −0.058* −0.056 (−0.78) (−1.81) (−1.38) Target leverage 0.008 −0.017 −0.011 (0.39) (−0.54) (−0.27) Bidder profitability −0.002 0.037 0.056* (−0.10) (1.60) (1.86) Target profitability −0.001 −0.030** −0.017 (−0.14) (−1.98) (−0.89) Bidder capital stock 0.013 0.008 0.037 (0.40) (0.16) (0.60) Target capital stock 0.019 0.068 0.002 (0.56) (1.32) (0.04) Bidder runup 0.026*** 0.017 −0.020 (2.72) (1.18) (−1.09) Target runup −0.036*** −0.033*** −0.044*** (−4.37) (−2.64) (−2.77) Bidder return volatility −0.546 −0.548 −0.005 (−1.08) (−0.72) (−0.00) Target return volatility −0.527 −0.743 −1.136 (−1.34) (−1.25) (−1.49) Bidder E-index presence 0.032* 0.021 0.021 (1.91) (0.82) (0.65) Target E-index presence −0.000 0.012 0.043 (−0.01) (0.33) (0.92) Bidder E-index×presence −0.004 −0.008 −0.002 (−0.80) (−0.98) (−0.24) Target E-index×presence −0.004 −0.010 −0.018 (−0.52) (−0.85) (−1.22) Inverse Mills ratio 0.090** 0.067* 0.015 (3.20) (1.89) (0.78) Number of observation 480 480 480 Adjusted R-squared 0.210 0.160 0.160 Panel B. Efficiency gain projections and operating performance Dependent variable ΔROS [−1, 1] ΔROS [−1, 3] ΔCOGS [−1, 1] ΔCOGS [−1, 3] ΔSG&A [−1, 1] ΔSG&A [−1, 3] Intercept 0.101 −0.082 −0.001 −0.005 −0.116** −0.012 (1.23) (−0.79) (−0.02) (−0.06) (−1.96) (−0.18) PV of projected efficiency gains 0.223** 0.130* −0.106** −0.123** −0.051 −0.099* (2.37) (1.69) (−2.13) (−2.36) (−1.25) (−1.84) Implied change in HHI 0.037 0.045 0.013 0.026 0.070** −0.008 (1.37) (0.86) (0.42) (0.61) (2.37) (−0.24) Product similarity −0.017 −0.042 −0.006 −0.023 −0.015 0.036 (−0.31) (−0.62) (−0.14) (−0.41) (−0.39) (0.89) Technology similarity 0.005 0.016 −0.028 −0.033 −0.008 −0.006 (0.23) (0.53) (−1.50) (−1.30) (−0.45) (−0.33) Same BEA 0.006 0.052*** −0.013 −0.021** 0.016** −0.006 (0.62) (3.95) (−1.59) (−1.97) (2.17) (−0.74) Deregulation 0.041 0.026 −0.045* −0.036 0.047** 0.022 (1.34) (0.68) (−1.91) (−1.14) (2.14) (0.95) Number of competing bidders −0.025 −0.018 0.034** −0.004 −0.009 0.005 (−1.22) (−0.69) (2.14) (−0.17) (−0.63) (0.33) % Cash 0.106*** 0.101** −0.053* −0.068* 0.014 0.003 (2.70) (2.02) (−1.74) (−1.65) (0.51) (0.10) % Equity 0.077** 0.124*** −0.021 −0.022 0.027 −0.041 (2.07) (2.63) (−0.73) (−0.56) (1.01) (−1.44) Log(bidder MV equity) −0.007 0.014 −0.001 0.002 0.011** −0.003 (−1.04) (1.51) (−0.16) (0.25) (2.14) (−0.57) Log(bidder MV equity/ target MV equity) −0.027** 0.003 0.010 0.014 0.014 −0.008 (−2.23) (0.21) (1.01) (1.12) (1.56) (−0.81) Bidder Tobin’s Q −0.004 −0.012*** 0.005** 0.007** −0.003 0.002 (−1.40) (−3.28) (2.23) (2.37) (−1.19) (0.77) Target Tobin’s Q 0.006** 0.005 −0.003 −0.004* 0.000 0.001 (2.51) (1.44) (−1.41) (−1.66) (−0.10) (0.63) Bidder leverage 0.008 0.035 −0.031 −0.054* 0.034* 0.035 (0.30) (0.97) (−1.38) (−1.80) (1.67) (1.58) Target leverage 0.023 0.053 −0.005 −0.003 0.051** 0.006 (0.83) (1.47) (−0.22) (−0.09) (2.52) (0.28) Bidder profitability −0.298*** −0.670*** 0.030** 0.047** 0.134*** 0.202*** (−15.47) (−27.21) (1.99) (2.30) (9.63) (13.64) Target profitability 0.008 0.047*** 0.098*** 0.098*** −0.003 −0.051*** (0.65) (3.13) (10.69) (7.84) (−0.31) (−5.59) Bidder capital stock 0.168*** 0.032 −0.114*** −0.088** −0.072** 0.009 (4.08) (0.61) (−3.56) (−2.02) (−2.43) (0.29) Target capital stock −0.142*** −0.009 0.119*** 0.069 0.054* −0.041 (−3.27) (−0.16) (3.52) (1.49) (1.72) (−1.22) Bidder stock price runup −0.022 0.009 0.018 −0.013 −0.019* 0.000 (−1.44) (0.47) (1.49) (−0.82) (−1.68) (−0.04) Target stock price runup 0.005 −0.001 −0.022** −0.014 −0.007 −0.010 (0.48) (−0.06) (−2.48) (−1.20) (−0.90) (−1.21) Bidder return volatility 0.197 −3.232*** −0.267 −0.367 −1.258** 2.170*** (0.29) (−3.68) (−0.50) (−0.51) (−2.54) (4.11) Target return volatility −1.044** −1.293** 0.748* 1.020* −0.093 −0.270 (−2.04) (−1.98) (1.88) (1.89) (−0.25) (−0.69) Bidder E-index presence 0.010 0.003 0.024 0.021 −0.010 −0.012 (0.43) (0.09) (1.37) (0.86) (−0.62) (−0.71) Target E-index presence 0.018 −0.042 −0.030 −0.035 −0.023 0.023 (0.58) (−1.04) (−1.21) (−1.06) (−1.03) (0.95) Bidder E-index×presence −0.003 0.005 −0.007 −0.003 0.009* 0.004 (−0.42) (0.59) (−1.23) (−0.43) (1.78) (0.75) Target E-index×presence −0.008 0.011 0.007 0.008 0.008 −0.007 (−0.89) (0.90) (0.91) (0.83) (1.26) (−0.94) Inverse Mills ratio 0.074** 0.055 0.027 0.036 −0.059** −0.040 (2.05) (1.20) (0.95) (0.98) (−2.27) (−1.43) Number of observation 399 376 399 376 399 376 Adjusted R-squared 0.550 0.765 0.406 0.287 0.281 0.444 Panel B. Efficiency gain projections and operating performance Dependent variable ΔROS [−1, 1] ΔROS [−1, 3] ΔCOGS [−1, 1] ΔCOGS [−1, 3] ΔSG&A [−1, 1] ΔSG&A [−1, 3] Intercept 0.101 −0.082 −0.001 −0.005 −0.116** −0.012 (1.23) (−0.79) (−0.02) (−0.06) (−1.96) (−0.18) PV of projected efficiency gains 0.223** 0.130* −0.106** −0.123** −0.051 −0.099* (2.37) (1.69) (−2.13) (−2.36) (−1.25) (−1.84) Implied change in HHI 0.037 0.045 0.013 0.026 0.070** −0.008 (1.37) (0.86) (0.42) (0.61) (2.37) (−0.24) Product similarity −0.017 −0.042 −0.006 −0.023 −0.015 0.036 (−0.31) (−0.62) (−0.14) (−0.41) (−0.39) (0.89) Technology similarity 0.005 0.016 −0.028 −0.033 −0.008 −0.006 (0.23) (0.53) (−1.50) (−1.30) (−0.45) (−0.33) Same BEA 0.006 0.052*** −0.013 −0.021** 0.016** −0.006 (0.62) (3.95) (−1.59) (−1.97) (2.17) (−0.74) Deregulation 0.041 0.026 −0.045* −0.036 0.047** 0.022 (1.34) (0.68) (−1.91) (−1.14) (2.14) (0.95) Number of competing bidders −0.025 −0.018 0.034** −0.004 −0.009 0.005 (−1.22) (−0.69) (2.14) (−0.17) (−0.63) (0.33) % Cash 0.106*** 0.101** −0.053* −0.068* 0.014 0.003 (2.70) (2.02) (−1.74) (−1.65) (0.51) (0.10) % Equity 0.077** 0.124*** −0.021 −0.022 0.027 −0.041 (2.07) (2.63) (−0.73) (−0.56) (1.01) (−1.44) Log(bidder MV equity) −0.007 0.014 −0.001 0.002 0.011** −0.003 (−1.04) (1.51) (−0.16) (0.25) (2.14) (−0.57) Log(bidder MV equity/ target MV equity) −0.027** 0.003 0.010 0.014 0.014 −0.008 (−2.23) (0.21) (1.01) (1.12) (1.56) (−0.81) Bidder Tobin’s Q −0.004 −0.012*** 0.005** 0.007** −0.003 0.002 (−1.40) (−3.28) (2.23) (2.37) (−1.19) (0.77) Target Tobin’s Q 0.006** 0.005 −0.003 −0.004* 0.000 0.001 (2.51) (1.44) (−1.41) (−1.66) (−0.10) (0.63) Bidder leverage 0.008 0.035 −0.031 −0.054* 0.034* 0.035 (0.30) (0.97) (−1.38) (−1.80) (1.67) (1.58) Target leverage 0.023 0.053 −0.005 −0.003 0.051** 0.006 (0.83) (1.47) (−0.22) (−0.09) (2.52) (0.28) Bidder profitability −0.298*** −0.670*** 0.030** 0.047** 0.134*** 0.202*** (−15.47) (−27.21) (1.99) (2.30) (9.63) (13.64) Target profitability 0.008 0.047*** 0.098*** 0.098*** −0.003 −0.051*** (0.65) (3.13) (10.69) (7.84) (−0.31) (−5.59) Bidder capital stock 0.168*** 0.032 −0.114*** −0.088** −0.072** 0.009 (4.08) (0.61) (−3.56) (−2.02) (−2.43) (0.29) Target capital stock −0.142*** −0.009 0.119*** 0.069 0.054* −0.041 (−3.27) (−0.16) (3.52) (1.49) (1.72) (−1.22) Bidder stock price runup −0.022 0.009 0.018 −0.013 −0.019* 0.000 (−1.44) (0.47) (1.49) (−0.82) (−1.68) (−0.04) Target stock price runup 0.005 −0.001 −0.022** −0.014 −0.007 −0.010 (0.48) (−0.06) (−2.48) (−1.20) (−0.90) (−1.21) Bidder return volatility 0.197 −3.232*** −0.267 −0.367 −1.258** 2.170*** (0.29) (−3.68) (−0.50) (−0.51) (−2.54) (4.11) Target return volatility −1.044** −1.293** 0.748* 1.020* −0.093 −0.270 (−2.04) (−1.98) (1.88) (1.89) (−0.25) (−0.69) Bidder E-index presence 0.010 0.003 0.024 0.021 −0.010 −0.012 (0.43) (0.09) (1.37) (0.86) (−0.62) (−0.71) Target E-index presence 0.018 −0.042 −0.030 −0.035 −0.023 0.023 (0.58) (−1.04) (−1.21) (−1.06) (−1.03) (0.95) Bidder E-index×presence −0.003 0.005 −0.007 −0.003 0.009* 0.004 (−0.42) (0.59) (−1.23) (−0.43) (1.78) (0.75) Target E-index×presence −0.008 0.011 0.007 0.008 0.008 −0.007 (−0.89) (0.90) (0.91) (0.83) (1.26) (−0.94) Inverse Mills ratio 0.074** 0.055 0.027 0.036 −0.059** −0.040 (2.05) (1.20) (0.95) (0.98) (−2.27) (−1.43) Number of observation 399 376 399 376 399 376 Adjusted R-squared 0.550 0.765 0.406 0.287 0.281 0.444 Table III. Projected operating efficiency gains and merging firms’ valuation and performance This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and target’s combined CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20] in Panel A and bidder’s change in ROS [−1, 1], ROS [−1, 3], COGS-to-sales [−1, 1], COGS-to-sales [−1, 3], SG&A-to-sales [−1, 1], and SG&A-to-sales [−1, 3] in Panel B. The sample contains all mergers with available efficiency projections in Panel A and all completed mergers in Panel B. The main dependent variable is the value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I for the definitions of the rest of the variables. The inverse Mills ratio is computed from estimates of regression in Column 1 of Table II. The regressions are estimated with year-fixed effects. Standard errors are reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. Panel A. Efficiency gain projections and announcement returns Dependent variable Combined CAR [−1, 1] Combined CAR [−10, 10] Combined CAR [−20, 20] Intercept 0.146** 0.234** 0.164 (2.38) (2.51) (1.38) PV of projected efficiency gains 0.217*** 0.279*** 0.344*** (3.28) (3.32) (4.99) Implied change in HHI −0.005 0.039 0.080 (−0.16) (0.85) (1.36) Product similarity 0.022 0.051 0.042 (0.56) (0.84) (0.54) Technology similarity 0.016 0.001 −0.036 (0.94) (0.05) (−1.06) Same BEA −0.003 −0.007 −0.008 (−0.46) (−0.60) (−0.54) Deregulation −0.015 −0.021 0.002 (−1.17) (−1.10) (0.08) Number of competing bidders −0.010 −0.012 −0.024 (−1.08) (−0.85) (−1.26) % cash 0.047* 0.057 0.041 (1.89) (1.50) (0.84) % equity −0.025 −0.029 −0.039 (−1.01) (−0.79) (−0.83) Log(bidder MV equity) −0.011** −0.016** −0.010 (−2.01) (−1.96) (−0.98) Log(bidder MV of equity/target MV equity) 0.005 −0.005 −0.003 (0.51) (−0.34) (−0.19) Bidder Tobin’s Q −0.001 −0.006** −0.007* (−0.76) (−2.08) (−1.78) Target Tobin’s Q −0.001 −0.003 0.000 (−0.37) (−0.84) (−0.01) Bidder leverage −0.016 −0.058* −0.056 (−0.78) (−1.81) (−1.38) Target leverage 0.008 −0.017 −0.011 (0.39) (−0.54) (−0.27) Bidder profitability −0.002 0.037 0.056* (−0.10) (1.60) (1.86) Target profitability −0.001 −0.030** −0.017 (−0.14) (−1.98) (−0.89) Bidder capital stock 0.013 0.008 0.037 (0.40) (0.16) (0.60) Target capital stock 0.019 0.068 0.002 (0.56) (1.32) (0.04) Bidder runup 0.026*** 0.017 −0.020 (2.72) (1.18) (−1.09) Target runup −0.036*** −0.033*** −0.044*** (−4.37) (−2.64) (−2.77) Bidder return volatility −0.546 −0.548 −0.005 (−1.08) (−0.72) (−0.00) Target return volatility −0.527 −0.743 −1.136 (−1.34) (−1.25) (−1.49) Bidder E-index presence 0.032* 0.021 0.021 (1.91) (0.82) (0.65) Target E-index presence −0.000 0.012 0.043 (−0.01) (0.33) (0.92) Bidder E-index×presence −0.004 −0.008 −0.002 (−0.80) (−0.98) (−0.24) Target E-index×presence −0.004 −0.010 −0.018 (−0.52) (−0.85) (−1.22) Inverse Mills ratio 0.090** 0.067* 0.015 (3.20) (1.89) (0.78) Number of observation 480 480 480 Adjusted R-squared 0.210 0.160 0.160 Panel A. Efficiency gain projections and announcement returns Dependent variable Combined CAR [−1, 1] Combined CAR [−10, 10] Combined CAR [−20, 20] Intercept 0.146** 0.234** 0.164 (2.38) (2.51) (1.38) PV of projected efficiency gains 0.217*** 0.279*** 0.344*** (3.28) (3.32) (4.99) Implied change in HHI −0.005 0.039 0.080 (−0.16) (0.85) (1.36) Product similarity 0.022 0.051 0.042 (0.56) (0.84) (0.54) Technology similarity 0.016 0.001 −0.036 (0.94) (0.05) (−1.06) Same BEA −0.003 −0.007 −0.008 (−0.46) (−0.60) (−0.54) Deregulation −0.015 −0.021 0.002 (−1.17) (−1.10) (0.08) Number of competing bidders −0.010 −0.012 −0.024 (−1.08) (−0.85) (−1.26) % cash 0.047* 0.057 0.041 (1.89) (1.50) (0.84) % equity −0.025 −0.029 −0.039 (−1.01) (−0.79) (−0.83) Log(bidder MV equity) −0.011** −0.016** −0.010 (−2.01) (−1.96) (−0.98) Log(bidder MV of equity/target MV equity) 0.005 −0.005 −0.003 (0.51) (−0.34) (−0.19) Bidder Tobin’s Q −0.001 −0.006** −0.007* (−0.76) (−2.08) (−1.78) Target Tobin’s Q −0.001 −0.003 0.000 (−0.37) (−0.84) (−0.01) Bidder leverage −0.016 −0.058* −0.056 (−0.78) (−1.81) (−1.38) Target leverage 0.008 −0.017 −0.011 (0.39) (−0.54) (−0.27) Bidder profitability −0.002 0.037 0.056* (−0.10) (1.60) (1.86) Target profitability −0.001 −0.030** −0.017 (−0.14) (−1.98) (−0.89) Bidder capital stock 0.013 0.008 0.037 (0.40) (0.16) (0.60) Target capital stock 0.019 0.068 0.002 (0.56) (1.32) (0.04) Bidder runup 0.026*** 0.017 −0.020 (2.72) (1.18) (−1.09) Target runup −0.036*** −0.033*** −0.044*** (−4.37) (−2.64) (−2.77) Bidder return volatility −0.546 −0.548 −0.005 (−1.08) (−0.72) (−0.00) Target return volatility −0.527 −0.743 −1.136 (−1.34) (−1.25) (−1.49) Bidder E-index presence 0.032* 0.021 0.021 (1.91) (0.82) (0.65) Target E-index presence −0.000 0.012 0.043 (−0.01) (0.33) (0.92) Bidder E-index×presence −0.004 −0.008 −0.002 (−0.80) (−0.98) (−0.24) Target E-index×presence −0.004 −0.010 −0.018 (−0.52) (−0.85) (−1.22) Inverse Mills ratio 0.090** 0.067* 0.015 (3.20) (1.89) (0.78) Number of observation 480 480 480 Adjusted R-squared 0.210 0.160 0.160 Panel B. Efficiency gain projections and operating performance Dependent variable ΔROS [−1, 1] ΔROS [−1, 3] ΔCOGS [−1, 1] ΔCOGS [−1, 3] ΔSG&A [−1, 1] ΔSG&A [−1, 3] Intercept 0.101 −0.082 −0.001 −0.005 −0.116** −0.012 (1.23) (−0.79) (−0.02) (−0.06) (−1.96) (−0.18) PV of projected efficiency gains 0.223** 0.130* −0.106** −0.123** −0.051 −0.099* (2.37) (1.69) (−2.13) (−2.36) (−1.25) (−1.84) Implied change in HHI 0.037 0.045 0.013 0.026 0.070** −0.008 (1.37) (0.86) (0.42) (0.61) (2.37) (−0.24) Product similarity −0.017 −0.042 −0.006 −0.023 −0.015 0.036 (−0.31) (−0.62) (−0.14) (−0.41) (−0.39) (0.89) Technology similarity 0.005 0.016 −0.028 −0.033 −0.008 −0.006 (0.23) (0.53) (−1.50) (−1.30) (−0.45) (−0.33) Same BEA 0.006 0.052*** −0.013 −0.021** 0.016** −0.006 (0.62) (3.95) (−1.59) (−1.97) (2.17) (−0.74) Deregulation 0.041 0.026 −0.045* −0.036 0.047** 0.022 (1.34) (0.68) (−1.91) (−1.14) (2.14) (0.95) Number of competing bidders −0.025 −0.018 0.034** −0.004 −0.009 0.005 (−1.22) (−0.69) (2.14) (−0.17) (−0.63) (0.33) % Cash 0.106*** 0.101** −0.053* −0.068* 0.014 0.003 (2.70) (2.02) (−1.74) (−1.65) (0.51) (0.10) % Equity 0.077** 0.124*** −0.021 −0.022 0.027 −0.041 (2.07) (2.63) (−0.73) (−0.56) (1.01) (−1.44) Log(bidder MV equity) −0.007 0.014 −0.001 0.002 0.011** −0.003 (−1.04) (1.51) (−0.16) (0.25) (2.14) (−0.57) Log(bidder MV equity/ target MV equity) −0.027** 0.003 0.010 0.014 0.014 −0.008 (−2.23) (0.21) (1.01) (1.12) (1.56) (−0.81) Bidder Tobin’s Q −0.004 −0.012*** 0.005** 0.007** −0.003 0.002 (−1.40) (−3.28) (2.23) (2.37) (−1.19) (0.77) Target Tobin’s Q 0.006** 0.005 −0.003 −0.004* 0.000 0.001 (2.51) (1.44) (−1.41) (−1.66) (−0.10) (0.63) Bidder leverage 0.008 0.035 −0.031 −0.054* 0.034* 0.035 (0.30) (0.97) (−1.38) (−1.80) (1.67) (1.58) Target leverage 0.023 0.053 −0.005 −0.003 0.051** 0.006 (0.83) (1.47) (−0.22) (−0.09) (2.52) (0.28) Bidder profitability −0.298*** −0.670*** 0.030** 0.047** 0.134*** 0.202*** (−15.47) (−27.21) (1.99) (2.30) (9.63) (13.64) Target profitability 0.008 0.047*** 0.098*** 0.098*** −0.003 −0.051*** (0.65) (3.13) (10.69) (7.84) (−0.31) (−5.59) Bidder capital stock 0.168*** 0.032 −0.114*** −0.088** −0.072** 0.009 (4.08) (0.61) (−3.56) (−2.02) (−2.43) (0.29) Target capital stock −0.142*** −0.009 0.119*** 0.069 0.054* −0.041 (−3.27) (−0.16) (3.52) (1.49) (1.72) (−1.22) Bidder stock price runup −0.022 0.009 0.018 −0.013 −0.019* 0.000 (−1.44) (0.47) (1.49) (−0.82) (−1.68) (−0.04) Target stock price runup 0.005 −0.001 −0.022** −0.014 −0.007 −0.010 (0.48) (−0.06) (−2.48) (−1.20) (−0.90) (−1.21) Bidder return volatility 0.197 −3.232*** −0.267 −0.367 −1.258** 2.170*** (0.29) (−3.68) (−0.50) (−0.51) (−2.54) (4.11) Target return volatility −1.044** −1.293** 0.748* 1.020* −0.093 −0.270 (−2.04) (−1.98) (1.88) (1.89) (−0.25) (−0.69) Bidder E-index presence 0.010 0.003 0.024 0.021 −0.010 −0.012 (0.43) (0.09) (1.37) (0.86) (−0.62) (−0.71) Target E-index presence 0.018 −0.042 −0.030 −0.035 −0.023 0.023 (0.58) (−1.04) (−1.21) (−1.06) (−1.03) (0.95) Bidder E-index×presence −0.003 0.005 −0.007 −0.003 0.009* 0.004 (−0.42) (0.59) (−1.23) (−0.43) (1.78) (0.75) Target E-index×presence −0.008 0.011 0.007 0.008 0.008 −0.007 (−0.89) (0.90) (0.91) (0.83) (1.26) (−0.94) Inverse Mills ratio 0.074** 0.055 0.027 0.036 −0.059** −0.040 (2.05) (1.20) (0.95) (0.98) (−2.27) (−1.43) Number of observation 399 376 399 376 399 376 Adjusted R-squared 0.550 0.765 0.406 0.287 0.281 0.444 Panel B. Efficiency gain