Acquisitions of family owned firms: boon or bust?

Transcription

1 J Econ Finan (2014) 38: DOI /s Acquisitions of family owned firms: boon or bust? Kimberly C. Gleason & Anita K. Pennathur & Joan Wiggenhorn Published online: 26 November 2011 # Springer Science+Business Media, LLC 2011 Abstract Using a hand-collected dataset, we examine the acquisitions of 307 family owned targets for the period to study the conflicting roles of entrenchment and alignment of interests in these firms. We find that bidders experience the strongest announcement market reaction at the medium levels of family ownership. However, bidder returns are negatively impacted when acquiring a public target family firm, even controlling for the percent owned by the family. We also find that overall firms that acquire family owned targets experience significant large negative returns in the long run. However, the long run results are also sensitive to the ownership structure of the family firm target with medium levels resulting in insignificant returns in the long run. Keywords Ownership Structure. Mergers. Acquisitions. Family Firms 1 Introduction The family owned firm has historically been the most prevalent form of enterprise in the United States, comprising, according to the Family Business Forum (Accessed at 78% K. C. Gleason Department of Finance, University of Pittsburgh, Mervis Hall, Pittsburg, PA 15260, USA gleasonk@pitt.edu A. K. Pennathur Department of Finance, Florida Atlantic University, LA 475, 2912 College Avenue, Davie, FL 33314, USA pennathu@fau.edu J. Wiggenhorn (*) College of Business, Florida Institute of Technology, 150 W. University Boulevard, Melbourne, FL 32901, USA jwiggenhorn@fit.edu

2 270 J Econ Finan (2014) 38: of all new jobs created and 90% of all businesses in the United States. A study by the American Management Association finds that fewer than 30% of family businesses survive into the second generation, 10% to the third and only 4% operate at the fourth generation and above (Prachyakom 2010). Accordingly, we focus on the sale of 307 family businesses by either the founder or by subsequent generations and investigate the returns to the bidders when these family-owned firms are acquired. Our unique, hand-collected dataset allows an in-depth insight into the characteristics of the deal, such as the age of the family firm, ownership levels, method of payment, and generational distance between the founder and current family-manager. The literature reveals opposing conclusions pertaining to ownership structure and firm value. According to traditional agency theory, an alignment of interests should exist in a family owned business, which should presumably lead to superior firm performance (Jensen and Meckling 1976; Fama and Jensen 1983; Stulz 1988). Other research focuses on entrenchment issues (McConnell and Servaes 1990; Slovin and Sushka 1993), and still others such as Chang (1998) and Fuller et al. (2002) focus on returns to shareholders when private firms are acquired. We extend the literature by focusing on the acquirers of family firms, both public and privately held. In the case of alignment of interests, any change in family interest via a merger could negatively impact firm value. Alternately, a family owned firm could have an entrenched family that is detrimental to firm value. In these cases, any change in ownership structure should alleviate this entrenchment problem, resulting in a higher firm value. While earlier studies generally show that bidders experience negative, or at best, insignificant gains (Jensen and Ruback 1983; Weston et al. 2001; Mulherin and Boone 2000), we find that bidders do have positive announcement abnormal returns from the acquisition at the high and medium levels of family ownership. However, bidder returns are negatively impacted when acquiring a public target family firm, even controlling for the percent owned by the family. Thus, bidder gains can be explained for the high levels of family ownership where not many of the targets are publicly traded and where information asymmetry and entrenchment are likely to be very high. However, using the Fama-French three factor model, long run returns are significantly and largely negative, suggesting that the absorption of family firms is not easily accomplished. The remainder of the paper is organized as follows. We review the related literature and form hypotheses in Section 2 while Section 3 describes our data and methodology. Section 4 reports our results and interpretations, and Section 5 summarizes and concludes the paper. 2 Review of the literature and hypotheses 2.1 Acquisition announcement returns The majority of the literature finds negative or at best insignificant abnormal announcement returns for acquirers (Roll 1986; Jensen 1986; Morck et al. 1990). For studies dealing with our time period, Weston et al. (2001), Schwert (1996), Mulherin

3 J Econ Finan (2014) 38: and Boone (2000) and Swanstrom (2006) all find negative returns. On the other hand, there has been some evidence of positive announcement returns for acquiring firms. Chang (1998) argues that competition for private targets may be limited due to the difficulty of obtaining information on these targets, and correspondingly high search costs. Fuller et al. (2002) call this the liquidity effect, where private firms cannot be bought or sold as effortlessly as public firms. A similar argument could be made for family owned firms. 2.2 Family firm valuation In analyzing the returns to bidders of family firms, it is first important to understand how family firm valuation varies with the level of family ownership. There are three significant competing literature strands. The alignment theme predicates a positive relation between firm value and insider ownership (Jensen and Meckling 1976; Fama and Jensen 1983; DeAngelo and DeAngelo 1985). This agency argument is that the value of family owned firms should be high due to the alignment of ownermanager interests with that of the firm. Another stream of literature, the entrenchment hypothesis, asserts that while ownership does provide some agency resolution, concentration of such ownership will allow owner-managers (families) to become entrenched and reluctant to relinquish power or to allow change (Stulz 1988; Slovin and Sushka 1993). Further evidence of the entrenchment problem is provided by Yermack (1996), who examines the connection between board composition and firm value, proxied by Tobin s Q, to report that the presence of a founding family CEO negatively impacts firm value. The third possibility, which we find more compelling, predicts a curvilinear relationship between ownership concentration and firm value. Both Morck et al. (1988) and McConnell and Servaes (1990) find that Tobin s Q, first increases and then decreases, with an increase in ownership. Also similar to Stulz (1988), McConnell and Servaes (1990) find that this association reaches a maximum prior to a 50% ownership. Further evidence supporting a curvilinear valuation is found by Anderson et al. (2003), Anderson and Reeb (2003), Maury (2006), and also Claessens et al. (2002). Ali et al. (2007) find that the traditional agency problem is reduced by family ownership, but there arises a conflict between the controlling and non-controlling shareholders. From these studies, an important conclusion, therefore, is that there exists an optimal inside voting right percentage that maximizes the value of the firm, and this becomes the basis of our first hypothesis. 2.3 Family firm characteristics Since some studies find that returns vary according to characteristics of the target firm (Chang 1998; Gleason et al. 2005), we examine other family factors beside the percentage ownership. We first investigate whether generational distance from the founder affects the abnormal returns to the bidder and target. Superior performance is found when the family firm is still controlled by the founder (Barontini and Caprio 2006; Miller et al. 2007; Villalonga and Amit 2006). A higher agency cost for subsequent generations with decreasing firm valuations is found by Blanco-Mazagatos et al. (2007). Bennedsen et al. (2007) find that family successions have a large

4 272 J Econ Finan (2014) 38: negative causal impact on firm performance. Thus, family firms still controlled by the founder should result in higher returns. However, we must also consider whether family ownership is simply a proxy for a closely held firm. Fuller et al. (2002) find that bidder returns are positive when the target is private, but not when the target is public. Chang (1998) similarly finds positive returns only when the target is private and when purchased with stock. Faccio et al. (2006) find that the five day cumulative abnormal return for acquirers of listed targets is an insignificant 0.38% while acquirers of unlisted targets earn a significant abnormal return of 1.48%. Similar results are found by Capron and Shen (2007). Therefore, we investigate the returns to bidder shareholders by whether the family owned targets are publicly or privately owned. Returns to bidders of public family owned firms should be lower than those of privately held firms, given that the returns for publicly held family owned firms should reflect lower search costs, lower information asymmetry, and a more efficient asset market than those of private family owned firms. 2.4 Method of payment Further, the method of payment can also provide a signal about the amount of information asymmetry. This is validated by empirical research reporting a negative market reaction to stock acquisitions and a positive reaction to cash payments (Travlos 1987; Brown and Ryngaert 1991; Swanstrom 2006). However, Chang (1998) finds that private firms that are acquired by stock elicit a positive return for the acquirer. One explanation he provides is the formation of a blockholder in the acquiring firm. Since all of our firms have family ownership of at least 20%, then all target firms have the potential to be vigilant blockholders. 2.5 Long run performance The last question to be answered is if acquisition activity affects a bidder s financial performance in the long run. Franks et al. (1991) examine 400 mergers from 1975 through 1984 and find the abnormal announcement returns are negative and insignificant, except for all equity bids that have significant negative abnormal returns. Agrawal et al. (1992) find that the cumulative abnormal returns are a statistically significant 4.94% at the end of the second year and 10.26% after 5 years. Loderer and Martin (1992) find support for negative abnormal returns in the second and third year after an acquisition. On the other hand, Dube and Glascock (2006) find no long run risk adjusted changes in the stock performance of acquiring firms. In addition, various factors have been found to affect the long term stock performance. Loughran and Vijh (1997) find that for the 5 year period following the acquisition, firms that complete stock mergers earn a significant 25%, while firms that used complete cash offers earned a significant 61.7%. However, in the work most closely related to ours, Chang (1998) never investigates the long run returns of acquirers of privately owned firms. Since the majority of our targets are privately held, it is possible that with a new blockholder, the long run returns will be positive.