projections and operating performance Dependent variable ΔROS [−1, 1] ΔROS [−1, 3] ΔCOGS [−1, 1] ΔCOGS [−1, 3] ΔSG&A [−1, 1] ΔSG&A [−1, 3] Intercept 0.101 −0.082 −0.001 −0.005 −0.116** −0.012 (1.23) (−0.79) (−0.02) (−0.06) (−1.96) (−0.18) PV of projected efficiency gains 0.223** 0.130* −0.106** −0.123** −0.051 −0.099* (2.37) (1.69) (−2.13) (−2.36) (−1.25) (−1.84) Implied change in HHI 0.037 0.045 0.013 0.026 0.070** −0.008 (1.37) (0.86) (0.42) (0.61) (2.37) (−0.24) Product similarity −0.017 −0.042 −0.006 −0.023 −0.015 0.036 (−0.31) (−0.62) (−0.14) (−0.41) (−0.39) (0.89) Technology similarity 0.005 0.016 −0.028 −0.033 −0.008 −0.006 (0.23) (0.53) (−1.50) (−1.30) (−0.45) (−0.33) Same BEA 0.006 0.052*** −0.013 −0.021** 0.016** −0.006 (0.62) (3.95) (−1.59) (−1.97) (2.17) (−0.74) Deregulation 0.041 0.026 −0.045* −0.036 0.047** 0.022 (1.34) (0.68) (−1.91) (−1.14) (2.14) (0.95) Number of competing bidders −0.025 −0.018 0.034** −0.004 −0.009 0.005 (−1.22) (−0.69) (2.14) (−0.17) (−0.63) (0.33) % Cash 0.106*** 0.101** −0.053* −0.068* 0.014 0.003 (2.70) (2.02) (−1.74) (−1.65) (0.51) (0.10) % Equity 0.077** 0.124*** −0.021 −0.022 0.027 −0.041 (2.07) (2.63) (−0.73) (−0.56) (1.01) (−1.44) Log(bidder MV equity) −0.007 0.014 −0.001 0.002 0.011** −0.003 (−1.04) (1.51) (−0.16) (0.25) (2.14) (−0.57) Log(bidder MV equity/ target MV equity) −0.027** 0.003 0.010 0.014 0.014 −0.008 (−2.23) (0.21) (1.01) (1.12) (1.56) (−0.81) Bidder Tobin’s Q −0.004 −0.012*** 0.005** 0.007** −0.003 0.002 (−1.40) (−3.28) (2.23) (2.37) (−1.19) (0.77) Target Tobin’s Q 0.006** 0.005 −0.003 −0.004* 0.000 0.001 (2.51) (1.44) (−1.41) (−1.66) (−0.10) (0.63) Bidder leverage 0.008 0.035 −0.031 −0.054* 0.034* 0.035 (0.30) (0.97) (−1.38) (−1.80) (1.67) (1.58) Target leverage 0.023 0.053 −0.005 −0.003 0.051** 0.006 (0.83) (1.47) (−0.22) (−0.09) (2.52) (0.28) Bidder profitability −0.298*** −0.670*** 0.030** 0.047** 0.134*** 0.202*** (−15.47) (−27.21) (1.99) (2.30) (9.63) (13.64) Target profitability 0.008 0.047*** 0.098*** 0.098*** −0.003 −0.051*** (0.65) (3.13) (10.69) (7.84) (−0.31) (−5.59) Bidder capital stock 0.168*** 0.032 −0.114*** −0.088** −0.072** 0.009 (4.08) (0.61) (−3.56) (−2.02) (−2.43) (0.29) Target capital stock −0.142*** −0.009 0.119*** 0.069 0.054* −0.041 (−3.27) (−0.16) (3.52) (1.49) (1.72) (−1.22) Bidder stock price runup −0.022 0.009 0.018 −0.013 −0.019* 0.000 (−1.44) (0.47) (1.49) (−0.82) (−1.68) (−0.04) Target stock price runup 0.005 −0.001 −0.022** −0.014 −0.007 −0.010 (0.48) (−0.06) (−2.48) (−1.20) (−0.90) (−1.21) Bidder return volatility 0.197 −3.232*** −0.267 −0.367 −1.258** 2.170*** (0.29) (−3.68) (−0.50) (−0.51) (−2.54) (4.11) Target return volatility −1.044** −1.293** 0.748* 1.020* −0.093 −0.270 (−2.04) (−1.98) (1.88) (1.89) (−0.25) (−0.69) Bidder E-index presence 0.010 0.003 0.024 0.021 −0.010 −0.012 (0.43) (0.09) (1.37) (0.86) (−0.62) (−0.71) Target E-index presence 0.018 −0.042 −0.030 −0.035 −0.023 0.023 (0.58) (−1.04) (−1.21) (−1.06) (−1.03) (0.95) Bidder E-index×presence −0.003 0.005 −0.007 −0.003 0.009* 0.004 (−0.42) (0.59) (−1.23) (−0.43) (1.78) (0.75) Target E-index×presence −0.008 0.011 0.007 0.008 0.008 −0.007 (−0.89) (0.90) (0.91) (0.83) (1.26) (−0.94) Inverse Mills ratio 0.074** 0.055 0.027 0.036 −0.059** −0.040 (2.05) (1.20) (0.95) (0.98) (−2.27) (−1.43) Number of observation 399 376 399 376 399 376 Adjusted R-squared 0.550 0.765 0.406 0.287 0.281 0.444 The most important finding in Panel A is that efficiency gain projections are strongly and highly statistically significantly (at 1% level) related to merging firms’ combined announcement returns for all three announcement windows. This relation is economically important: a one-standard-deviation increase in the scaled value of operating efficiency projections (0.114) is associated with an average increase of 2.5% in combined cumulative abnormal return over the [−1, 1] announcement window, ceteris paribus, an average increase of 3.2% in the announcement return over the [−3, 3] window, and an average increase of 3.9% in the return over the [−20, 20] window. The results in our sample of horizontal mergers are consistent with Bernile and Bauguess’ (2014) results in a sample that includes also vertical and diversifying mergers. Other variables that are significantly related to merging firms’ announcement returns include bidder’s market capitalization (a negative relation, significant at 5% level in two specifications out of three), bidder’s Tobin’s Q (a negative relation, significant at 10% level in two specifications), and target’s stock price runup (a negative relation, highly signifiant in all three specifications). The coefficients on the inverse Mills ratio are statistically significant at 5% level in two specifications, suggesting that self-selection in managers’ decisions to disclose efficiency gain forecasts impacts the relation between these forecasts and merging firms’ returns around deal announcements. Having established that operating efficiency gain projections are positively associated with merging firms’ announcement returns, we proceed to examine whether these projections materialize in the form of improved operating performance, i.e., improved post-merger profitability and lower post-merger cost ratios. In Panel B of Table III, we estimate, within a subsample of 399 completed horizontal mergers with efficiency gain projections, regressions similar to those in Panel A, where the dependent variables are post-merger changes in profitability and in cost ratios. In particular, in the first column of Panel B, the dependent variable is the difference between the merged firm’s ROS in the first year following merger completion and bidder’s ROS in the year prior to merger announcement. In the second column, it is the difference between the merged firm’s average ROS in three post-merger-completion years and bidder’s ROS in the last pre-merger year. The relation between operating efficiency gain projections and post-merger changes in profitability is statistically significant in both specifications (at 5% level for 1-year change in ROS and at 10% level for 3-year change in ROS). It is also economically sizable: a one-standard-deviation increase in projected operating efficiency gains is associated with a 2.5 (1.5) percentage point increase in ROS over the 1-year (3-year) horizon. These figures are non-negligible compared with the standard deviation of post-merger changes in ROS (11% and 18%, respectively, for the two horizons). Additional variables that are significantly associated with post-merger changes in profitability at both horizons include: percentage of cash in merger offers (positively); bidder’s pre-merger profitability (negatively); and target’s stock return volatility (negatively). To dig deeper into which components of post-merger changes in profitability are associated with operating efficiency gain projections, we replace the dependent variable by post-merger changes in bidder’s cost ratios. In Columns 3 and 4, we examine changes in COGS-to-sales ratio, which is inversely related to gross profitability, over 1-year and 3-year horizons. In Columns 5 and 6, we examine changes in SG&A-to-revenues ratio, which is related to various overhead expenses, over 1-year and 3-year horizons. The relation between operating efficiency gain projections and post-merger change in COGS-to-revenues ratio is significantly negative and economically sizable: a one-standard-deviation increase in efficiency gain forecast is associated with 1.2 (1.4) percentage point reduction in COGS-to-revenues ratio over 1-year (3-year) horizon, which corresponds to roughly 15% of its standard deviation. The relation between efficiency gain projections and the change in SG&A-to-revenues ratio is statistically significant at 10% level only at the 3-year horizon and is somewhat less economically significant: it ranges between 0.6 percentage points at the 1-year horizon and 1.1 percentage points at the 3-year horizon, figures that correspond to 10–13% of its standard deviation. Overall, the results in Panel B show that the positive association between forecasted efficiency gains and merging firms’ announcement returns is at least partially due to increased post-merger profitability, which, in turn is mostly driven by increased gross profitability and, to a lesser extent, by reduced overhead expenses. To summarize, the evidence in both panels of Table III suggests that insiders’ operating efficiency gain forecasts are a strong determinant of stock and operating performance following horizontal mergers. This evidence is consistent with the first premise of our analysis: insiders’ projections contain relevant information about valuation and performance effects of mergers. This information, however, need not necessarily be about operating efficiencies stemming from horizontal mergers, which is the second premise of our empirical strategy. It is possible that managers provide efficiency gain projections in order to disguise anticompetitive effects of mergers and reduce the likelihood of antitrust authorities blocking the deal. As we argue below, insiders’ incentives to disguise market power gains as operating efficiencies work against us finding relations between operating efficiency gain forecasts and announcement returns of related firms, posited in Predictions 1–3 in Section 3. 5. Effects of Operating Efficiency Gains on Related Firms 5.1 Identifying Supply Chain Linkages We use data on customer–supplier links provided by Cohen and Frazzini (2008). These data employ the disclosure of significant customers available in Compustat Industry Segment Files to identify firm-level customer–supplier relations. SFAS 131 requires firms to report the identity of their principal customers, i.e., customers that account for at least 10% of their sales. The set of firm i’s customers is obtained by matching its principal customers on Compustat Industry Segment Files with CRSP/Compustat.9 In addition, if firm i is a principal customer of firm j, then firm j is classified as a supplier of firm i. Similar data are used in Fee and Thomas (2004) and Hertzel, Officer, and Rodgers (2008). A limitation of this method of identifying firms along merging firms’ supply chains is that it potentially biases our samples of customers and suppliers. In particular, larger firms are more likely to be principal customers of other firms, which biases the sample of customers toward relatively large ones. Smaller firms, on the other hand, are more likely to count on principal customers for large proportions of their sales, biasing the subsample of suppliers toward relatively small ones. 5.2 Summary Statistics of Rivals, Customers, and Suppliers Table IV presents summary statistics of merging firms’ rivals (Panel A), customers (Panel B), and suppliers (Panel C). Table IV. Summary statistics for rivals, customers, and suppliers This table presents summary statistics of characteristics of product market rivals, customers, and suppliers of merging firms in Panels A, B, and C, respectively. Rivals are all non-merging firms in bidder’s and/or targets’ TNIC industry. The identification of customers and suppliers is based on Compustat Industry Segment database and is described in Section 5.1. Rivals’, customers’, and suppliers’ CARs and accounting variables are defined analogously to bidders’ variables, as discussed in Panel D of Table I. In Panels B and C, bidder’s (target’s) customer [supplier] is an indicator variable equaling one if a firm is bidder’s (target’s) customer [supplier]. All accounting variables are from the fiscal year preceding the year of the M&A announcement. Panel A. Summary statistics for rivals Mean Standard deviation Q1 Median Q3 Number of observation Rival CAR [−1, 1] −0.36% 4.45% −2.08% 0.00% 1.81% 5,269 Rival CAR [−10, 10] −1.27% 28.33% −17.87% −0.97% 15.11% 5,269 Rival CAR [−20, 20] 1.31% 56.49% −32.86% −1.40% 32.57% 5,269 Rival–bidder product similarity 9.34% 8.43% 2.70% 7.06% 12.97% 5,269 Rival–target product similarity 10.25% 9.51% 3.14% 7.42% 14.20% 5,269 Rival assets 7,269 34,051 198 709 2,873 5,269 Rival assets to bidder assets 4.19 41.90 0.05 0.25 1.07 5,269 Rival assets to target assets 9.14 70.28 0.17 0.65 2.56 5,269 Rivals sales 2,295 8,347 45 199 1,292 5,269 Rival sales to bidder sales 4.18 49.96 0.05 0.24 1.05 5,269 Rival sales to target sales 9.90 101.17 0.15 0.61 2.37 5,269 Rival Tobin’s Q 1.38 2.61 0.20 0.71 1.52 5,269 Rival leverage 0.32 0.25 0.09 0.30 0.51 5,269 Rival profitability 0.05 1.58 0.05 0.17 0.28 5,269 Panel B. Summary statistics for customers Customer CAR [−1, 1] 0.02% 4.15% −1.94% 0.00% 2.27% 560 Customer CAR [−10, 10] 0.34% 28.33% −15.23% −0.74% 18.00% 560 Customer CAR [−20, 20] 1.09% 56.83% −28.53% −1.23% 36.84% 560 Customer assets 62,367 108,119 5,078 21,561 67,628 560 Customer assets to bidder assets 12.95 32.51 0.75 3.03 8.35 560 Customer assets to target assets 17.95 45.92 1.34 5.88 17.14 560 Customer sales 39,530 50,901 5,054 19,525 51,028 560 Customer sales to bidder sales 11.22 31.00 0.90 5.01 7.64 560 Customer sales to target sales 16.88 37.06 1.42 5.85 15.33 560 Customer Tobin’s Q 1.81 2.16 0.72 1.26 2.14 560 Customer leverage 0.28 0.22 0.10 0.21 0.42 560 Customer profitability 0.11 1.12 0.05 0.09 0.16 560 Bidder customer 0.57 0.50 0.00 1.00 1.00 560 Target customer 0.51 0.50 0.00 1.00 1.00 560 Panel A. Summary statistics for rivals Mean Standard deviation Q1 Median Q3 Number of observation Rival CAR [−1, 1] −0.36% 4.45% −2.08% 0.00% 1.81% 5,269 Rival CAR [−10, 10] −1.27% 28.33% −17.87% −0.97% 15.11% 5,269 Rival CAR [−20, 20] 1.31% 56.49% −32.86% −1.40% 32.57% 5,269 Rival–bidder product similarity 9.34% 8.43% 2.70% 7.06% 12.97% 5,269 Rival–target product similarity 10.25% 9.51% 3.14% 7.42% 14.20% 5,269 Rival assets 7,269 34,051 198 709 2,873 5,269 Rival assets to bidder assets 4.19 41.90 0.05 0.25 1.07 5,269 Rival assets to target assets 9.14 70.28 0.17 0.65 2.56 5,269 Rivals sales 2,295 8,347 45 199 1,292 5,269 Rival sales to bidder sales 4.18 49.96 0.05 0.24 1.05 5,269 Rival sales to target sales 9.90 101.17 0.15 0.61 2.37 5,269 Rival Tobin’s Q 1.38 2.61 0.20 0.71 1.52 5,269 Rival leverage 0.32 0.25 0.09 0.30 0.51 5,269 Rival profitability 0.05 1.58 0.05 0.17 0.28 5,269 Panel B. Summary statistics for customers Customer CAR [−1, 1] 0.02% 4.15% −1.94% 0.00% 2.27% 560 Customer CAR [−10, 10] 0.34% 28.33% −15.23% −0.74% 18.00% 560 Customer CAR [−20, 20] 1.09% 56.83% −28.53% −1.23% 36.84% 560 Customer assets 62,367 108,119 5,078 21,561 67,628 560 Customer assets to bidder assets 12.95 32.51 0.75 3.03 8.35 560 Customer assets to target assets 17.95 45.92 1.34 5.88 17.14 560 Customer sales 39,530 50,901 5,054 19,525 51,028 560 Customer sales to bidder sales 11.22 31.00 0.90 5.01 7.64 560 Customer sales to target sales 16.88 37.06 1.42 5.85 15.33 560 Customer Tobin’s Q 1.81 2.16 0.72 1.26 2.14 560 Customer leverage 0.28 0.22 0.10 0.21 0.42 560 Customer profitability 0.11 1.12 0.05 0.09 0.16 560 Bidder customer 0.57 0.50 0.00 1.00 1.00 560 Target customer 0.51 0.50 0.00 1.00 1.00 560 Panel C. Summary statistics for suppliers Mean Standard deviation Q1 Median Q3 Number of observation Supplier CAR (−1, 1) −1.57% 5.98% −4.12% −0.74% 1.54% 1,523 Supplier CAR (−10, 10) −3.09% 33.46% −28.68% −3.62% 13.18% 1,523 Supplier CAR (−20, 20) −4.26% 63.03% −49.79% −4.19% 27.23% 1,523 Supplier assets 1,828 10,093 43 145 767 1,523 Supplier assets to bidder assets 0.38 3.31 0.01 0.02 0.07 1,523 Supplier assets to target assets 0.83 6.30 0.01 0.03 0.16 1,523 Supplier sales 1,176 4,162 32 118 514 1,523 Supplier sales to bidder sales 0.32 5.40 0.01 0.02 0.05 1,523 Supplier sales to target sales 0.94 10.59 0.01 0.03 0.12 1,523 Supplier Tobin’s Q 2.71 3.80 0.84 1.44 3.01 1,523 Supplier leverage 0.19 0.22 0.01 0.12 0.32 1,523 Supplier profitability 0.03 1.58 −0.02 0.07 0.16 1,523 Bidder supplier 0.70 0.46 0.00 1.00 1.00 1,523 Target supplier 0.38 0.49 0.00 0.00 1.00 1,523 Panel C. Summary statistics for suppliers Mean Standard deviation Q1 Median Q3 Number of observation Supplier CAR (−1, 1) −1.57% 5.98% −4.12% −0.74% 1.54% 1,523 Supplier CAR (−10, 10) −3.09% 33.46% −28.68% −3.62% 13.18% 1,523 Supplier CAR (−20, 20) −4.26% 63.03% −49.79% −4.19% 27.23% 1,523 Supplier assets 1,828 10,093 43 145 767 1,523 Supplier assets to bidder assets 