5 J Econ Finan (2014) 38: Formal hypotheses H1: Acquirer announcement returns will be curvilinear dependent on the level of family ownership of the target. H2: Acquirer announcement returns will be highest when the family firm is still controlled by the founder. H3: Acquirer announcement returns will be higher for privately held targets than for public targets. H4: Acquirer announcement returns will be higher for acquisitions using stock rather than cash. H5: Acquirer long run returns will be positive. 3 Data and methodology 3.1 Data We obtain announcements of acquisitions of family owned targets from the Securities Data Corporation (SDC) International Mergers and Acquisitions database. Specifically, we search for announcements of acquisitions of family owned targets by publicly traded US firms. A firm is defined to be family owned by SDC if the family ownership is at least 20%. This definition is similar to that employed by La Porta et al. (1999). We identify a total of 307 acquisitions of family owned firms over the sample period of We then verify SDC announcements with Lexis Nexus and Infotrac searches. The sample is limited to bidder firms for which return data are available from the University of Chicago Center for Research on Security Prices (CRSP) database. The final sample contains 307 acquisitions for which bidder abnormal returns can be calculated, and 306 completed deals for which sufficient accounting data and supplemental information on the bidder, deal, and family owners of the target are available. 1 Our sample period spans the period of January 1, 1984 to December 21, 2000, allowing us to identify a wide array of acquisitions of family owned targets and to perform long run analyses as well as announcement returns. Next, we cull target specific data which include the value of the transaction, the method of payment, the percent of ownership at the time of the acquisition, the age of the company, the relationship between the founder and the family member who sold the firm, and the public or private status of the target. We also collect additional information about the acquirer which includes the size of the firm, a profitability measure and leverage data. The sources for the data are : Securities and Data Corporation, DataStream, First Call, Market Guide, Dow Jones News Retrieval, Securities and Exchange Commission filings, the Wall Street Journal and New York Times Indices, the Directory of Corporate Affiliations, W.T Grimm s Mergerstat Review, Dun and Bradstreet Million Dollar Directory, Mergers and Acquisitions journal, and target company web sites. We provide the results of this unique and exhaustive data collection in Tables 1 and 2. 1 We find only four cases of multiple bidders and no hostile takeovers or white knight situations.

6 274 J Econ Finan (2014) 38: Table 1 Deal characteristics A. Financial Variables Mean (median) Maximum (minimum) Standard deviation Value of transaction/total (11.40) (2.00) 9.98 acquirer assets (%) Average age of family (54.50) (3.00) owned firm Percent family ownership 60 (55) 100 (20) 28 B. Method of payment Number Percent of total Cash Stock Mixed Total C. Public Status Number Percent of total Public Private Total This table provides information on the characteristics of the acquisition transactions of family owned firms. Panel A provides financial characteristics of the deals. Panel B provides information on the method of payment in the transaction. Panel C provides the breakdown of the sample into publicly owned and privately held family firms Table 1 presents the deal characteristics for the target family owned firm. The value of the transaction, on average (median), is 15.96% (11.40%) of the acquirer s assets. The target companies are on average (median) (54.50) years old. Mean (median) family ownership is 60% (55.0%), demonstrating a fairly large family stake in the target companies. We are able to identify at least 39% of the acquisitions of family firms were financed via stock while cash transactions accounted for only 10% of the deals. Moreover, approximately three fourth (77%) of the family firms are privately owned. Panel A of Table 2 presents the ownership structure that we are able to verify for 191 family-owned firms in our sample. It is interesting to note that while ownership levels seem to be concentrated in the ranges of 20 39% (yielding 29.84% of firms with ownership data), ownership again peaks at the 50 59% level of family ownership (19.37% of the firms with ownership data), and then again at the 100% fully family owned (21.47% of the firms with ownership data). Thus, it appears that majority ownership is an important issue for many of the firms in our sample. Our unique, hand-collected dataset also allows us an insight into the generational distance from the founder for 149 firms. The existing CEO s relationship to the family founder is detailed in Panel B of Table 2. Most of the firms for which we are able to acquire this information seem to be first-generation firms, and 61 firms have a founder in place at the time of the merger. Only 15 firms show a fourth-generation link to the founder.

7 J Econ Finan (2014) 38: Table 2 Family characteristics A. Percent ownership of selling family Ownership distribution Number of firms Percent of firms with available information 20 29% % % % % % % % % Total of firms with data NA 116 Total 307 B. Relationship of selling CEO Relationship of selling CEO to the Number of firms Percent of firms with available information founder of the firm Founder Child Grandchild Subsequent Total of firms with data NA 158 Total 307 C. Role of founding family post acquisition Subsequent role of founding family Number of firms Percent of firms with available information New executive role Board Member Consultant No role Old management remains in place Total of firms with data NA 197 Total 307 This table provides details on the characteristics of the target family involvement. Target specific data are culled from a search of a number of sources, namely, Securities and Data Corporation (SDC) International Mergers and Acquisitions Database, DataStream, Lexus/Nexus, First Call, Market Guide, Dow Jones News Retrieval, Securities and Exchange Commission filings, the Wall Street Journal and New York Times Indices, the Directory of Corporate Affiliations, W.T.Grimm s Mererstat Review, Dun and Bradstreet Million Dollar Directory, Mergers and Acquisitons Journal, and target company websites. Targets are both public and private firms The subsequent role of the founding family is documented for 110 firms in our sample, and we present these results in Panel C of Table 2 Interestingly, for those