0.38 3.31 0.01 0.02 0.07 1,523 Supplier assets to target assets 0.83 6.30 0.01 0.03 0.16 1,523 Supplier sales 1,176 4,162 32 118 514 1,523 Supplier sales to bidder sales 0.32 5.40 0.01 0.02 0.05 1,523 Supplier sales to target sales 0.94 10.59 0.01 0.03 0.12 1,523 Supplier Tobin’s Q 2.71 3.80 0.84 1.44 3.01 1,523 Supplier leverage 0.19 0.22 0.01 0.12 0.32 1,523 Supplier profitability 0.03 1.58 −0.02 0.07 0.16 1,523 Bidder supplier 0.70 0.46 0.00 1.00 1.00 1,523 Target supplier 0.38 0.49 0.00 0.00 1.00 1,523 Table IV. Summary statistics for rivals, customers, and suppliers This table presents summary statistics of characteristics of product market rivals, customers, and suppliers of merging firms in Panels A, B, and C, respectively. Rivals are all non-merging firms in bidder’s and/or targets’ TNIC industry. The identification of customers and suppliers is based on Compustat Industry Segment database and is described in Section 5.1. Rivals’, customers’, and suppliers’ CARs and accounting variables are defined analogously to bidders’ variables, as discussed in Panel D of Table I. In Panels B and C, bidder’s (target’s) customer [supplier] is an indicator variable equaling one if a firm is bidder’s (target’s) customer [supplier]. All accounting variables are from the fiscal year preceding the year of the M&A announcement. Panel A. Summary statistics for rivals Mean Standard deviation Q1 Median Q3 Number of observation Rival CAR [−1, 1] −0.36% 4.45% −2.08% 0.00% 1.81% 5,269 Rival CAR [−10, 10] −1.27% 28.33% −17.87% −0.97% 15.11% 5,269 Rival CAR [−20, 20] 1.31% 56.49% −32.86% −1.40% 32.57% 5,269 Rival–bidder product similarity 9.34% 8.43% 2.70% 7.06% 12.97% 5,269 Rival–target product similarity 10.25% 9.51% 3.14% 7.42% 14.20% 5,269 Rival assets 7,269 34,051 198 709 2,873 5,269 Rival assets to bidder assets 4.19 41.90 0.05 0.25 1.07 5,269 Rival assets to target assets 9.14 70.28 0.17 0.65 2.56 5,269 Rivals sales 2,295 8,347 45 199 1,292 5,269 Rival sales to bidder sales 4.18 49.96 0.05 0.24 1.05 5,269 Rival sales to target sales 9.90 101.17 0.15 0.61 2.37 5,269 Rival Tobin’s Q 1.38 2.61 0.20 0.71 1.52 5,269 Rival leverage 0.32 0.25 0.09 0.30 0.51 5,269 Rival profitability 0.05 1.58 0.05 0.17 0.28 5,269 Panel B. Summary statistics for customers Customer CAR [−1, 1] 0.02% 4.15% −1.94% 0.00% 2.27% 560 Customer CAR [−10, 10] 0.34% 28.33% −15.23% −0.74% 18.00% 560 Customer CAR [−20, 20] 1.09% 56.83% −28.53% −1.23% 36.84% 560 Customer assets 62,367 108,119 5,078 21,561 67,628 560 Customer assets to bidder assets 12.95 32.51 0.75 3.03 8.35 560 Customer assets to target assets 17.95 45.92 1.34 5.88 17.14 560 Customer sales 39,530 50,901 5,054 19,525 51,028 560 Customer sales to bidder sales 11.22 31.00 0.90 5.01 7.64 560 Customer sales to target sales 16.88 37.06 1.42 5.85 15.33 560 Customer Tobin’s Q 1.81 2.16 0.72 1.26 2.14 560 Customer leverage 0.28 0.22 0.10 0.21 0.42 560 Customer profitability 0.11 1.12 0.05 0.09 0.16 560 Bidder customer 0.57 0.50 0.00 1.00 1.00 560 Target customer 0.51 0.50 0.00 1.00 1.00 560 Panel A. Summary statistics for rivals Mean Standard deviation Q1 Median Q3 Number of observation Rival CAR [−1, 1] −0.36% 4.45% −2.08% 0.00% 1.81% 5,269 Rival CAR [−10, 10] −1.27% 28.33% −17.87% −0.97% 15.11% 5,269 Rival CAR [−20, 20] 1.31% 56.49% −32.86% −1.40% 32.57% 5,269 Rival–bidder product similarity 9.34% 8.43% 2.70% 7.06% 12.97% 5,269 Rival–target product similarity 10.25% 9.51% 3.14% 7.42% 14.20% 5,269 Rival assets 7,269 34,051 198 709 2,873 5,269 Rival assets to bidder assets 4.19 41.90 0.05 0.25 1.07 5,269 Rival assets to target assets 9.14 70.28 0.17 0.65 2.56 5,269 Rivals sales 2,295 8,347 45 199 1,292 5,269 Rival sales to bidder sales 4.18 49.96 0.05 0.24 1.05 5,269 Rival sales to target sales 9.90 101.17 0.15 0.61 2.37 5,269 Rival Tobin’s Q 1.38 2.61 0.20 0.71 1.52 5,269 Rival leverage 0.32 0.25 0.09 0.30 0.51 5,269 Rival profitability 0.05 1.58 0.05 0.17 0.28 5,269 Panel B. Summary statistics for customers Customer CAR [−1, 1] 0.02% 4.15% −1.94% 0.00% 2.27% 560 Customer CAR [−10, 10] 0.34% 28.33% −15.23% −0.74% 18.00% 560 Customer CAR [−20, 20] 1.09% 56.83% −28.53% −1.23% 36.84% 560 Customer assets 62,367 108,119 5,078 21,561 67,628 560 Customer assets to bidder assets 12.95 32.51 0.75 3.03 8.35 560 Customer assets to target assets 17.95 45.92 1.34 5.88 17.14 560 Customer sales 39,530 50,901 5,054 19,525 51,028 560 Customer sales to bidder sales 11.22 31.00 0.90 5.01 7.64 560 Customer sales to target sales 16.88 37.06 1.42 5.85 15.33 560 Customer Tobin’s Q 1.81 2.16 0.72 1.26 2.14 560 Customer leverage 0.28 0.22 0.10 0.21 0.42 560 Customer profitability 0.11 1.12 0.05 0.09 0.16 560 Bidder customer 0.57 0.50 0.00 1.00 1.00 560 Target customer 0.51 0.50 0.00 1.00 1.00 560 Panel C. Summary statistics for suppliers Mean Standard deviation Q1 Median Q3 Number of observation Supplier CAR (−1, 1) −1.57% 5.98% −4.12% −0.74% 1.54% 1,523 Supplier CAR (−10, 10) −3.09% 33.46% −28.68% −3.62% 13.18% 1,523 Supplier CAR (−20, 20) −4.26% 63.03% −49.79% −4.19% 27.23% 1,523 Supplier assets 1,828 10,093 43 145 767 1,523 Supplier assets to bidder assets 0.38 3.31 0.01 0.02 0.07 1,523 Supplier assets to target assets 0.83 6.30 0.01 0.03 0.16 1,523 Supplier sales 1,176 4,162 32 118 514 1,523 Supplier sales to bidder sales 0.32 5.40 0.01 0.02 0.05 1,523 Supplier sales to target sales 0.94 10.59 0.01 0.03 0.12 1,523 Supplier Tobin’s Q 2.71 3.80 0.84 1.44 3.01 1,523 Supplier leverage 0.19 0.22 0.01 0.12 0.32 1,523 Supplier profitability 0.03 1.58 −0.02 0.07 0.16 1,523 Bidder supplier 0.70 0.46 0.00 1.00 1.00 1,523 Target supplier 0.38 0.49 0.00 0.00 1.00 1,523 Panel C. Summary statistics for suppliers Mean Standard deviation Q1 Median Q3 Number of observation Supplier CAR (−1, 1) −1.57% 5.98% −4.12% −0.74% 1.54% 1,523 Supplier CAR (−10, 10) −3.09% 33.46% −28.68% −3.62% 13.18% 1,523 Supplier CAR (−20, 20) −4.26% 63.03% −49.79% −4.19% 27.23% 1,523 Supplier assets 1,828 10,093 43 145 767 1,523 Supplier assets to bidder assets 0.38 3.31 0.01 0.02 0.07 1,523 Supplier assets to target assets 0.83 6.30 0.01 0.03 0.16 1,523 Supplier sales 1,176 4,162 32 118 514 1,523 Supplier sales to bidder sales 0.32 5.40 0.01 0.02 0.05 1,523 Supplier sales to target sales 0.94 10.59 0.01 0.03 0.12 1,523 Supplier Tobin’s Q 2.71 3.80 0.84 1.44 3.01 1,523 Supplier leverage 0.19 0.22 0.01 0.12 0.32 1,523 Supplier profitability 0.03 1.58 −0.02 0.07 0.16 1,523 Bidder supplier 0.70 0.46 0.00 1.00 1.00 1,523 Target supplier 0.38 0.49 0.00 0.00 1.00 1,523 There are 5,981 identifiable product market rivals (i.e., firms operating in TNIC industry of either bidder or target) in the 480 horizontal mergers with operating efficiency projections, out of which 5,269 have stock market and basic accounting data. As evident from Panel A, the market’s assessment of the effects of horizontal mergers on rivals is mixed and depends on the announcement return window. Within the [−1, 1] window, the mean return is insignificantly different from zero, while the median return is zero. Within the [−10, 10] window, the mean and median returns are around −1% and are statistically significant. Within the [−20, 20] window, the mean return in insignificantly positive 1.3%, while the median return is −1.4%. Value-weighted mean returns (not reported) are different. They range between 0.7% and 2.5% for the three announcement windows and are statistically significant.10 This discrepancy between equally weighted and value-weighted mean returns suggests that rival’s size is an important determinant of its reaction to merger announcement—a conjecture we investigate below. Importantly, it is difficult to draw conclusions regarding the effect of operating efficiency gains on rivals from the univariate statistics in Panel A. The reason is that there are multiple potential effects of a horizontal merger on rivals. On the one hand, operating efficiencies are expected to have a negative effect on rivals. On the other hand, there may be various positive effects of a merger on rivals. First, increased industry concentration is expected to positively impact rivals’ values through the market power channel. Second, the information revelation hypothesis posits that merger announcement may reveal positive information regarding merging firms’ stand-alone values. If this information is not entirely firm-specific and contains an industry component, then merger announcements may lead to a positive revaluation of rivals. Third, merger announcements impact the perceived likelihood that target’s rivals would become acquisition targets in the future. On the one hand, according to the deal anticipation hypothesis (e.g., Song and Walkling, 2000), merger announcements increase the likelihood of additional mergers in the industry, raising rivals’ values. On the other hand, if a bidder was contemplating a merger with multiple potential targets, then an announcement of the merger with one of them could reduce the perceived likelihood of other firms becoming acquisition targets, reducing their values. The evidence (e.g., Song and Walkling, 2000; Otchere and Ip, 2006; Cai, Song, and Walkling, 2011) suggests that the former effect dominates, and merger announcements raise the likelihood of rivals becoming future targets. The fact that the sign of predicted effect of operating efficiencies is different from the signs of predicted effects of increased market power, industry-level value-relevant information, and increased deal anticipation underscores the importance of measuring expected merger efficiency gains directly and relating them to rivals’ announcement returns. Not surprisingly, rivals exhibit high average product similarity with both the bidder and target: 9.3% and 10.3% on average, respectively. These means are comparable to mean bidder-target product similarity of 12.6%. A median rival’s book assets and sales are four times smaller than those of the bidder and around 40% smaller than those of the target. As follows from Panel B, mean announcement returns of customers are insignificantly positive, while the medians are negative but low in absolute value. Similar to the discussion above, it is difficult to draw conclusions from these univariate results, since the operating efficiency gains on the one hand and higher market power on the other hand are expected to have the opposite effects on customers. Consistent with Compustat Segment data being able to identify only the most important customers, those present in our sample are large—the median customer’s book assets are three times larger than those of the bidder and six times larger than those of the target. In 57% of the cases, the customer’s supplier is the bidder, while in 51% of the cases it is the target (in 8% of the cases the customer sources from both the bidder and target). Panel C shows that the mean and median announcement returns of suppliers are negative, large in absolute value, and statistically significant. Unlike customers, which are likely to benefit from operating efficiencies stemming from a merger, the effect of operating efficiencies on suppliers is ambiguous, as follows from empirical Prediction 3. This, coupled with the negative expected effect on suppliers of the merged firm’s increased market power, may explain the negative mean and median announcement returns to suppliers. Notably, unreported weighted-average mean returns to suppliers are close to zero and insignificant within most announcement windows, consistent with the evidence reported in Fee and Thomas (2004) and Shahrur (2005). We investigate the effect of suppliers’ size on their reactions to horizontal merger announcements below. Consistent with the Compustat-Segment-based sample being biased toward smaller suppliers, the median ratios of supplier’s assets to bidder’s assets and to target’s assets are 0.07 and 0.16, and similarly for sales-based ratios. In 70% of the cases, suppliers’ customers are bidders, while in 38% of the observations their customers are targets. To summarize, given that announcement returns of related firms reflect the net effect of merger-related operating efficiencies and changes in industry structure, as well as information revelation and deal anticipation effects, it is infeasible to draw unambiguous inferences about the effects of operating efficiencies on rivals, customers, and suppliers from mean and median announcement returns. Therefore, in the next subsection, we examine the relations between projected operating efficiency gains on the one hand and changes in values of related firms on the other hand, while controlling for other potential effects of mergers. 5.3 Effects of Operating Efficiency Gains on Rivals Table V reports estimates of regressions in which the dependent variable is a rival’s announcement return within each of the three announcement return windows. The independent variables include: estimated present value of projected operating efficiency gains; the number of rivals, proxying for pre-merger competitive structure of the bidder’s and target’s industry; implied change in merging firms’ industry HHI, proxying for expected change in industry participants’ market power; measures of product and technological similarity between the bidder and target, as well as the geographical overlap between them; deal characteristics; and rival’s characteristics, including rival’s product similarity to bidder and target. Similar to regressions of merging firms’ announcement returns in Table III, we also control for potential endogeneity of insiders’ choice to provide projections by including the inverse Mills ratio from the first-stage regression of propensity to disclose forecasts. Similar to regressions in Table II, rivals’ return regressions are estimated using OLS with year-fixed effects. Since there are multiple rivals in each deal, standard errors are clustered by merger. Table V. Rivals’ returns and projected operating efficiency gains This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s rival’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20]. The sample is all firms operating in the same TNIC industry as bidder or target in mergers with available efficiency projections. The main independent variable is the estimated value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I for the definitions of the rest of the variables. The inverse Mills ratio is computed using estimates of regression in Column 1 of Table II. The regressions are estimated with year fixed effects. Standard errors are clustered by merger. Standard errors are reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. Dependent variable Rival CAR [−1, 1] Rival CAR [−10, 10] Rival CAR [−20, 20] Intercept −0.024*** −0.139*** −0.112* (−5.29) (−4.70) (−1.89) PV of efficiency projections −0.056** −0.130*** −0.230*** (−2.35) (−2.95) (−2.60) Number of rivals 0.011* 0.012 0.013 (1.91) (1.58) (1.05) Implied change in HHI 0.015 0.020 −0.020 (0.95) (0.20) (−0.10) Bidder–target product similarity 0.019*** 0.091** 0.125 (2.63) (1.99) (1.36) Bidder–target technology similarity −0.004 −0.017 −0.010 (−1.33) (−0.85) (−0.24) Bidder–target same BEA 0.001 −0.003 −0.018 (0.92) (−0.31) (−0.97) Rival–bidder product similarity 0.001 0.007 −0.030 (0.06) (0.13) (−0.27) Rival–target product similarity 0.023*** 0.142*** 0.203* (2.80) (2.71) (1.93) Deregulation 0.005** 0.034** 0.051* (2.43) (2.54) (1.91) Number of competing bidders −0.002 −0.007 −0.018 (−1.08) (−0.60) (−0.78) Target sales/bidder sales −0.001 −0.005 −0.005 (−1.35) (−1.20) (−0.58) Log(rival sales) 0.002*** 0.013*** 0.016*** (6.45) (5.86) (3.81) Rival sales/bidder sales 0.000 0.000 0.000 (0.56) (0.86) (0.69) Rival sales/target sales 0.000 0.000 0.000 (0.18) (0.84) (0.16) Rival Tobin’s Q 0.000 0.000 0.004 (−1.41) (−0.07) (1.08) Rival leverage 0.000 0.024 0.042 (−0.04) (1.31) (1.11) Rival profitability 0.000 0.005* 0.009 (0.80) (1.80) (1.55) Inverse Mills ratio −0.001 0.001 0.004 (−0.78) (0.06) (0.19) Number of observation 4,955 4,955 4,955 Adjusted R-squared 0.060 0.058 0.037 Dependent variable Rival CAR [−1, 1] Rival CAR [−10, 10] Rival CAR [−20, 20] Intercept −0.024*** −0.139*** −0.112* (−5.29) (−4.70) (−1.89) PV of efficiency projections −0.056** −0.130*** −0.230*** (−2.35) (−2.95) (−2.60) Number of rivals 0.011* 0.012 0.013 (1.91) (1.58) (1.05) Implied change in HHI 0.015 0.020 −0.020 (0.95) (0.20) (−0.10) Bidder–target product similarity 0.019*** 0.091** 0.125 (2.63) (1.99) (1.36) Bidder–target technology similarity −0.004 −0.017 −0.010 (−1.33) (−0.85) (−0.24) Bidder–target same BEA 0.001 −0.003 −0.018 (0.92) (−0.31) (−0.97) Rival–bidder product similarity 0.001 0.007 −0.030 (0.06) (0.13) (−0.27) Rival–target product similarity 0.023*** 0.142*** 0.203* (2.80) (2.71) (1.93) Deregulation 0.005** 0.034** 0.051* (2.43) (2.54) (1.91) Number of competing bidders −0.002 −0.007 −0.018 (−1.08) (−0.60) (−0.78) Target sales/bidder sales −0.001 −0.005 −0.005 (−1.35) (−1.20) (−0.58) Log(rival sales) 0.002*** 0.013*** 0.016*** (6.45) (5.86) (3.81) Rival sales/bidder sales 0.000 0.000 0.000 (0.56) (0.86) (0.69) Rival sales/target sales 0.000 0.000 0.000 (0.18) (0.84) (0.16) Rival Tobin’s Q 0.000 0.000 0.004 (−1.41) (−0.07) (1.08) Rival leverage 0.000 0.024 0.042 (−0.04) (1.31) (1.11) Rival profitability 0.000 0.005* 0.009 (0.80) (1.80) (1.55) Inverse Mills ratio −0.001 0.001 0.004 (−0.78) (0.06) (0.19) Number of observation 4,955 4,955 4,955 Adjusted R-squared 0.060 0.058 0.037 Table V. Rivals’ returns and projected operating efficiency gains This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s rival’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20]. The sample is all firms operating in the same TNIC industry as bidder or target in mergers with available efficiency projections. The main independent variable is the estimated value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I for the definitions of the rest of the variables. The inverse Mills ratio is computed using estimates of regression in Column 1 of Table II. The regressions are estimated with year fixed effects. Standard errors are clustered by merger. Standard errors are reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. Dependent variable Rival CAR [−1, 1] Rival CAR [−10, 10] Rival CAR [−20, 20] Intercept −0.024*** −0.139*** −0.112* (−5.29) (−4.70) (−1.89) PV of efficiency projections −0.056** −0.130*** −0.230*** (−2.35) (−2.95) (−2.60) Number of rivals 0.011* 0.012 0.013 (1.91) (1.58) (1.05) Implied change in HHI 0.015 0.020 −0.020 (0.95) (0.20) (−0.10) Bidder–target product similarity 0.019*** 0.091** 0.125 (2.63) (1.99) (1.36) Bidder–target technology similarity −0.004 −0.017 −0.010 (−1.33) (−0.85) (−0.24) Bidder–target same BEA 0.001 −0.003 −0.018 (0.92) (−0.31) (−0.97) Rival–bidder product similarity 0.001 0.007 −0.030 (0.06) (0.13) (−0.27) Rival–target product similarity 0.023*** 0.142*** 0.203* (2.80) (2.71) (1.93) Deregulation 0.005** 0.034** 0.051* (2.43) (2.54) (1.91) Number of competing bidders −0.002 −0.007 −0.018 (−1.08) (−0.60) (−0.78) Target sales/bidder sales −0.001 −0.005 −0.005 (−1.35) (−1.20) (−0.58) Log(rival sales) 0.002*** 0.013*** 0.016*** (6.45) (5.86) (3.81) Rival sales/bidder sales 0.000 0.000 0.000 (0.56) (0.86) (0.69) Rival sales/target sales 0.000 0.000 0.000 (0.18) (0.84) (0.16) Rival Tobin’s Q 0.000 0.000 0.004 (−1.41) (−0.07) (1.08) Rival leverage 0.000 0.024 0.042 (−0.04) (1.31) (1.11) Rival profitability 0.000 0.005* 0.009 (0.80) (1.80) (1.55) Inverse Mills ratio −0.001 0.001 0.004 (−0.78) (0.06) (0.19) Number of observation 4,955 4,955 4,955 Adjusted R-squared 0.060 0.058 0.037 Dependent variable Rival CAR [−1, 1] Rival CAR [−10, 10] Rival CAR [−20, 20] Intercept −0.024*** −0.139*** −0.112* (−5.29) (−4.70) (−1.89) PV of efficiency projections −0.056** −0.130*** −0.230*** (−2.35) (−2.95) (−2.60) Number of rivals 0.011* 0.012 0.013 (1.91) (1.58) (1.05) Implied change in HHI 0.015 0.020 −0.020 (0.95) (0.20) (−0.10) Bidder–target product similarity 0.019*** 0.091** 0.125 (2.63) (1.99) (1.36) Bidder–target technology similarity −0.004 −0.017 −0.010 (−1.33) (−0.85) (−0.24) Bidder–target same BEA 0.001 −0.003 −0.018 (0.92) (−0.31) (−0.97) Rival–bidder product similarity 0.001 0.007 −0.030 (0.06) (0.13) (−0.27) Rival–target product similarity 0.023*** 0.142*** 0.203* (2.80) (2.71) (1.93) Deregulation 0.005** 0.034** 0.051* (2.43) (2.54) (1.91) Number of competing bidders −0.002 −0.007 −0.018 (−1.08) (−0.60) (−0.78) Target sales/bidder sales −0.001 −0.005 −0.005 (−1.35) (−1.20) (−0.58) Log(rival sales) 0.002*** 0.013*** 0.016*** (6.45) (5.86) (3.81) Rival sales/bidder sales 0.000 0.000 0.000 (0.56) (0.86) (0.69) Rival sales/target sales 0.000 0.000 0.000 (0.18) (0.84) (0.16) Rival Tobin’s Q 0.000 0.000 0.004 (−1.41) (−0.07) (1.08) Rival leverage 0.000 0.024 0.042 (−0.04) (1.31) (1.11) Rival profitability 0.000 0.005* 0.009 (0.80) (1.80) (1.55) Inverse Mills ratio −0.001 0.001 0.004 (−0.78) (0.06) (0.19) Number of observation 4,955 4,955 4,955 Adjusted R-squared 0.060 0.058 0.037 Consistent with Prediction 1, the relation between operating efficiency gain projections and rivals’ announcement returns is negative and significant at least at 5% level for all three announcement windows. This relation is also economically significant: a one-standard-deviation increase in projected operating efficiencies is associated with 1.3–2.6% reduction in rivals’ market reaction. Merger-related changes in industry structure are not significantly associated with rivals’ announcement returns. However, product similarity between bidder and target, as well as that between rival and target, are positively associated with rival’s reaction to deal announcement, potentially consistent with larger impact of mergers between firms with more similar/substitutable products on industry participants’ market power. Rivals’ returns are also higher in mergers that occurred following deregulation events, consistent with firms being able to exploit market power to a larger extent in less regulated industries. Rivals’ announcement returns are positively associated with their size, proxied by the logarithm of sales, consistent with rivals’ value-weighted mean announcement returns being higher than equally weighted mean returns. The coefficients on the inverse Mills ratio are insignificant in all three specifications, suggesting that self-selection of insiders into disclosing forecasts does not impact the relation between projected efficiencies and rivals’ values. The value of operating efficiency gains that a merger can generate likely depends on bidder’s and target’s production functions. In particular, Devos, Kadapakkam, and Krishnamurthy (2009) find that operating efficiencies are mostly realized through the merging firms’ ability to reduce joint capital expenditures. It follows that operating efficiency gains are more likely to be present in mergers between more capital-intensive bidders and targets, and insiders’ efficiency gain forecasts are expected to be more informative in mergers among relatively capital-intensive firms. In other words, we expect the market to capitalize a larger fraction of insiders’ efficiency gain forecasts, resulting in a stronger negative relation between projected gains and rivals’ announcement returns, in mergers among relatively capital-intensive bidders and targets. In Table VI, we test this conjecture. In Panel A, we split the sample of mergers into two subsamples. The first (second) subsample includes mergers involving bidders whose capital-stock-to-assets ratio in the last pre-merger year is above (below) median among all bidders that year. We then estimate regressions as in Table V for the two subsamples. To save space, we only report the coefficient on efficiency gain projections. Table VI. Rivals’ returns and projected operating efficiency gains—subsample analysis This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s rival’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20], and which are estimated for rivals within various subsamples of mergers. In Panel A, the high (low) subsample contains rivals of bidders and/or targets in mergers in which bidder’s capital-stock-to-value ratio exceeds (falls short of) that of all bidders in M&As announced in the same year. In Panel B, the high (low) subsample contains rivals of bidders and/or targets in mergers in which target’s capital-stock-to-value ratio exceeds (falls short of) that of all targets in M&As announced in the same year. In Panel C, the high (low) subsample contains rivals in mergers in which the capitalization by the market of projected efficiency gains, computed as the ratio of bidder’s and target’s combined CAR over the corresponding announcement window and the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization, is above (below) median for all M&As announced in the same year. In Panel D, the high (low) subsample contains rivals in mergers in which the ratio of the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization and the implied change in the merging firms’ industry HHI, is above (below) median for all M&As announced in the same year. In Panel E, the high (low) subsample contains customers in mergers in which target’s book assets is above (below) median for all M&As announced in the same year. The main independent variable is the value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I for the definitions of the rest of the variables. The regressions are estimated with year-fixed effects. Standard errors are clustered by merger and reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. We only report the coefficients and associated standard errors on the estimated value of projected efficiency gains. The difference column represents the difference in coefficient estimates between two subsamples. Its significance is computed using Wald test. Dependent variable Rival CAR [−1, 1] Rival CAR [−10, 10] Rival CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.122*** 0.004 −0.126*** −0.175** −0.077 −0.098 −0.332** −0.131 −0.201 PV of projected efficiency gains (−3.59) (0.53) (−3.62) (−2.12) (−1.63) (−1.03) (−2.00) (−1.38) (−1.05) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.074*** −0.028*** −0.046** −0.214*** −0.010 −0.204** −0.360** −0.009 −0.351* PV of projected efficiency gains (−4.59) (−3.22) (−2.51) (−2.60) (−0.22) (−2.17) (−2.17) (−0.09) (−1.81) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.067*** −0.034*** −0.033* −0.178* −0.097** −0.081 −0.273* −0.090 −0.183 PV of projected efficiency gains (−4.22) (−2.84) (−1.66) (−1.72) (−2.13) (−0.72) (−1.81) (−1.11) (−1.07) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.054*** −0.032 −0.022 −0.124* −0.083 −0.041 −0.213* −0.098 −0.115 PV of projected efficiency gains (−2.87) (−1.52) (−0.78) (−1.72) (−1.25) (−0.42) (−1.71) (−0.94) (−0.71) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.042** −0.017 −0.025 −0.100 −0.059 −0.041 −0.198* −0.118 −0.080 PV of projected efficiency gains (−2.03) (−1.44) (−1.05) (−1.35) (−1.11) (−0.45) (−1.77) (−1.30) (−0.56) Dependent variable Rival CAR [−1, 1] Rival CAR [−10, 10] Rival CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.122*** 0.004 −0.126*** −0.175** −0.077 −0.098 −0.332** −0.131 −0.201 PV of projected efficiency gains (−3.59) (0.53) (−3.62) (−2.12) (−1.63) (−1.03) (−2.00) (−1.38) (−1.05) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.074*** −0.028*** −0.046** −0.214*** −0.010 −0.204** −0.360** −0.009 −0.351* PV of projected efficiency gains (−4.59) (−3.22) (−2.51) (−2.60) (−0.22) (−2.17) (−2.17) (−0.09) (−1.81) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.067*** −0.034*** −0.033* −0.178* −0.097** −0.081 −0.273* −0.090 −0.183 PV of projected efficiency gains (−4.22) (−2.84) (−1.66) (−1.72) (−2.13) (−0.72) (−1.81) (−1.11) (−1.07) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.054*** −0.032 −0.022 −0.124* −0.083 −0.041 −0.213* −0.098 −0.115 PV of projected efficiency gains (−2.87) (−1.52) (−0.78) (−1.72) (−1.25) (−0.42) (−1.71) (−0.94) (−0.71) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.042** −0.017 −0.025 −0.100 −0.059 −0.041 −0.198* −0.118 −0.080 PV of projected efficiency gains (−2.03) (−1.44) (−1.05) (−1.35) (−1.11) (−0.45) (−1.77) (−1.30) (−0.56) Table VI. Rivals’ returns and projected operating efficiency gains—subsample analysis This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s rival’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20], and which are estimated for rivals within various subsamples of mergers. In Panel A, the high (low) subsample contains rivals of bidders and/or targets in mergers in which bidder’s capital-stock-to-value ratio exceeds (falls short of) that of all bidders in M&As announced in the same year. In Panel B, the high (low) subsample contains rivals of bidders and/or targets in mergers in which target’s capital-stock-to-value ratio exceeds (falls short of) that of all targets in M&As announced in the same year. In Panel C, the high (low) subsample contains rivals in mergers in which the capitalization by the market of projected efficiency gains, computed as the ratio of bidder’s and target’s combined CAR over the corresponding announcement window and the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization, is above (below) median for all M&As announced in the same year. In Panel D, the high (low) subsample contains rivals in mergers in which the ratio of the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization and the implied change in the merging firms’ industry HHI, is above (below) median for all M&As announced in the same year. In Panel E, the high (low) subsample contains customers in mergers in which target’s book assets is above (below) median for all M&As announced in the same year. The main independent variable is the value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I for the definitions of the rest of the variables. The regressions are estimated with year-fixed effects. Standard errors are clustered by merger and reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. We only report the coefficients and associated standard errors on the estimated value of projected efficiency gains. The difference column represents the difference in coefficient estimates between two subsamples. Its significance is computed using Wald test. Dependent variable Rival CAR [−1, 1] Rival CAR [−10, 10] Rival CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.122*** 0.004 −0.126*** −0.175** −0.077 −0.098 −0.332** −0.131 −0.201 PV of projected efficiency gains (−3.59) (0.53) (−3.62) (−2.12) (−1.63) (−1.03) (−2.00) (−1.38) (−1.05) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.074*** −0.028*** −0.046** −0.214*** −0.010 −0.204** −0.360** −0.009 −0.351* PV of projected efficiency gains (−4.59) (−3.22) (−2.51) (−2.60) (−0.22) (−2.17) (−2.17) (−0.09) (−1.81) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.067*** −0.034*** −0.033* −0.178* −0.097** −0.081 −0.273* −0.090 −0.183 PV of projected efficiency gains (−4.22) (−2.84) (−1.66) (−1.72) (−2.13) (−0.72) (−1.81) (−1.11) (−1.07) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.054*** −0.032 −0.022 −0.124* −0.083 −0.041 −0.213* −0.098 −0.115 PV of projected efficiency gains (−2.87) (−1.52) (−0.78) (−1.72) (−1.25) (−0.42) (−1.71) (−0.94) (−0.71) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.042** −0.017 −0.025 −0.100 −0.059 −0.041 −0.198* −0.118 −0.080 PV of projected efficiency gains (−2.03) (−1.44) (−1.05) (−1.35) (−1.11) (−0.45) (−1.77) (−1.30) (−0.56) Dependent variable Rival CAR [−1, 1] Rival CAR [−10, 10] Rival CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.122*** 0.004 −0.126*** −0.175** −0.077 −0.098 −0.332** −0.131 −0.201 PV of projected efficiency gains (−3.59) (0.53) (−3.62) (−2.12) (−1.63) (−1.03) (−2.00) (−1.38) (−1.05) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.074*** −0.028*** −0.046** −0.214*** −0.010 −0.204** −0.360** −0.009 −0.351* PV of projected efficiency gains (−4.59) (−3.22) (−2.51) (−2.60) (−0.22) (−2.17) (−2.17) (−0.09) (−1.81) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.067*** −0.034*** −0.033* −0.178* −0.097** −0.081 −0.273* −0.090 −0.183 PV of projected efficiency gains (−4.22) (−2.84) (−1.66) (−1.72) (−2.13) (−0.72) (−1.81) (−1.11) (−1.07) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.054*** −0.032 −0.022 −0.124* −0.083 −0.041 −0.213* −0.098 −0.115 PV of projected efficiency gains (−2.87) (−1.52) (−0.78) (−1.72) (−1.25) (−0.42) (−1.71) (−0.94) (−0.71) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on −0.042** −0.017 −0.025 −0.100 −0.059 −0.041 −0.198* −0.118 −0.080 PV of projected efficiency gains (−2.03) (−1.44) (−1.05) (−1.35) (−1.11) (−0.45) (−1.77) (−1.30) (−0.56) Consistent with the market capitalizing a larger fraction of insiders’ forecasts in mergers involving capital-intensive bidders, the coefficients on efficiency gain projections are more negative in the subsample of mergers involving bidders with above-median capital stock than in the subsample of mergers involving low-capital-stock bidders. These coefficients are significantly negative in the former subsample for all three announcement windows, while they are insignificant for all three windows in the latter subsample. For the [−1, 1] window, the difference between the coefficients in the two subsamples is significant at 1% level. In Panel B, we split the sample to above-median and below-median target’s capital-stock-to-assets ratio. The results are more striking than those in Panel A. For all three event windows, the coefficient on efficiency gain projections is significantly negative in the capital-intensive-target subsample, whereas it is never significantly negative in the low-capital-intensity subsample. Moreover, the differences in the estimated coefficients between the two subsamples are statistically significant for all three announcement windows, at least at 10% level. A more direct way to infer the proportion of insiders’ projected operating efficiency gains that the market capitalizes is to compare projected gains with ex-post (realized) changes in merging firms’ values around merger announcements. Note that our measure of projected efficiency gains is scaled by merging firms’ pre-merger market capitalization. Bidder’s and target’s value-weighted mean announcement return is the change in bidder’s and target’s combined value around merger announcement, also scaled by their pre-merger combined market capitalization. Thus, to estimate the proportion of projected synergies that is capitalized by the market, for every merger we divide merging firms’ value-weighted announcement return by scaled projected operating efficiency gains. We winsorize this ratio at zero and at one.11 In panel C, we split the sample to subsamples with above-median and below-median proportions of projected efficiency gains capitalized by the market. The relation between efficiency gain forecasts and rivals’ announcement returns is more negative within the high-capitalization subsample than within the low-capitalization one for all three announcement windows, the difference in the coefficients being significant at 10% level in the [−1, 1] window. A potential issue with splitting the sample of mergers based on the ratio of the market reaction to merger announcement and efficiency gain projection is that the numerator of this ratio is a noisy proxy for the market’s capitalization of operating efficiency gain and can take negative values. In Panel D, we split the sample based on the ratio of forecasted efficiency gains and implied change in HHI in the merging firms’ industry. This ratio can be interpreted as a measure of the importance of the operating efficiency effect relative to the market power effect. We expect the association between efficiency gain projections and rivals’ announcement returns to be stronger in the subsample with above-median ratio of projected efficiency gains and implied change in HHI. The results are consistent with this conjecture. This association is significantly negative in the subsample of mergers with above-median efficiency gains-to-change in HHI ratio and it is substantially weaker and insignificant in the low efficiency gains-to-change in HHI subsample. In Panel E, we examine the effect of the size of the target firm on the relation between efficiency gains and rivals’ announcement returns. Efficiency gains in mergers involving larger targets are expected to have a more significant impact on all related firms, rivals in particular. In addition, as in the majority of large mergers insiders disclose operating efficiency gain projections, the selection bias is less likely to impact the results in the subsample of large deals. The results show that the magnitude of the negative coefficients on the efficiency gains projection in the above-median-size subsample is more than twice the magnitude of the coefficients in the below-median-size subsample. The coefficients are significant at least at 10% level for two announcement windows in the former subsample, whereas they are insignificant in the latter one. Overall, the results in Table VI suggest that market participants do not treat all insiders’ efficiency gain forecasts as equal. The negative relation between rivals’ announcement returns and insiders’ efficiency gain forecasts is stronger within subsamples of mergers in which the market believes that these forecasts are more likely to materialize, those that are less likely to capture non-efficiency-gain effects of mergers, and those that are likely to have a larger impact on rivals. 5.4 Customers Table VII reports estimates of regressions in which the dependent variable is the announcement returns of merging firms’ customers. The explanatory variables and the structure of the table are similar to those in Table V, with the sole addition of two indicator variables equalling one if a customer sources from bidder/target. Table VII. Customers’ returns and projected operating efficiency gains This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s customer’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20]. The sample is all principal customers of bidder and/or target, as defined in Compustat Industry Segment database. The main independent variable is the estimated value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I, and Table IV for the definitions of the rest of the variables. The inverse Mills ratio is computed using estimates of regression in Column 1 of Table II. The regressions are estimated with year-fixed effects. Standard errors are clustered by merger. Standard errors are reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. Dependent variable Customer CAR [−1, 1] Customer CAR [−10, 10] Customer CAR [−20, 20] Intercept −0.022 −0.142 −0.191 (−1.45) (−1.38) (−0.93) PV of projected efficiency gains 0.005 0.031** 0.099** (1.15) (2.28) (2.45) Number of rivals 0.001 0.005* 0.012** (1.42) (1.94) (2.38) Implied change in HHI −0.015 −0.318** −0.799*** (−0.77) (−2.44) (−3.06) Bidder–target product similarity 0.023 −0.061 −0.153 (1.32) (−0.52) (−0.65) Bidder–target technology similarity 0.011 0.031 0.067 (1.61) (0.66) (0.71) Same BEA −0.003 −0.007 −0.005 (−0.80) (−0.29) (−0.11) Deregulation 0.001 0.022 0.036 (0.23) (0.53) (0.43) Number of competing bidders 0.005 0.024 0.035 (1.18) (0.80) (0.59) Target sales/bidder sales 0.000 −0.007 −0.011 (0.25) (−0.60) (−0.51) Log(customer sales) 0.002* 0.013* 0.016 (1.67) (1.70) (1.07) Customer sales/bidder sales −0.009** −0.072*** −0.131*** (−2.54) (−3.18) (−2.87) Customer sales/target sales 0.005*** 0.036*** 0.079*** (2.75) (3.03) (3.29) Customer Tobin’s Q −0.002** −0.018*** −0.024* (−2.09) (−2.74) (−1.89) Customer leverage −0.007 −0.070 −0.117 (−0.73) (−1.14) (−0.94) Customer profitability 0.019 0.114 0.065 (1.23) (1.10) (0.31) Bidder’s customer −0.001 −0.047 −0.111 (−0.15) (−1.00) (−1.18) Target’s customer −0.009 −0.060 −0.104 (−1.39) (−1.32) (−1.13) Inverse Mills ratio 0.003 0.053* 0.123** (0.65) (1.91) (2.21) Number of observation 560 560 560 Adjusted R-squared 0.027 0.037 0.035 Dependent variable Customer CAR [−1, 1] Customer CAR [−10, 10] Customer CAR [−20, 20] Intercept −0.022 −0.142 −0.191 (−1.45) (−1.38) (−0.93) PV of projected efficiency gains 0.005 0.031** 0.099** (1.15) (2.28) (2.45) Number of rivals 0.001 0.005* 0.012** (1.42) (1.94) (2.38) Implied change in HHI −0.015 −0.318** −0.799*** (−0.77) (−2.44) (−3.06) Bidder–target product similarity 0.023 −0.061 −0.153 (1.32) (−0.52) (−0.65) Bidder–target technology similarity 0.011 0.031 0.067 (1.61) (0.66) (0.71) Same BEA −0.003 −0.007 −0.005 (−0.80) (−0.29) (−0.11) Deregulation 0.001 0.022 0.036 (0.23) (0.53) (0.43) Number of competing bidders 0.005 0.024 0.035 (1.18) (0.80) (0.59) Target sales/bidder sales 0.000 −0.007 −0.011 (0.25) (−0.60) (−0.51) Log(customer sales) 0.002* 0.013* 0.016 (1.67) (1.70) (1.07) Customer sales/bidder sales −0.009** −0.072*** −0.131*** (−2.54) (−3.18) (−2.87) Customer sales/target sales 0.005*** 0.036*** 0.079*** (2.75) (3.03) (3.29) Customer Tobin’s Q −0.002** −0.018*** −0.024* (−2.09) (−2.74) (−1.89) Customer leverage −0.007 −0.070 −0.117 (−0.73) (−1.14) (−0.94) Customer profitability 0.019 0.114 0.065 (1.23) (1.10) (0.31) Bidder’s customer −0.001 −0.047 −0.111 (−0.15) (−1.00) (−1.18) Target’s customer −0.009 −0.060 −0.104 (−1.39) (−1.32) (−1.13) Inverse Mills ratio 0.003 0.053* 0.123** (0.65) (1.91) (2.21) Number of observation 560 560 560 Adjusted R-squared 0.027 0.037 0.035 Table VII. Customers’ returns and projected operating efficiency gains This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s customer’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20]. The sample is all principal customers of bidder and/or target, as defined in Compustat Industry Segment database. The main independent variable is the estimated value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I, and Table IV for the definitions of the rest of the variables. The inverse Mills ratio is computed using estimates of regression in Column 1 of Table II. The regressions are estimated with year-fixed effects. Standard errors are clustered by merger. Standard errors are reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. Dependent variable Customer CAR [−1, 1] Customer CAR [−10, 10] Customer CAR [−20, 20] Intercept −0.022 −0.142 −0.191 (−1.45) (−1.38) (−0.93) PV of projected efficiency gains 0.005 0.031** 0.099** (1.15) (2.28) (2.45) Number of rivals 0.001 0.005* 0.012** (1.42) (1.94) (2.38) Implied change in HHI −0.015 −0.318** −0.799*** (−0.77) (−2.44) (−3.06) Bidder–target product similarity 0.023 −0.061 −0.153 (1.32) (−0.52) (−0.65) Bidder–target technology similarity 0.011 0.031 0.067 (1.61) (0.66) (0.71) Same BEA −0.003 −0.007 −0.005 (−0.80) (−0.29) (−0.11) Deregulation 0.001 0.022 0.036 (0.23) (0.53) (0.43) Number of competing bidders 0.005 0.024 0.035 (1.18) (0.80) (0.59) Target sales/bidder sales 0.000 −0.007 −0.011 (0.25) (−0.60) (−0.51) Log(customer sales) 0.002* 0.013* 0.016 (1.67) (1.70) (1.07) Customer sales/bidder sales −0.009** −0.072*** −0.131*** (−2.54) (−3.18) (−2.87) Customer sales/target sales 0.005*** 0.036*** 0.079*** (2.75) (3.03) (3.29) Customer Tobin’s Q −0.002** −0.018*** −0.024* (−2.09) (−2.74) (−1.89) Customer leverage −0.007 −0.070 −0.117 (−0.73) (−1.14) (−0.94) Customer profitability 0.019 0.114 0.065 (1.23) (1.10) (0.31) Bidder’s customer −0.001 −0.047 −0.111 (−0.15) (−1.00) (−1.18) Target’s customer −0.009 −0.060 −0.104 (−1.39) (−1.32) (−1.13) Inverse Mills ratio 0.003 0.053* 0.123** (0.65) (1.91) (2.21) Number of observation 560 560 560 Adjusted R-squared 0.027 0.037 0.035 Dependent variable Customer CAR [−1, 1] Customer CAR [−10, 10] Customer CAR [−20, 20] Intercept −0.022 −0.142 −0.191 (−1.45) (−1.38) (−0.93) PV of projected efficiency gains 0.005 0.031** 0.099** (1.15) (2.28) (2.45) Number of rivals 0.001 0.005* 0.012** (1.42) (1.94) (2.38) Implied change in HHI −0.015 −0.318** −0.799*** (−0.77) (−2.44) (−3.06) Bidder–target product similarity 0.023 −0.061 −0.153 (1.32) (−0.52) (−0.65) Bidder–target technology similarity 0.011 0.031 0.067 (1.61) (0.66) (0.71) Same BEA −0.003 −0.007 −0.005 (−0.80) (−0.29) (−0.11) Deregulation 0.001 0.022 0.036 (0.23) (0.53) (0.43) Number of competing bidders 0.005 0.024 0.035 (1.18) (0.80) (0.59) Target sales/bidder sales 0.000 −0.007 −0.011 (0.25) (−0.60) (−0.51) Log(customer sales) 0.002* 0.013* 0.016 (1.67) (1.70) (1.07) Customer sales/bidder sales −0.009** −0.072*** −0.131*** (−2.54) (−3.18) (−2.87) Customer sales/target sales 0.005*** 0.036*** 0.079*** (2.75) (3.03) (3.29) Customer Tobin’s Q −0.002** −0.018*** −0.024* (−2.09) (−2.74) (−1.89) Customer leverage −0.007 −0.070 −0.117 (−0.73) (−1.14) (−0.94) Customer profitability 0.019 0.114 0.065 (1.23) (1.10) (0.31) Bidder’s customer −0.001 −0.047 −0.111 (−0.15) (−1.00) (−1.18) Target’s customer −0.009 −0.060 −0.104 (−1.39) (−1.32) (−1.13) Inverse Mills ratio 0.003 0.053* 0.123** (0.65) (1.91) (2.21) Number of observation 560 560 560 Adjusted R-squared 0.027 0.037 0.035 Customers’ announcement returns are increasing in the efficiency gain projections. This relation is statistically significant at 5% level for [−10, 10] and [−20, 20] announcement windows. It is also economically sizable: a one-standard-deviation increase in projected efficiency gains is associated with 0.3% increase in the [−10, 10] return and with 1.1% increase in the [−20, 20] return. Interestingly, unlike in the case of rivals, the market does not immediately incorporate the information in insiders’ efficiency gain forecasts into customers’ prices, as the relation between forecasted gains and customers’ market reaction is insignificant for [−1, 1] announcement window and it is three times larger for [−20, 20] window than for [−10, 10] one. The number of product market rivals in bidder’s and target’s industry is positively related to customers’ announcement returns for two out of three announcement windows, consistent with the weaker (negative) market power effect on customers in more competitive industries. Similarly, implied change in industry HHI is negatively related to customers’ returns, consistent with larger increases in merging firms’ market power being more detrimental to their customers. These findings, which are in line with the market power effect of horizontal M&As on merging firms’ customers, are different from the corresponding results in Fee and Thomas (2004) and Shahrur (2005), who report a (mostly insignificant) positive relation between market power measures and customers’ announcement returns. To attempt to reconcile our findings with those of Fee and Thomas (2004) and Shahrur (2005), we estimate regressions as in Table VII, while excluding projected efficiency gains and the inverse Mills ratio from the set of explanatory variables. In these regressions, which are found in the Online Appendix, the coefficient on the implied change in bidder’s and target’s industry HHI is about one-third smaller in magnitude than in the regressions reported in Table VII for all three announcement windows, and becomes largely statistically insignificant (it remains significant at 10% level only for the [−20, 20] announcement window). This result further illustrates the importance of controlling for operating efficiency gains when examining the effects of changes in merging firms’ market power along their supply chains. In addition, customers’ announcement returns are positively related to their size, as proxied by sales, consistent with smaller customers being hurt more by merging firms’ increased market power. The coefficients on the inverse Mills ratio are statistically significant at least at 10% level in two specifications, suggesting that endogeneity of the decision to disclose projections impacts the relation between forecasted efficiency gains and customers’ market reactions. Similar to the analysis of rivals, we split the sample of customers into subsamples of mergers with: above-median and below-median bidder’s and target’s capital-stock-to-assets ratio; above-median and below-median ratio of merging firms’ announcement returns to forecasted efficiency gains; above-median and below-median ratio of efficiency gain forecast to implied change in HHI; and above-median and below-median target size. Similar to the case of rivals, the results suggest that not all insiders’ projections of efficiency gains have the same impact on merging firm customers’ values. The positive relation between projected gains and customers’ announcement returns tends to be significantly stronger in mergers involving capital-intensive bidders than in mergers in which bidders have below-median capital stock: the differences between the coefficients on efficiency gain projections within the two subsamples are significant for all three announcement windows. However, the results of sorting mergers by target’s capital intensity do not reveal differences in the relation between efficiency gain projections and customers’ announcement returns. Splitting the sample based on ex-post capitalization of these projections by the market shows that this relation is positive and significant at 10% level within the subsample of mergers with above-median capitalization of projections for all three announcement windows, whereas it is statistically insignificant within the subsample of mergers with below-median projection capitalization for all three announcement windows. We obtain similar results when we split the sample by the ratio of projected efficiency gains to implied change in HHI. Target-size-based split suggests that the effect of operating efficiency gains on customers tends to be somewhat more pronounced for mergers involving larger targets, but in both subsamples the relation is significant for only one announcement window. 