8 276 J Econ Finan (2014) 38: firms for which data was available, most families remain with the firm in one role or the other, with only 10 firms choosing to have no role whatsoever. 3.2 Announcement returns methodology Event study methodology is used to identify the wealth effects to bidders subsequent to the announcements of acquisitions of family owned targets. Daily excess (abnormal) returns are calculated by subtracting the expected returns from the actual returns for each day of the window in question. The ordinary least squares market model is used to specify the expected returns generating process. We use the market model parameter over the estimation period from t= 110 to t= 11 relative to the announcement day t=0. The standardized cross-sectional method of Boehmer et al. (1991) with Scholes and Williams (1977) betas is used to test for significance. For each firm, the abnormal return is calculated as the difference between the firm s actual return and the expected return generated by the market model. Thus, the abnormal returns are calculated as: Where: AR ij ¼ R ij ER ij AR ij is the abnormal return on day i for each firm j R ij is the actual return on day i for each firm j, and ER ij is the expected return generated from the market model on day i for each firm j. The average excess return for any day is calculated by summing over the ARs for the N firms in the sample and dividing by N. The cumulative average excess returns (CARs) over a multi-day event period are calculated by summing the average excess returns over the T day event window. 3.3 Long term stock performance methodology The long run stock performance is examined using the Fama-French 3 factor model. The BHAR methodology suggested by Lyon et al. (1999) has been found to suffer from the cross correlation problem arising because matching on firm-specific characteristics cannot completely remove the correlation between event firms returns. More recently, long run performance has been measured using the Fama- French 3 factor model (Yook 2010; Hyland 2008; Ferreira et al. 2008). Yook (2010) suggests that the BHAR methodology presents a modeling problem since no benchmark provides a perfect estimate of the expected return and errors from the difference between the return of an event firm and its benchmark are compounded in computing long-term buy-and-hold returns (Yook 2010, ). The returns are calculated using the Fama-French three factor model using the formula: R p;t R f; t ¼ a þ b R m;t R f; t þ ssmbt þ hhml t þ " p;t Where R p,t is the event portfolio s return in month t; R f,t is the 1 month Treasury bill rate, observed at the beginning of the month; R m,t is the monthly market return in month t; SMT t is the average return on a small market capitalization portfolio minus

9 J Econ Finan (2014) 38: the average return on a large market-capitalization portfolio; HML t is the average monthly return on a high book-to-market equity portfolio minus the average monthly return on a low book to market equity portfolio. (Fama and French 1993). 4 Empirical evidence 4.1 Abnormal announcement returns Table 3 presents the cumulative abnormal returns for the three day window ( 1, +1), the two day window ( 1,0), and the event day (0,0) where day 0 is the day of the initial public announcement of the acquisition. We also provide the number of cumulative abnormal returns which are positive and negative for the ( 1, +1) window. For the 306 bidders on family owned firms for which returns data was available, we find that the bidders exhibit a positive abnormal return on the announcement of the acquisition, with a CAR of 0.98%, significant at the 1% level. Thus, the acquisition of the family-owned firm is a wealth-generating event for the bidders. Further analysis is done to try to determine the source of the positive returns. 4.2 Returns by level of family ownership To further examine the nature and extent of entrenchment and ownership concentration, we parse our sample by percent of family ownership. The research on entrenchment of blockholders clearly demonstrates that blockholders can be either beneficial or detrimental to the firm performance, depending on the concentration of holdings. Therefore, we partition the targets into three levels of family ownership: high (76 100%), medium (51 75%) and low (20 50%). The breakpoints of 50% and 75% were found by Stulz (1988) and McConnell and Servaes (1990) respectively and 20% is the minimum required by the Securities Data Corporation (SDC) International Mergers and Acquisitions database. Table 4, Table 3 Abnormal returns to bidders of family owned firms CAR event windows # of firms ( 1, +1) (z-stat) ( 1, 0) (z-stat) (0,0) (z-stat) +/ ( 1, +1) (sign test) (2.99)*** 0.60 (1.99)** 0.44 (2.73)*** 161/135 (2.43)** This table presents the announcement period cumulative abnormal returns (CARs) for ( 1, +1), ( 1,0), and (0, 0) event windows for the total sample of bidders of family owned targets. Excess returns are calculated using the market model estimated from 110 to 11 days prior to the event announcement. CARs represent the cumulative market model-adjusted change over the relevant event window. The CRSP equally-weighted market index is used. The Z statistics (given in parentheses) are based on the standardized cross-sectional method. The number of positive and negative CARs for the ( 1, +1) window are reported in the last column, with the test statistic for the nonparametric generalized sign test reported in parenthesis under +/ Significance levels are given by: *significant at.10% **significant at 5% ***significant at 1%

10 278 J Econ Finan (2014) 38: Table 4 Subsamples for cumulative abnormal returns Panel A. Cumulative abnormal returns by family ownership CAR Event Windows # of firms ( 1, +1) (z-stat) ( 1, 0) (z-stat) (0, 0) (z-stat) +/ (sign test) High ownership (2.80)*** 1.04 (2.45)** 0.72 (2.533)*** 50:30 (1.74)** Medium ownership (3.09)*** 1.97 (3.36)*** 1.41 (3.67)*** 28:19 (1.64) Low ownership ( 1.96)** 0.98 ( 2.61)*** 0.56 ( 1.86)* 18:37 ( 2.09)*** Difference between the means: high vs. medium (t statistic) ( 2.36)** ( 1.88)* ( 1.94)* Difference between the means: medium vs. low (t statistic) (3.42)*** (2.56)** (2.39)** Panel B. Cumulative abnormal returns by generational distance Bidders n ( 1,+1) (z-stat) ( 1,0) (z-stat) (0,0) (z-stat) +/ ( 1, +1) (sign test) Founder (1.91)* 1.38 (1.83)* 0.26 (0.42) 31 :19 (2.11)** Parent (0.69) 0.04 ( 0.47) 0.34 (0.44) 23 :20 (0.92) Grandchild or subsequent (2.04)** 0.75 (1.78)* 0.35 (1.37) 22 :18 (0.88) Excess returns are calculated using the market model estimated from 110 to 11 days prior to the event announcement. CARs represent the cumulative market model-adjusted change over the relevant event window. The Z statistics (given in parentheses) are based on the standardized cross-sectional method. The number of positive and negative CARs for the ( 1, +1) window are reported in the last column, with the test statistic for the nonparametric generalized sign test reported in parenthesis under +/. Panel A provides CARs for the bidders by percent of family ownership of the family owned targets. High ownership is defined by the family owners of the target owning between 76 and 100% of the target shares at the time of the acquisition. Medium ownership is defined by the target family controlling between 51 and 75% of the shares of the target at the time of the acquisition. Low ownership is defined by the family owners of the target controlling between 20 and 50% of the target shares at the time of the acquisition. Panel B provides CARs based on the generational distance of the selling CEO to the founder. Significance levels are given by: *significant at.10% **significant at 5% ***significant at 1%

11 J Econ Finan (2014) 38: Panel A provides evidence on the market reaction based on ownership concentration. The returns to bidders of targets with high ownership levels (76% to 100%) are CARs of (1.65%) and significant at the 1% level. It appears that the market perceives that perhaps the bidder has the potential to liberate assets of the target, which may be under- or inefficiently utilized by the current entrenched owners. When ownership concentration is in the medium range of 51 75%, the bidder experiences a positive CAR of 2.25%, which is significant at the 1% level. The presence of strong target monitors through the incentive alignment mechanism of family ownership leads to positive cash flow implications to bidder shareholders. At low levels of ownership (20 50%), bidders experience a negative CAR of 0.78%, significant at the 5% level. Moreover, the difference between the means tests show that the medium level of ownership has statistically higher abnormal returns than either the low or the high ownership cases. Thus, Hypothesis 1 is confirmed that there is a curvilinear relationship between the percent ownership in the family firm and the abnormal returns of the acquirer. Taken together, the results for the high, medium, and low ownership concentrations have the following implications. Bidders have the ability to generate positive value by removing an entrenched blockholder at the highest level of family ownership, although the costs of engaging in corporate governance restructuring, and potential problems with information asymmetry inherent in a closely held firm, offset the potential gains to some extent. Bidders experience the strongest market reaction when the target ownership is in the medium range, and where the gains from monitoring are not yet offset by the costs of entrenchment. We see that bidders of medium ownership firms experience the most significant positive returns, implying that these target firms have the best performance, and most likely, the most competent monitoring. We find that at the low levels of family ownership, bidders CARs are negative. This finding is in congruence with the majority of the research that finds insignificant or negative bidder returns at announcement of an acquisition. 4.3 Returns by generational distance from founder We also examine abnormal returns by generational distance from the founder and present these results in Panel B of Table 4. Recall that the returns shown are abnormal returns where AR ij =R ij ER ij and R ij is the actual return, ER ij is the market model expected return for each firm, and AR ij is the excess or abnormal return. Bidders record the strongest abnormal returns when purchasing from the founder. The abnormal or excess return is 1.91% and it is statistically significant. This abnormal return is supported by the literature; however our finding of a statistically significant 1.19% return for the grandchildren of the founder is contrary to the literature. While it is true that the acquisitions when the founder is in control have the highest returns, the fact that the grandchildren have higher returns than their parents is in dispute of the entrenchment hypothesis. Thus Hypothesis 2 is only partially supported. 4.4 Returns by public vs. private target status Panel A of Table 5 provides announcement abnormal returns for bidders of public and privately held targets. Abnormal returns for bidders of public targets are

12 280 J Econ Finan (2014) 38: Table 5 Subsamples for cumulative abnormal returns Panel A. Cumulative abnormal returns by target public status n ( 1,+1) (z-stat) ( 1, 0) (z-stat) (0, 0) (z-stat) +/ ( 1, +1) (sign test) Public ( 0.01) 0.44 ( 1.35) 0.49 ( 1.61) 32:39 ( 0.14) Private (3.40)*** 0.84 (2.89)*** 0.68 (3.78)*** 140:95 (3.93)*** Difference between the means: private vs. public (t statistic) (2.06)** (2.72)** (2.14)** Panel B. Cumulative abnormal returns by method of payment Cash (1.70)* 0.66 (1.47) 0.18 (0.95) 62:49 (1.84)* Stock (1.98)** 0.64 (0.31) 0.97 (1.78)* 22:22 (0.52) Mixed (2.04)** 1.85 (1.67) 1.44 (1.99)** 18:13 (1.91)* Terms Not Disclosed (0.70) 0.03 (0.07) 0.19 (0.54) 18:16 (0.60) Difference between the means: cash vs. stock (t statistic) ( 0.42) (0.04) ( 0.91) Difference between the means: cash vs. mixed (t statistic) Difference between the means: stock vs. mixed (t statistic) ( 1.69)* ( 1.72)* ( 2.20)** ( 1.21) ( 1.70)* ( 1.37) Excess returns are calculated using the market model estimated from 110 to 11 days prior to the event announcement. CARs represent the cumulative market model-adjusted change over the relevant event window. The Z statistics (given in parentheses) are based on the standardized cross-sectional method. The number of positive and negative CARs for the ( 1, +1) window are reported in the last column, with the test statistic for the nonparametric generalized sign test reported in parenthesis under +/. Panel A provides CARs for the bidders by target firm status. Panel B provides CARs based on the method of payment Significance levels are given by: *significant at.10% **significant at 5% ***significant at 1% negative and insignificantly different from zero, consistent with an efficient takeover market documented in previous research. However, the abnormal returns of 1.20% to bidders of privately held family owned firms are highly significant and positive. The z statistic is 3.40 with a p value of 1%. In addition, the difference between the means for private versus public targets has a t statistic of 2.06 and is significant at the 5% level, so Hypothesis 3 is confirmed. These results are consistent with value generation arising from the acquisition of targets in less liquid takeover markets with incomplete information described by Morck et al. (1988) and Fuller et al. (2002). 4.5 Returns by method of payment Next, we partition the sample by method of payment. Our results, shown in Panel B of Table 5, show positive statistically significant event window CARs of.82% for bidders with cash transactions, 1.36% for bidders in pure stock financed transactions, and 2.31% for mixed transactions, which are partially stock financed. Unlike Chang (1998), who finds insignificant returns to bidders in cash transactions, we find marginally significant CARs of 0.82% for bidders in pure cash financed transactions.