5.5 Suppliers Our third empirical prediction states that the effect of operating efficiencies on suppliers depends on the source of efficiency gains. When operating efficiencies stem from cost savings, their effect on suppliers is expected to be positive, as reduction in marginal costs of production raises the merged firm’s equilibrium demand for inputs. In contrast, when operating efficiencies are due to productivity gains, their effect on suppliers is expected to be negative because increased productivity reduces the demand for inputs for a given level of output. Notably, it appears from reading the text in insiders’ projections of operating efficiencies, such as those found in the Appendix, that in the majority of cases, projected operating efficiencies are expected to take the form of cost savings. Thus, while the sign of the average effect of operating efficiencies on suppliers is ultimately an empirical question, we expect the overall relation between operating efficiency gain forecasts and suppliers’ market reactions to be positive. In Table IX, we report results of estimating regressions of suppliers’ announcement returns on forecasted efficiency gains and on control variables similar to those in the rivals’ and customers’ regressions. Table VIII. Customers’ returns and projected operating efficiency gains—subsample analysis This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s customer’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20], and which are estimated for customers within various subsamples of mergers. In Panel A, the high (low) subsample contains customers of bidders and/or targets in mergers in which bidder’s capital-stock-to-value ratio exceeds (falls short of) that of all bidders in M&As announced in the same year. In Panel B, the high (low) subsample contains customers of bidders and/or targets in mergers in which target’s capital-stock-to-value ratio exceeds (falls short of) that of all targets in M&As announced in the same year. In Panel C, the high (low) subsample contains customers in mergers in which the capitalization by the market of projected efficiency gains, computed as the ratio of bidder’s and target’s combined CAR over the corresponding announcement window and the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization, is above (below) median for all M&As announced in the same year. In Panel D, the high (low) subsample contains customers in mergers in which the ratio of the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization and the implied change in the merging firms’ industry HHI, is above (below) median for all M&As announced in the same year. In Panel E, the high (low) subsample contains customers in mergers in which target’s book assets is above (below) median for all M&As announced in the same year. The main independent variable is the value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I for the definitions of the rest of the variables. The regressions are estimated with year-fixed effects. Standard errors are clustered by merger and reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. We only report the coefficients and associated standard errors on the estimated value of projected efficiency gains. The difference column represents the difference in coefficient estimates between two subsamples. Its significance is computed using Wald test. Dependent variable Customer CAR [−1, 1] Customer CAR [−10, 10] Customer CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.010** −0.003 0.012*** 0.068** −0.005 0.073** 0.134** −0.037 0.171* PV of projected efficiency gains (2.47) (−1.41) (2.84) (2.40) (−0.35) (2.32) (2.34) (−0.45) (1.71) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.005 0.006 −0.002 0.036* 0.028 0.008 0.111* 0.097 0.014 PV of projected efficiency gains (1.16) (1.23) (−0.23) (1.82) (1.39) (0.28) (1.71) (1.60) (0.16) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.008* 0.003 0.005 0.035* 0.023 0.012 0.152** 0.069 0.083 PV of projected operating efficiency gains (1.91) (1.12) (1.09) (1.72) (1.20) (0.44) (2.40) (0.99) (0.88) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.010** 0.002 0.008* 0.031* 0.019 0.012 0.180*** 0.075 0.105 PV of projected efficiency gains (2.19) (1.20) (1.65) (1.90) (1.11) (0.51) (2.68) (1.19) (1.14) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.007** 0.005* 0.002 0.047 0.020 0.027 0.130 0.105 0.025 PV of projected efficiency gains (1.99) (1.70) (0.44) (1.55) (1.20) (0.78) (1.50) (0.99) (0.18) Dependent variable Customer CAR [−1, 1] Customer CAR [−10, 10] Customer CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.010** −0.003 0.012*** 0.068** −0.005 0.073** 0.134** −0.037 0.171* PV of projected efficiency gains (2.47) (−1.41) (2.84) (2.40) (−0.35) (2.32) (2.34) (−0.45) (1.71) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.005 0.006 −0.002 0.036* 0.028 0.008 0.111* 0.097 0.014 PV of projected efficiency gains (1.16) (1.23) (−0.23) (1.82) (1.39) (0.28) (1.71) (1.60) (0.16) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.008* 0.003 0.005 0.035* 0.023 0.012 0.152** 0.069 0.083 PV of projected operating efficiency gains (1.91) (1.12) (1.09) (1.72) (1.20) (0.44) (2.40) (0.99) (0.88) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.010** 0.002 0.008* 0.031* 0.019 0.012 0.180*** 0.075 0.105 PV of projected efficiency gains (2.19) (1.20) (1.65) (1.90) (1.11) (0.51) (2.68) (1.19) (1.14) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.007** 0.005* 0.002 0.047 0.020 0.027 0.130 0.105 0.025 PV of projected efficiency gains (1.99) (1.70) (0.44) (1.55) (1.20) (0.78) (1.50) (0.99) (0.18) Table VIII. Customers’ returns and projected operating efficiency gains—subsample analysis This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s customer’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20], and which are estimated for customers within various subsamples of mergers. In Panel A, the high (low) subsample contains customers of bidders and/or targets in mergers in which bidder’s capital-stock-to-value ratio exceeds (falls short of) that of all bidders in M&As announced in the same year. In Panel B, the high (low) subsample contains customers of bidders and/or targets in mergers in which target’s capital-stock-to-value ratio exceeds (falls short of) that of all targets in M&As announced in the same year. In Panel C, the high (low) subsample contains customers in mergers in which the capitalization by the market of projected efficiency gains, computed as the ratio of bidder’s and target’s combined CAR over the corresponding announcement window and the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization, is above (below) median for all M&As announced in the same year. In Panel D, the high (low) subsample contains customers in mergers in which the ratio of the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization and the implied change in the merging firms’ industry HHI, is above (below) median for all M&As announced in the same year. In Panel E, the high (low) subsample contains customers in mergers in which target’s book assets is above (below) median for all M&As announced in the same year. The main independent variable is the value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I for the definitions of the rest of the variables. The regressions are estimated with year-fixed effects. Standard errors are clustered by merger and reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. We only report the coefficients and associated standard errors on the estimated value of projected efficiency gains. The difference column represents the difference in coefficient estimates between two subsamples. Its significance is computed using Wald test. Dependent variable Customer CAR [−1, 1] Customer CAR [−10, 10] Customer CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.010** −0.003 0.012*** 0.068** −0.005 0.073** 0.134** −0.037 0.171* PV of projected efficiency gains (2.47) (−1.41) (2.84) (2.40) (−0.35) (2.32) (2.34) (−0.45) (1.71) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.005 0.006 −0.002 0.036* 0.028 0.008 0.111* 0.097 0.014 PV of projected efficiency gains (1.16) (1.23) (−0.23) (1.82) (1.39) (0.28) (1.71) (1.60) (0.16) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.008* 0.003 0.005 0.035* 0.023 0.012 0.152** 0.069 0.083 PV of projected operating efficiency gains (1.91) (1.12) (1.09) (1.72) (1.20) (0.44) (2.40) (0.99) (0.88) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.010** 0.002 0.008* 0.031* 0.019 0.012 0.180*** 0.075 0.105 PV of projected efficiency gains (2.19) (1.20) (1.65) (1.90) (1.11) (0.51) (2.68) (1.19) (1.14) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.007** 0.005* 0.002 0.047 0.020 0.027 0.130 0.105 0.025 PV of projected efficiency gains (1.99) (1.70) (0.44) (1.55) (1.20) (0.78) (1.50) (0.99) (0.18) Dependent variable Customer CAR [−1, 1] Customer CAR [−10, 10] Customer CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.010** −0.003 0.012*** 0.068** −0.005 0.073** 0.134** −0.037 0.171* PV of projected efficiency gains (2.47) (−1.41) (2.84) (2.40) (−0.35) (2.32) (2.34) (−0.45) (1.71) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.005 0.006 −0.002 0.036* 0.028 0.008 0.111* 0.097 0.014 PV of projected efficiency gains (1.16) (1.23) (−0.23) (1.82) (1.39) (0.28) (1.71) (1.60) (0.16) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.008* 0.003 0.005 0.035* 0.023 0.012 0.152** 0.069 0.083 PV of projected operating efficiency gains (1.91) (1.12) (1.09) (1.72) (1.20) (0.44) (2.40) (0.99) (0.88) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.010** 0.002 0.008* 0.031* 0.019 0.012 0.180*** 0.075 0.105 PV of projected efficiency gains (2.19) (1.20) (1.65) (1.90) (1.11) (0.51) (2.68) (1.19) (1.14) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.007** 0.005* 0.002 0.047 0.020 0.027 0.130 0.105 0.025 PV of projected efficiency gains (1.99) (1.70) (0.44) (1.55) (1.20) (0.78) (1.50) (0.99) (0.18) Table IX. Suppliers’ returns and projected operating efficiency gains This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s supplier’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20]. The sample is all firms to which bidder and/or target is a principal customer, as defined in Compustat Industry Segment database. The main independent variable is the estimated value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I, and Table IV for the definitions of the rest of the variables. The inverse Mills ratio is computed using estimates of regression in Column 1 of Table II. The regressions are estimated with year-fixed effects. Standard errors are clustered by merger. Standard errors are reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. Dependent variable Supplier CAR [−1, 1] Supplier CAR [−10, 10] Supplier CAR [−20, 20] Intercept −0.001 −0.110* −0.136 (−0.09) (−1.83) (−1.19) PV of projected efficiency gains 0.053* 0.158* 0.235* (1.70) (1.90) (1.70) Number of rivals −0.001 −0.006 −0.009 (−1.17) (−1.16) (−1.36) Implied change in HHI −0.052*** −0.220** −0.269 (−2.70) (−2.04) (−1.31) Bidder–target product similarity 0.044* 0.323** 0.633*** (1.93) (2.52) (2.60) Bidder–target technology similarity −0.017*** −0.046 −0.044 (−3.04) (−1.46) (−0.74) Same BEA 0.006* −0.002 −0.019 (1.74) (−0.09) (−0.50) Deregulation −0.004 −0.020 −0.022 (−0.89) (−0.71) (−0.42) Number of competing bidders −0.001 0.016 0.039 (−0.23) (0.85) (1.06) Target sales/bidder sales −0.005** −0.016 −0.029 (−2.40) (−1.42) (−1.32) Log(supplier sales) 0.002** 0.014*** 0.017* (2.26) (2.86) (1.95) Supplier sales/bidder sales 0.000 −0.001 0.000 (−0.54) (−0.22) (0.02) Supplier sales/target sales 0.000 −0.001 −0.003 (−0.88) (−1.20) (−1.28) Supplier Tobin’s Q −0.001** −0.003 −0.004 (−2.40) (−1.14) (−0.76) Supplier leverage 0.008 0.048 0.075 (0.99) (1.10) (0.92) Supplier profitability 0.000 0.000 −0.001 (−0.19) (0.07) (−0.11) Bidder’s supplier −0.001 0.016 0.034 (−0.18) (0.45) (0.51) Target’s supplier −0.002 0.024 0.057 (−0.28) (0.75) (0.94) Inverse Mills ratio 0.000 0.035 0.087 (0.00) (1.22) (1.58) Number of observation 1,523 1,523 1,523 Adjusted R-squared 0.050 0.029 0.015 Dependent variable Supplier CAR [−1, 1] Supplier CAR [−10, 10] Supplier CAR [−20, 20] Intercept −0.001 −0.110* −0.136 (−0.09) (−1.83) (−1.19) PV of projected efficiency gains 0.053* 0.158* 0.235* (1.70) (1.90) (1.70) Number of rivals −0.001 −0.006 −0.009 (−1.17) (−1.16) (−1.36) Implied change in HHI −0.052*** −0.220** −0.269 (−2.70) (−2.04) (−1.31) Bidder–target product similarity 0.044* 0.323** 0.633*** (1.93) (2.52) (2.60) Bidder–target technology similarity −0.017*** −0.046 −0.044 (−3.04) (−1.46) (−0.74) Same BEA 0.006* −0.002 −0.019 (1.74) (−0.09) (−0.50) Deregulation −0.004 −0.020 −0.022 (−0.89) (−0.71) (−0.42) Number of competing bidders −0.001 0.016 0.039 (−0.23) (0.85) (1.06) Target sales/bidder sales −0.005** −0.016 −0.029 (−2.40) (−1.42) (−1.32) Log(supplier sales) 0.002** 0.014*** 0.017* (2.26) (2.86) (1.95) Supplier sales/bidder sales 0.000 −0.001 0.000 (−0.54) (−0.22) (0.02) Supplier sales/target sales 0.000 −0.001 −0.003 (−0.88) (−1.20) (−1.28) Supplier Tobin’s Q −0.001** −0.003 −0.004 (−2.40) (−1.14) (−0.76) Supplier leverage 0.008 0.048 0.075 (0.99) (1.10) (0.92) Supplier profitability 0.000 0.000 −0.001 (−0.19) (0.07) (−0.11) Bidder’s supplier −0.001 0.016 0.034 (−0.18) (0.45) (0.51) Target’s supplier −0.002 0.024 0.057 (−0.28) (0.75) (0.94) Inverse Mills ratio 0.000 0.035 0.087 (0.00) (1.22) (1.58) Number of observation 1,523 1,523 1,523 Adjusted R-squared 0.050 0.029 0.015 Table IX. Suppliers’ returns and projected operating efficiency gains This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s supplier’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20]. The sample is all firms to which bidder and/or target is a principal customer, as defined in Compustat Industry Segment database. The main independent variable is the estimated value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I, and Table IV for the definitions of the rest of the variables. The inverse Mills ratio is computed using estimates of regression in Column 1 of Table II. The regressions are estimated with year-fixed effects. Standard errors are clustered by merger. Standard errors are reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. Dependent variable Supplier CAR [−1, 1] Supplier CAR [−10, 10] Supplier CAR [−20, 20] Intercept −0.001 −0.110* −0.136 (−0.09) (−1.83) (−1.19) PV of projected efficiency gains 0.053* 0.158* 0.235* (1.70) (1.90) (1.70) Number of rivals −0.001 −0.006 −0.009 (−1.17) (−1.16) (−1.36) Implied change in HHI −0.052*** −0.220** −0.269 (−2.70) (−2.04) (−1.31) Bidder–target product similarity 0.044* 0.323** 0.633*** (1.93) (2.52) (2.60) Bidder–target technology similarity −0.017*** −0.046 −0.044 (−3.04) (−1.46) (−0.74) Same BEA 0.006* −0.002 −0.019 (1.74) (−0.09) (−0.50) Deregulation −0.004 −0.020 −0.022 (−0.89) (−0.71) (−0.42) Number of competing bidders −0.001 0.016 0.039 (−0.23) (0.85) (1.06) Target sales/bidder sales −0.005** −0.016 −0.029 (−2.40) (−1.42) (−1.32) Log(supplier sales) 0.002** 0.014*** 0.017* (2.26) (2.86) (1.95) Supplier sales/bidder sales 0.000 −0.001 0.000 (−0.54) (−0.22) (0.02) Supplier sales/target sales 0.000 −0.001 −0.003 (−0.88) (−1.20) (−1.28) Supplier Tobin’s Q −0.001** −0.003 −0.004 (−2.40) (−1.14) (−0.76) Supplier leverage 