13 J Econ Finan (2014) 38: However, the tests for difference between the means shows that there is no statistically significant difference between cash and stock, but mixed is significantly higher than cash for all windows and higher than stock for the window ( 1, 0). Thus Hypothesis 4 is rejected. 4.6 Announcement returns cross-sectional regression results Thus far, we have examined the univariate results for acquirers of family firms. We next test the results in a multivariate framework. We estimate bidder returns as a function of several target and transaction characteristics such as the percent of family ownership, ROA, leverage, log of market value, 2 relative value of the transaction, and an interaction variable for method of payment (non-cash) and whether the firm is privately owned. In addition, in a manner similar to Anderson andreeb(2003), we model the non-linear relation via the variable ownership squared. We use dummy variables to investigate the impact of method of payment (cash, stock, or mixed) and for whether the target is a public or private family owned firm. We provide six model specifications to illustrate the robustness of model and variable explanatory power. Table 6 presents the results of cross-sectional results which are qualitatively similar to our univariate results. In the first four models, we use the CARs for the full sample, while the fifth model is for private targets only and the sixth for public targets only. The first five models have statistical significance overall as well as several statistically significant variables. We are able to confirm the findings in prior research regarding the non-linear relationship between ownership and firm value. The coefficient estimate of percent ownership is positive and the coefficient estimate of percent ownership squared is negative, both of them statistically significant in the five models. This suggests that indeed, medium range levels of ownership are lucrative to bidders; however, these gains from incremental increases in ownership begin to decline at higher levels of ownership. We also find that target size is a significant determinant of bidder abnormal returns, as bidder abnormal returns decrease as target size increases. Similarly, bidder returns are negatively impacted when acquiring a public target family firm, even controlling for the percent owned by the family. The relative value of the transaction to the size of the bidder also marginally impacts bidder returns. Furthermore, the coefficient estimates on the interaction of non-cash method of payment and private firm are significant in three out of the four models. Bidders returns are higher when they acquire private family firms and pay with stock or some combination of stock and cash. We also perform separate regressions for public only targets and for private only targets. For the privately held targets, the signs and significance of the variables remain the same, but for the publicly held targets, there are no longer significant independent variables and the model itself is not significant as shown by the F statistic, suggesting that it is the privately held targets which are driving the overall results. 2 Since there is a large variation in the size of the firms, we use a log transformation to normalize the variable (McDonald, 2009)

14 282 J Econ Finan (2014) 38: Table 6 Cross-sectional regression results Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Dependent variable = CARs to acquirer Private only Public only Constant (1.57) (1.73)* (1.70)* (2.35)** (1.89)* ( 0.41) Pct own (1.85)* (1.90)* (1.88)* (1.87)* (2.07)** (0.39) Pct own ( 1.84)* ( 1.99)** ( 1.77)* ( 2.04)** ( 2.32)** ( 0.85) Log(market value) ( 2.57)** ( 2.51)** ( 2.55)** ( 2.97)*** ( 3.20)*** ( 0.15) ROE ( 1.30) ( 1.26) ( 1.47) ( 1.36) ( 1.41) (0.63) Leverage (debt/total assets) ( 0.69) (0.75) ( 0.92) (0.63) Public= ( 2.25)** ( 2.41)** ( 3.21)*** Cash= (0.56) (0.59) ( 0.16) (0.88) Value/AT (1.82)* (1.69)* (1.77)* (0.86) (1.42) (0.22) Private*Cash (1.36) (1.69)* (2.23)** (2.36)** R % 20.4% 19.9% 17.4% 13.4% 7.1% F 4.32*** 4.36*** 3.87*** 3.81*** 4.00*** 0.46 N This table provides the results of cross-sectional regression with ( 1, 0) bidder event period returns as the dependent variable. Independent variables are: percent family ownership, percent family ownership squared, logarithm of market value, (Since there is a large variation in the size of the firms, we use a log transformation to normalize the variable (McDonald 2009)) Return on Equity (ROE), Leverage (Debt/Assets Ratio), Value/AT (value of the transaction to the acquirer total assets), a dummy variable that equals one if the target is a public family firm, a dummy variable if the method of payment is pure cash, and an interaction variable for private and non-cash method of payment. The z statistic is given in the parentheses Significance levels are given by: *significant at.10% **significant at 5% ***significant at 1%

15 J Econ Finan (2014) 38: Long run returns In Table 7, we report the analysis of Fama-French 3 factor long horizon returns. Our results indicate that firms that acquire family owned targets have insignificant returns the first year, but have significant large negative returns for years 2 and 5. Thus Hypothesis 5 is not proved. By 5 years following the acquisition, bidders had a significant 44.5% return. The results in Table 7 also suggest that in the later years, acquirers of public family owned targets did not perform as badly as did the private firms although the difference between the means test is statistically insignificant. Table 7 Long run returns to acquirers of family owned firms Panel A: Total sample n=274 firms that are remaining at end of 5 years 6 months (z-stat) 1 year (z-stat) 2 years (z-stat) 5 years (z-stat) 0.51 (0.31) 2.51 ( 1.10) ( 6.48)*** ( 11.56)*** Panel B: By status of target Public status of target Public targets N= ( 0.52) 6.82 ( 1.89)* ( 3.39)*** ( 5.49)*** Private targets N= (0.54) 1.10 ( 0.39) ( 5.41)*** ( 10.00)*** Difference between the means: private vs. public (t statistic) 1.68* Panel C: By target ownership Ownership percentage High 6.87 ( 2.82)*** 5.64 ( 1.28) ( 2.11)** ( 5.48)*** Medium 4.81 (1.88)* 2.04 (0.59) 1.81 ( 0.38) 6.91 ( 1.17) Low 0.88 (0.24) 7.65 ( 1.51) ( 2.16)** ( 1.87)* Difference between the means: high vs. medium (t statistic) Difference between the means: medium vs. low (t statistic) Difference between the means: low vs. high (t statistic) ( 3.22)*** ( 2.04)** ( 2.22)** ( 3.27)*** (2.25)** (2.50)** (2.06)** (2.25)** (1.88)* ( 0.24) ( 1.06) ( 1.48) The returns are calculated using the Fama-French three factor model using the formula: R p;t R f; t ¼ a þ b R m; t R f ; t þ ssmbt þ hhml t þ " p;t Where R p,t is the event portfolio s return in month t; R f,t is the 1 month Treasury bill rate, observed at the beginning of the month; R m,t is the monthly market return in month t; SMT t is the average return on small market capitalization portfolio minus the average return on a large market-capitalization portfolio; HML t is the average monthly return on a high book-to-market equity portfolio minus the average monthly return on a low book to market equity portfolio.. The z statistic is given in the parentheses. High ownership is defined by the family owners of the target owning between 76 and 100% of the target shares at the time of the acquisition. Medium ownership is defined by the target family controlling between 51 and 75% of the shares of the target at the time of the acquisition. Low ownership is defined by the family owners of the target controlling between 20 and 50% of the target shares at the time of the acquisition. Significance levels are given by: *significant at.10% **significant at 5% ***significant at 1%

16 284 J Econ Finan (2014) 38: This result is particularly interesting since the announcement returns for privately held firms was significantly higher than that of public firms. A similar result was found for newly public firms that made acquisitions by Wiggenhorn et al. (2007) which suggests that the absorption of privately held firms presents additional challenges. Table 7, Panel C, shows the effect of the ownership percentage of the target company. For the high ownership the 6 month returns are 6.87% significant at the 1% level. By 2 years, the negative return has nearly doubled and by 5 years, the return is 40.87% again significant at the 1% level. The results for the low ownership show an insignificant.88% the first 6 months, but 15.11% and 15.91%, for the 2 and 5 year returns, both significant. The results for the medium ownership are quite different. At 6 months, the returns are 4.81% significant at the 10% level, and while the returns in years 2 and 5 are negative, the percent is much smaller and the results are statistically insignificant in any case. The difference between the means test show that the returns for acquirers of the medium ownership family firms, while negative, are significantly higher than those for either the high ownership or the low ownership. Thus, as can be readily seen, the beneficial results of acquiring family firms with the medium ownership level persist in the long run. 5 Conclusion Family owned firms comprise a vital and significant segment of our economy. We examine bidder returns when family owned businesses are acquired and the results of our study become increasingly important as the demographics of the American business community shift in the coming decades, and as more family owned firms come on the block. Contrary to previous research, we find that bidders accrue positive returns at the announcement of the merger. Consistent with earlier findings (Morck et al. 1988) that suggest a range of values of inside ownership could facilitate the development of appropriate monitoring mechanisms, we demonstrate that market reaction is most advantageous for the bidder of a firm possessing a midrange family ownership, where there is less entrenchment and greater transparency. Similar to Anderson and Reeb (2003), we are able to confirm this non-linear relation between market reaction and ownership concentration in our cross-sectional analysis. Interestingly, the bidder is unable to capture any of the gains of the acquisition when low ownership targets are acquired. However, bidder returns are negatively impacted when acquiring a public target family firm, even controlling for the percent owned by the family. Unlike Chang (1998), we do not find that stock transactions are significantly higher than cash payments, but we do find statistically higher returns for mixed payments. Our cross section results show that bidders returns are significantly higher when private targets are acquired and when private firms are acquired in stock or a combination of stock and cash transactions. Overall, we find support of the entrenchment hypothesis and document that the acquisition of a family owned firm, unlike most acquisitions, is viewed positively by the market. Our long term results show that for the first year, acquirers of family owned firms earn insignificant abnormal returns, but by year 2 and continuing to year 5, bidder