0.008 0.048 0.075 (0.99) (1.10) (0.92) Supplier profitability 0.000 0.000 −0.001 (−0.19) (0.07) (−0.11) Bidder’s supplier −0.001 0.016 0.034 (−0.18) (0.45) (0.51) Target’s supplier −0.002 0.024 0.057 (−0.28) (0.75) (0.94) Inverse Mills ratio 0.000 0.035 0.087 (0.00) (1.22) (1.58) Number of observation 1,523 1,523 1,523 Adjusted R-squared 0.050 0.029 0.015 Dependent variable Supplier CAR [−1, 1] Supplier CAR [−10, 10] Supplier CAR [−20, 20] Intercept −0.001 −0.110* −0.136 (−0.09) (−1.83) (−1.19) PV of projected efficiency gains 0.053* 0.158* 0.235* (1.70) (1.90) (1.70) Number of rivals −0.001 −0.006 −0.009 (−1.17) (−1.16) (−1.36) Implied change in HHI −0.052*** −0.220** −0.269 (−2.70) (−2.04) (−1.31) Bidder–target product similarity 0.044* 0.323** 0.633*** (1.93) (2.52) (2.60) Bidder–target technology similarity −0.017*** −0.046 −0.044 (−3.04) (−1.46) (−0.74) Same BEA 0.006* −0.002 −0.019 (1.74) (−0.09) (−0.50) Deregulation −0.004 −0.020 −0.022 (−0.89) (−0.71) (−0.42) Number of competing bidders −0.001 0.016 0.039 (−0.23) (0.85) (1.06) Target sales/bidder sales −0.005** −0.016 −0.029 (−2.40) (−1.42) (−1.32) Log(supplier sales) 0.002** 0.014*** 0.017* (2.26) (2.86) (1.95) Supplier sales/bidder sales 0.000 −0.001 0.000 (−0.54) (−0.22) (0.02) Supplier sales/target sales 0.000 −0.001 −0.003 (−0.88) (−1.20) (−1.28) Supplier Tobin’s Q −0.001** −0.003 −0.004 (−2.40) (−1.14) (−0.76) Supplier leverage 0.008 0.048 0.075 (0.99) (1.10) (0.92) Supplier profitability 0.000 0.000 −0.001 (−0.19) (0.07) (−0.11) Bidder’s supplier −0.001 0.016 0.034 (−0.18) (0.45) (0.51) Target’s supplier −0.002 0.024 0.057 (−0.28) (0.75) (0.94) Inverse Mills ratio 0.000 0.035 0.087 (0.00) (1.22) (1.58) Number of observation 1,523 1,523 1,523 Adjusted R-squared 0.050 0.029 0.015 The relation between efficiency gain projections and suppliers’ announcement returns is significantly positive at 10% level for all three announcement windows. The economic significance of this relation is substantially larger than that for customers: a one-standard-deviation increase in projected efficiency gains is associated with an increase of 0.6% in suppliers’ announcement return over [−1, 1] window, an increase of 1.8% over [−10, 10] window, and an increase of 2.7% over [−20, 20] window. Since suppliers in our sample are significantly smaller than customers, the larger announcement returns of suppliers are consistent with mergers having a larger impact on smaller firms along the merging firms’ supply chains. Similar to the case of customers, the market does not incorporate the information in insiders’ projections immediately into suppliers’ share prices. Suppliers’ returns are decreasing in the implied change in merging firms’ industry HHI, suggesting that larger increases in downstream industries’ market power are detrimental to suppliers. However, higher bidder-target product similarity is positively associated with suppliers’ returns. This is potentially inconsistent with the market power channel, since bidder-target product similarity may be positively related to merging firms’ ability to exercise market power. Similar to the case of customers, larger suppliers seem to react more positively to merger announcements than smaller ones. This is in line with Bhattacharyya and Nain (2011), who show that suppliers with larger bargaining power are hurt less by consolidation in downstream industry. Insignificant coefficients on the inverse Mills ratio suggest that self-selection does not seem to drive the relation between projected operating efficiency gains and suppliers’ announcement returns. Split-sample regressions, similar to those in Tables VI and VIII, are reported in Table X. Table X. Suppliers’ returns and projected operating efficiency gains—subsample analysis This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s supplier’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20], and which are estimated for suppliers within various subsamples of mergers. In Panel A, the high (low) subsample contains suppliers of bidders and/or targets in mergers in which bidder’s capital-stock-to-value ratio exceeds (falls short of) that of all bidders in M&As announced in the same year. In Panel B, the high (low) subsample contains suppliers of bidders and/or targets in mergers in which target’s capital-stock-to-value ratio exceeds (falls short of) that of all targets in M&As announced in the same year. In Panel C, the high (low) subsample contains suppliers in mergers in which the capitalization by the market of projected efficiency gains, computed as the ratio of bidder’s and target’s combined CAR over the corresponding announcement window and the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization, is above (below) median for all M&As announced in the same year. In Panel D, the high (low) subsample contains suppliers in mergers in which the ratio of the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization and the implied change in the merging firms’ industry HHI, is above (below) median for all M&As announced in the same year. In Panel E, the high (low) subsample contains suppliers in mergers in which target’s book assets is above (below) median for all M&As announced in the same year. The main independent variable is the value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I for the definitions of the rest of the variables. The regressions are estimated with year-fixed effects. Standard errors are clustered by merger and reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. We only report the coefficients and associated standard errors on the estimated value of projected efficiency gains. The difference column represents the difference in coefficient estimates between two subsamples. Its significance is computed using Wald test. Dependent variable Supplier CAR [−1, 1] Supplier CAR [−10, 10] Supplier CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.156*** 0.007 0.149** 0.532* −0.076 0.608* 0.757** −0.142 0.899 PV of projected efficiency gains (3.03) (0.17) (2.26) (1.81) (−0.42) (1.76) (2.39) (−0.30) (1.58) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.080 0.023 0.057 0.237 −0.019 0.256 0.481 −0.126 0.607 PV of projected efficiency gains (1.41) (0.57) (0.82) (1.03) (−0.06) (0.65) (1.10) (−0.21) (0.82) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.021 0.049* −0.029 0.178* −0.054 0.232** 0.363* −0.216 0.579** PV of projected efficiency gains (0.62) (1.69) (−0.65) (1.90) (−1.14) (2.21) (1.95) (−1.30) (2.32) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.035 0.040 −0.005 0.155** 0.011 0.144* 0.400** 0.018 0.382** PV of projected efficiency gains (0.89) (1.15) (−0.10) (1.99) (0.29) (−1.66) (2.11) (0.16) (−1.99) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.041 0.044 −0.003 0.110 0.099 0.011 0.187 0.291 −0.104 PV of projected efficiency gains (1.59) (1.10) (−0.06) (1.48) (1.03) (0.09) (1.38) (1.60) (−0.46) Dependent variable Supplier CAR [−1, 1] Supplier CAR [−10, 10] Supplier CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.156*** 0.007 0.149** 0.532* −0.076 0.608* 0.757** −0.142 0.899 PV of projected efficiency gains (3.03) (0.17) (2.26) (1.81) (−0.42) (1.76) (2.39) (−0.30) (1.58) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.080 0.023 0.057 0.237 −0.019 0.256 0.481 −0.126 0.607 PV of projected efficiency gains (1.41) (0.57) (0.82) (1.03) (−0.06) (0.65) (1.10) (−0.21) (0.82) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.021 0.049* −0.029 0.178* −0.054 0.232** 0.363* −0.216 0.579** PV of projected efficiency gains (0.62) (1.69) (−0.65) (1.90) (−1.14) (2.21) (1.95) (−1.30) (2.32) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.035 0.040 −0.005 0.155** 0.011 0.144* 0.400** 0.018 0.382** PV of projected efficiency gains (0.89) (1.15) (−0.10) (1.99) (0.29) (−1.66) (2.11) (0.16) (−1.99) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.041 0.044 −0.003 0.110 0.099 0.011 0.187 0.291 −0.104 PV of projected efficiency gains (1.59) (1.10) (−0.06) (1.48) (1.03) (0.09) (1.38) (1.60) (−0.46) Table X. Suppliers’ returns and projected operating efficiency gains—subsample analysis This table presents estimates of OLS regressions in which the dependent variable takes the form of bidder’s and/or target’s supplier’s CAR [−1, 1], CAR [−10, 10], or CAR [−20, 20], and which are estimated for suppliers within various subsamples of mergers. In Panel A, the high (low) subsample contains suppliers of bidders and/or targets in mergers in which bidder’s capital-stock-to-value ratio exceeds (falls short of) that of all bidders in M&As announced in the same year. In Panel B, the high (low) subsample contains suppliers of bidders and/or targets in mergers in which target’s capital-stock-to-value ratio exceeds (falls short of) that of all targets in M&As announced in the same year. In Panel C, the high (low) subsample contains suppliers in mergers in which the capitalization by the market of projected efficiency gains, computed as the ratio of bidder’s and target’s combined CAR over the corresponding announcement window and the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization, is above (below) median for all M&As announced in the same year. In Panel D, the high (low) subsample contains suppliers in mergers in which the ratio of the value of projected efficiency gains scaled by bidder’s and target’s combined market capitalization and the implied change in the merging firms’ industry HHI, is above (below) median for all M&As announced in the same year. In Panel E, the high (low) subsample contains suppliers in mergers in which target’s book assets is above (below) median for all M&As announced in the same year. The main independent variable is the value of projected efficiency gains estimated using 10-year horizon and bidder’s and target’s combined WACC, as described in Panel B of Table I. See Panels C–E of Table I for the definitions of the rest of the variables. The regressions are estimated with year-fixed effects. Standard errors are clustered by merger and reported in parentheses. ***, **, and * denote statistical significance at 1%, 5%, and 10% levels, respectively. We only report the coefficients and associated standard errors on the estimated value of projected efficiency gains. The difference column represents the difference in coefficient estimates between two subsamples. Its significance is computed using Wald test. Dependent variable Supplier CAR [−1, 1] Supplier CAR [−10, 10] Supplier CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.156*** 0.007 0.149** 0.532* −0.076 0.608* 0.757** −0.142 0.899 PV of projected efficiency gains (3.03) (0.17) (2.26) (1.81) (−0.42) (1.76) (2.39) (−0.30) (1.58) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.080 0.023 0.057 0.237 −0.019 0.256 0.481 −0.126 0.607 PV of projected efficiency gains (1.41) (0.57) (0.82) (1.03) (−0.06) (0.65) (1.10) (−0.21) (0.82) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.021 0.049* −0.029 0.178* −0.054 0.232** 0.363* −0.216 0.579** PV of projected efficiency gains (0.62) (1.69) (−0.65) (1.90) (−1.14) (2.21) (1.95) (−1.30) (2.32) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.035 0.040 −0.005 0.155** 0.011 0.144* 0.400** 0.018 0.382** PV of projected efficiency gains (0.89) (1.15) (−0.10) (1.99) (0.29) (−1.66) (2.11) (0.16) (−1.99) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.041 0.044 −0.003 0.110 0.099 0.011 0.187 0.291 −0.104 PV of projected efficiency gains (1.59) (1.10) (−0.06) (1.48) (1.03) (0.09) (1.38) (1.60) (−0.46) Dependent variable Supplier CAR [−1, 1] Supplier CAR [−10, 10] Supplier CAR [−20, 20] Panel A: Bidder capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.156*** 0.007 0.149** 0.532* −0.076 0.608* 0.757** −0.142 0.899 PV of projected efficiency gains (3.03) (0.17) (2.26) (1.81) (−0.42) (1.76) (2.39) (−0.30) (1.58) Panel B: Target capital stock-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.080 0.023 0.057 0.237 −0.019 0.256 0.481 −0.126 0.607 PV of projected efficiency gains (1.41) (0.57) (0.82) (1.03) (−0.06) (0.65) (1.10) (−0.21) (0.82) Panel C: % capitalization of efficiency gains-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.021 0.049* −0.029 0.178* −0.054 0.232** 0.363* −0.216 0.579** PV of projected efficiency gains (0.62) (1.69) (−0.65) (1.90) (−1.14) (2.21) (1.95) (−1.30) (2.32) Panel D: Ratio of efficiency gains estimate and implied change in HHI-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.035 0.040 −0.005 0.155** 0.011 0.144* 0.400** 0.018 0.382** PV of projected efficiency gains (0.89) (1.15) (−0.10) (1.99) (0.29) (−1.66) (2.11) (0.16) (−1.99) Panel E: Target size-based subsamples High Low Difference High Low Difference High Low Difference Coefficient on 0.041 0.044 −0.003 0.110 0.099 0.011 0.187 0.291 −0.104 PV of projected efficiency gains (1.59) (1.10) (−0.06) (1.48) (1.03) (0.09) (1.38) (1.60) (−0.46) The estimates of these regressions reveal that the relation between projected efficiency gains and suppliers’ announcement returns is more positive for mergers involving more capital-intensive bidders. The differences in the relation between projected gains and suppliers’ announcement returns between the two bidder-capital-intensity-based subsamples are statistically significant at least at 10% level for two announcement windows out of three. In addition, the relation is statistically significant for all announcement windows in the capital-intensive-bidder subsample and it is insignificant in the low-capital-intensity subsample for all announcement windows. The results are qualitatively similar for subsamples constructed based on targets’ capital-stock-to-assets ratios, but they are statistically insignificant. The relation between efficiency gain forecasts and suppliers’ announcement returns is significantly (at 10% level) stronger for mergers in which relatively high proportion of projected gains are capitalized than for mergers with lower-than-median capitalization of projections for two announcement windows out of three. This relation also tends to be stronger within the subsample of mergers with above-median ratio of efficiency gain projections to implied change in industry HHI. Splitting the sample by target size does not reveal significant differences in the effect of operating efficiency gains on suppliers between the two subsamples. This result is consistent with self-selection of efficiency gains disclosure having no material effect on suppliers’ announcement returns, as follows from the insignificant coefficient on the inverse Mills ratio in Table IX. Overall, the results in Table X suggest that, similar to the cases of rivals and customers, the impact of projected operating efficiency gains on suppliers depends on the degree to which the market believes that projected forecast would materialize and on the perceived importance of the operating efficiency effect relative to the market power effect. 6. Conclusions This paper is the first to examine directly the effects of operating efficiencies in horizontal mergers on firms that interact with acquirers and targets in product, input, and output markets. Our tests exploit a unique dataset of managerial forecasts disclosed during merger announcements, which reflect expected gains from merger-related operating efficiencies. We find that the relations between projected operating efficiencies on the one hand and announcement returns of related firms on the other hand are consistent with predictions of industrial organization theory. First, operating efficiencies are negatively related to announcement returns of merging firms’ rivals. Second, operating efficiencies have a positive effect on values of merging firms’ customers. Third, operating efficiencies are positively related to announcement returns of merging firms’ suppliers. These relations tend to be more pronounced in larger mergers and in mergers in which managerial projections of operating efficiency gains are more credible, i.e., in cases of more capital-intensive bidders and targets, mergers in which the market capitalizes a larger fraction of insiders’ projections, and those in which the ratio of projected efficiency gains to implied change in market power is larger. Our paper contributes to a growing literature that demonstrates the importance of product market linkages for firms’ performance and valuation. In particular, our empirical results suggest that the pass-through of efficiency gains along merging firms’ supply chains is as important as the effects of post-merger increase in market power in the merging firms’ industry. Supplementary Material Supplementary data are available at Review of Finance online. Footnotes We are grateful to Alex Edmans (the editor), Gerard Hoberg and David Robinson (the referees), Kenneth Ahern, Scott Bauguess, Nittai Bergman, Gregg Jarrell, Simon Kwan, Andres Liberman, Ernst Maug, Roni Michaely, Gordon Phillips, Frode Steen, Alminas Zaldokas, and seminar participants at Boston University, Tel Aviv University, University of Miami, 2011 Rothschild Caesarea Center Conference, 2012 American Finance Association Meetings, 2012 European Association for Research in Industrial Economics Meetings, 2013 Financial Intermediation Research Society Meetings, 2014 Western Finance Association Meetings, 2014 China International Conference in Finance, and 2014 European Finance Association Meetings for helpful comments and suggestions. We thank Jerry Hoberg and Gordon Phillips for sharing data on text-based industry classifications, and Lauren Cohen for sharing data on corporate customers linkages. We are responsible for any errors. 1 Our sample size generally compares favorably to other studies that employ managerial projections: the sample in Gilson, Hotchkiss, and Ruback (2000), who examine projections in the context of bankruptcy filings, contains sixty-three firms, and that in Houston, James, and Ryngaert (2001) contains forty-one bank mergers. Bernile and Bauguess’ (2014) sample of 894 mergers during the same sample period as ours includes horizontal as well as vertical and diversifying mergers. 2 The sources include: Business Wire: Dow Jones Business News; Dow Jones News Service; Mergers & Acquisitions Report; Mergers & Acquisitions: The Dealmakers Journal; Mergers & Acquisitions Litigation Reporter; PR Newswire; Reuters News; Seeking Acquisitions: FirstList; The Wall Street Journal. 3 If managers forecast annual gains of x dollars by year t, where t > 1 (or provide no timing, in which case we assume t = 4), we assume that the merged entity realizes x/2 in year t – 1, x/4 in year t – 2, and so forth. If managers forecast annual gains of x dollars by year t and y dollars by year t + i, where i > 1, we interpolate the expected gains for the intermediate years, assuming that the gains increase linearly over the missing forecast years. When a cumulative forecast is provided, we adopt the following convention. For a 3-year cumulative forecast equal to x, we assume the first year forecast equals x/7, the second 2x/7, and the third 4x/7. For a 5-year cumulative forecast that equals x, we assume the first year forecast equals x/23, the second 2x/23, the third 4x/23, and 8x/23 in each remaining year. For a 10-year cumulative forecast equal to x, we assume the first-year forecast equals x/63, the second 2x/63, the third 4x/63, and 8x/63 in each remaining year. 4 To calculate each merging firm’s cost of unlevered equity, we (1) estimate the firm’s equity beta, βE, using at least 60 and at most 250 daily firm stock returns and the CRSP value-weighted portfolio returns ending 60 days prior to when the target is “put in play”, (2) assume that the merging firms’ debt and preferred equity betas, βD and βP, are equal to 0.25, (3) calculate unlevered equity beta as βU=[(βEE+βPP+0.64βDD)/(E+P+0.64D)], (4) apply the CAPM equation to the unlevered equity beta, βU, assuming a 7.5% risk-premium, which is reasonable during our sample period, and setting the risk-free rate equal to the 10-year Treasury bond yield at the time of the merger announcement. 5 A similarity of zero means that there are no overlapping words in the two firms’ product descriptions, other than designated common words. A similarity of one means that the sets of words appearing in the two firms’ product descriptions are identical bar these common words. Importantly, this measure is purged of vertical relations using the Bureau of Economic Analysis input–output tables. 6 For each firm i in year t, we construct a vector Ci,t = [Ci,t,1,…,Ci,t,k,…,Ci,t,K], where k=1,…,K=36 is the number of two-digit technological subcategories, and Ci,t,k is the proportion of citations to firm i’s patents awarded in class k in year t out of all citations to firm i’s patents awarded in year t. To measure the number of citations to each patent, we consider only patents granted in year t that have an application year that precedes the granting year by at most 3 years. Then, for each patent we evaluate the total number of citations that the patent received within 3 years from the granting year. We calculate technological similarity of firm-pair i,j in year t as γi,j,t=∑k=1Kmin⁡[Ci,t,k,Cj,t,k]max⁡[∑k=1KCi,t,k,∑k=1KCj,t,k]. γi,j,t is bounded between 0 and 1. If γi,j,t equals 1, then firms i and j have the exact same proportions of citations to patents in each of thirty-six two-digit technological categories. If γi,j,t equals 0, then the two firms do not share citations to patents assigned to any two-digit technological category. If at least one of the merging firms does not have any citations to its patents (or no patents), then γi,j,t is set to 0. 7 A firm’s capital stock at the end of year t, which is defined as the first year the firm appears in Compustat, is its gross PP&E depreciated using three-digit SIC industry-wide depreciation rate in year t. Its capital stock at the end of any year τ>t is the sum of the capital stock in year τ−1 and the capital investment in year τ, computed as the difference between gross capital investment in year τ and that in year τ−1, and depreciated using industry-average depreciation rate in year τ. 8 We use the specification in Column 1 of Table II, in which the only instruments are bidder’s and target’s number of earnings guidances, to compute the inverse Mills ratio. The results are robust to computing them using the specification in Column 2 of Table II, which includes additional instruments. 9 The detailed description of the matching algorithm is available in Cohen and Frazzini (2008). 10 Positive value-weighted mean returns are consistent with prior studies (e.g., Eckbo, 1983; Fee and Thomas, 2004; Shahrur, 2005). 11 The mean ratio of scaled projected efficiency gains and merging firms’ combined announcement returns is around 45% for all three announcement windows. Appendix: Examples of News Stories and Press Releases Containing Insiders’ Operating Efficiency Forecasts In this Appendix, we include a number of extracts from news stories and press releases that contain managerial forecasts of merger-related gains. Example 1. Hilton to buy Promus Hotel for $4 billion. 7 September 1999, Reuters News: The [merger] will result in annual cost savings and operating efficiencies of about $55 million in the first year and $90 million thereafter. Example 2. Kroger to merge with Fred Meyer. 19 October 1998, Reuters News: Kroger plans to generate annual cost savings of approximately $225 million within 3 years, including approximately $75 million in the first year. Kroger plans to generate these savings through combined procurement of goods and services, reduced corporate overhead, in-market synergies, and consolidation of support services. Example 3. Amax Inc., Cyprus Minerals merger to create $5 billion company. 25 May 1993, Reuters News: The combination of the companies will present significant opportunities to reduce operating and corporate and divisional overhead costs, with anticipated annual cost savings of at least $100 million. Example 4. Bell Atlantic, GTE Merger—4 bn cost synergies within 3 years. 28 July 1998, Dow Jones News Service: The companies said in a joint press release Tuesday that they see the transaction producing cost synergies totaling $2 billion within 3 years of the deal’s completion. The merged company is also expected to generate an additional $2 billion in revenue synergies. Example 5. Food lion will buy hannaford for about $3.3 billion, plus debt. 19 August 1999, Dow Jones Business News: The combined Food Lion and Hannaford will have nearly $14 billion in pro-forma annual revenue. The combined company is expected to result in synergies estimated at about $40 million in the first year and about $75 million annually by the third year. References Ahern K. ( 2012 ): Bargaining power and industry dependence in mergers , Journal of Financial Economics 103 , 530 – 550 . Google Scholar CrossRef Search ADS Ahern K. , Harford J. ( 2014 ): The importance of industry links in merger waves , Journal of Finance 69 , 527 – 576 . Google Scholar CrossRef Search ADS Bebchuk L. , Cohen A. , Ferrell A. ( 2009 ): What matters in corporate governance? , Review of Financial Studies 22 , 783 – 827 . Google Scholar CrossRef Search ADS Bena J. , Li K. ( 2014 ): Corporate innovations and mergers and acquisitions , Journal of Finance 69 , 1923 – 1960 . Google Scholar CrossRef Search ADS Bernile G. , Bauguess S. ( 2014 ): Do merger-related operating synergies exist. Working paper, University of Miami. Bhattacharyya S. , Nain A. ( 2011 ): Horizontal acquisitions and buying power: a product market analysis , Journal of Financial Economics 99 , 97 – 115 . Google Scholar CrossRef Search ADS Cai J. , Song M. , Walkling R. ( 2011 ): Anticipation, acquisitions, and bidder returns: industry shocks and the transfer of information across rivals , Review of Financial Studies 24 , 2242 – 2285 . Google Scholar CrossRef Search ADS Cohen L. , Frazzini A. ( 2008 ): Economic links and predictable returns , Journal of Finance 63 , 1977 – 2011 . Google Scholar CrossRef Search ADS Devos E. , Kadapakkam P. , Krishnamurthy S. ( 2009 ): How do mergers create value? A comparison of taxes, market power, and efficiency improvements as explanations for synergies , Review of Financial Studies 22 , 1179 – 1211 . Google Scholar CrossRef Search ADS Eckbo E. ( 1983 ): Horizontal mergers, collusion, and stockholder wealth , Journal of Financial Economics 11 , 241 – 273 . Google Scholar CrossRef Search ADS Eckbo E. ( 1985 ): Mergers and the market concentration doctrine: evidence from the capital market , Journal of Business 58 , 325 – 349 . Google Scholar CrossRef Search ADS Eckbo E. ( 2014 ): Corporate takeovers and economic efficiency , Annual Review of Financial Economics 6 , 51 – 74 . Google Scholar CrossRef Search ADS Fan J. , Goyal V. ( 2006 ): On the patterns and wealth effects of vertical mergers , Journal of Business 79 , 877 – 902 . Google Scholar CrossRef Search ADS Fee E. , Thomas S. ( 2004 ): Sources of gains in horizontal mergers: evidence from customer, supplier, and rival firms , Journal of Financial Economics 74 , 423 – 460 . Google Scholar CrossRef Search ADS Gilson S. , Hotchkiss E. , Ruback R. ( 2000 ): Valuation of bankrupt firms , Review of Financial Studies 13 , 43 – 74 . Google Scholar CrossRef Search ADS Gorton G. , Kahl M. , Rosen R. ( 2009 ): Eat or be eaten: a theory of mergers and firm size , Journal of Finance 64 , 1291 – 1344 . Google Scholar CrossRef Search ADS Hall B. , Jaffe A. , Trajtenberg M. ( 2005 ): Market value and patent citations , Rand Journal of Economics 1 , 16 – 38 . Harford J. ( 2005 ): What drives merger waves? , Journal of Financial Economics 77 , 529 – 560 . Google Scholar CrossRef Search ADS Heckman J. ( 1979 ): Sample selection bias as a specification error , Econometrica 47 , 153 – 161 . Google Scholar CrossRef Search ADS Heckman J. , Navarro-Lozano S. ( 2004 ): Using matching, instrumental variables, and control functions to estimate economic choice models , Review of Economics and Statistics 86 , 30 – 57 . Google Scholar CrossRef Search ADS Hertzel M. , Officer M. , Rodgers K. ( 2008 ): Inter-firm linkages and the wealth effects of financial distress along the supply chain , Journal of Financial Economics 87 , 374 – 387 . Google Scholar CrossRef Search ADS Hirshleifer D. , Hsu P. H. , Li D. ( 2013 ): Innovative efficiency and stock returns , Journal of Financial Economics 107 , 632 – 654 . Google Scholar CrossRef Search ADS Hoberg G. , Phillips G. ( 2010a ): Real and financial industry booms and busts , Journal of Finance 65 , 45 – 86 . Google Scholar CrossRef Search ADS Hoberg G. , Phillips G. ( 2010b ): Product market synergies and competition in mergers and acquisitions: a text based analysis , Review of Financial Studies 23 , 3773 – 3811 . Google Scholar CrossRef Search ADS Houston J. , James C. , Ryngaert M. ( 2001 ): Where do merger gains come from? Bank mergers from the perspective of insiders and outsiders , Journal of Financial Economics 60 , 285 – 331 . Google Scholar CrossRef Search ADS Jaffe A. ( 1986 ): Technological opportunity and spillovers of R&D: evidence from firms’ patents, profits, and market value , American Economic Review 76 , 984 – 1001 . Jovanovic B. , Rousseau P. ( 2002 ): The Q-theory of mergers , American Economic Review 92 , 198 – 204 . Google Scholar CrossRef Search ADS Kaplan S. ( 2000 ): Introduction, in: Kaplan S. (ed.), Mergers and Productivity , University of Chicago Press/NBER , Chicago , pp. 1 – 8 . Google Scholar CrossRef Search ADS Kaplan S. , Ruback R. ( 1995 ): The valuation of cash flow forecasts: an empirical analysis , Journal of Finance 50 , 1059 – 1093 . Google Scholar CrossRef Search ADS Lambrecht B. ( 2004 ): The timing and terms of takeovers motivated by economies of scale , Journal of Financial Economics 72 , 41 – 62 . Google Scholar CrossRef Search ADS Li K. , Prabhala N. ( 2005 ) Self-selection models in corporate finance, Chapter 2, in: Eckbo E. (ed.), Handbook of Corporate Finance: Empirical Corporate Finance , Elsevier/North-Holland , New York , pp. 37 – 86 . Maksimovic V. , Phillips G. ( 2001 ): The market for corporate assets: who engages in mergers and asset sales and are there efficiency gains? , Journal of Finance 56 , 2019 – 2065 . Morck R. , Shleifer A. , Vishny R. ( 1990 ): Do managerial objectives drive bad acquisitions? , Journal of Finance 45 , 31 – 48 . Google Scholar CrossRef Search ADS Otchere I. , Ip E. ( 2006 ): Intra-industry effects of completed and cancelled cross border acquisitions in Australia: a test of the acquisition probability hypothesis , Pacific-Basin Finance Journal 14 , 209 – 230 . Google Scholar CrossRef Search ADS Perry M. , Porter R. ( 1985 ): Oligopoly and the incentive for horizontal merger , American Economic Review 75 , 219 – 227 . Rhodes-Kropf M. , Robinson D. ( 2008 ): The market for mergers and the boundaries of the firm , Journal of Finance 63 , 1169 – 1211 . Google Scholar CrossRef Search ADS Rhodes-Kropf M. , Viswanathan S. ( 2004 ): Market valuation and merger waves , Journal of Finance 59 , 2685 – 2718 . Google Scholar CrossRef Search ADS Roll R. ( 1986 ): The hubris hypothesis of corporate takeovers , Journal of Business 59 , 437 – 467 . Salant S. , Switzer S. , Reynolds R. ( 1983 ): Losses from horizontal merger: the effects of an exogenous change in industry structure on Cournot–Nash equilibrium , Quarterly Journal of Economics 98 , 185 – 199 . Google Scholar CrossRef Search ADS Shahrur H. ( 2005 ): Industry structure and horizontal takeovers: analysis of wealth effects on rivals, suppliers, and corporate customers , Journal of Financial Economics 76 , 61 – 98 . Google Scholar CrossRef Search ADS Shleifer A. , Vishny R. ( 2003 ): Stock market driven acquisitions , Journal of Financial Economics 70 , 295 – 311 . Google Scholar CrossRef Search ADS Song M. , Walkling R. ( 2000 ): Abnormal returns to rivals of acquisition targets: a test of the “acquisition probability hypothesis” , Journal of Financial Economics 55 , 143 – 171 . Google Scholar CrossRef Search ADS Stigler G. ( 1964 ): A theory of oligopoly , Journal of Political Economy 72 , 44 – 61 . Google Scholar CrossRef Search ADS Stillman R. ( 1983 ): Examining antitrust policy towards horizontal mergers , Journal of Financial Economics 11 , 225 – 240 . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of the European Finance Association. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Review of Finance Oxford University Press

The Effects of Horizontal Merger Operating Efficiencies on Rivals, Customers, and Suppliers