17 J Econ Finan (2014) 38: firms experience significant and large negative abnormal returns. Additionally, while the acquirer announcement returns are higher for privately held firms, just the opposite is true for the long term. Acquisitions of privately traded family owned firms results in bidder returns 15% lower in the long run, though they are also significantly negative. Long run returns are again better for medium ownership with returns being significantly positive the first 6 months and then slightly negative, but insignificant, for years 1 through 5. The hugely negative and significant returns for the acquirers of both the high and low ownership family firms indicate that acquirers would certainly do well to consider the ownership structure of the target firms. References Agrawal A, Jaffe J, Mandelker G (1992) The post-merger performance of acquiring firms: a reexamination of an anomaly. J Finance 47: Ali A, Chen T, Radhakrishan (2007) Corporate disclosures by family firms. J Account Econ 44: Anderson RC, Reeb DM (2003) Founding family ownership and firm performance. J Finance 58(3): Anderson RC, Mansi SA, Reeb DM (2003) Founding family ownership and the agency cost of debt. J Financ Econ 68(2): Barontini R, Caprio L (2006) The effect of family control on firm value and performance: Evidence from continental Europe. Eur Financ Manag 12(5): Bennedsen M, Nielsen KM, Perez-Gonzalez F, Wolfenzon D (2007) Inside the family firm: The role of families in succession decisions and performance. Q J Econ 122(2): Blanco-Mazagatos V, De Quevedo-Puente E, Castrillo L (2007) The trade-off between financial resources and agency costs in the family business: An exploratory study. Fam Bus Rev 20(3): Boehmer E, Musumeci J, Poulsen A (1991) Event study methodology under conditions of event-induced variance. J Financ Econ 30(2): Brown DT, Ryngaert MD (1991) The mode of acquisition in takeovers: taxes and asymmetric information. J Finance 46: Capron L, Shen J (2007) Acquisitions of private vs. public firms: private information, target selection and acquirer returns. Strateg Manag J 28: Chang S (1998) Takeovers of privately held targets, methods of payment, and bidder returns. J Finance 53 (2): Claessens S, Djankov S, Fan J, Lang L (2002) Disentangling the incentive and entrenchment effects of large shareholdings. J Finance 57(6): DeAngelo H, DeAngelo LE (1985) Managerial ownership of voting rights: a study of public corporations with dual classes of common stock. J Financ Econ 14:33 71 Dube S, Glascock J (2006) Effects of the method of payment and the mode of acquisition on performance and risk metrics. Int J Manag Financ 2: Faccio M, McConnell J, Stolin D (2006) Returns to acquirers of listed and unlisted targets. J Financ Quant Anal 41(1): Fama EF, French KR (1993) Common risk factors in the returns on stocks and bonds. J Financ Econ 33:3 56 Fama E, Jensen MC (1983) Separation of ownership and control. J Law Econ 26: Ferreira E, Sinha A, Varble D (2008) Long-run performance following quality management certification. Rev Quant Financ Account 30: Franks J, Jarris R, Titman S (1991) The postmerger share-price performance of acquiring firms. J Financ Econ 29(1):81 95 Fuller K, Netter J, Stegemoller M (2002) What do returns to acquiring firms tell us? Evidence from firms that make many acquisitions. J Finance 57: Gleason K, McNulty J, Pennathur A (2005) Returns to acquirers of privatizing financial services firms: an international examination. J Bank Financ 29: Hyland D (2008) The long-run performance of diversifying firms. J Econ Financ 32: Jensen M (1986) Agency costs of free cash flow, corporate finance and takeovers. Am Econ Rev 76:

18 286 J Econ Finan (2014) 38: Jensen MC, Meckling WH (1976) Theory of the firm: managerial behavior, agency costs, and ownership structure. J Financ Econ 3: Jensen M, Ruback R (1983) The market for corporate control. J Financ Econ 11:5 50 La Porta R, Lopez-De-Silanes F, Shleifer A (1999) Corporate ownership around the world. J Finance 54 (2): Loderer C, Martin K (1992) Postacquisition performance of acquiring firms. Financ Manag 21:69 79 Loughran T, Vijh A (1997) Do long-term shareholders benefit from corporate acquisitions? J Finance 52: Lyon J, Barber M, Tsai C (1999) Improved methods for tests of long-run abnormal stock returns. J Finance 54(1): Maury B (2006) Family ownership and firm performance: empirical evidence from Western European corporations. J Corp Financ 12: McConnell J, Servaes H (1990) Additional evidence on equity ownership and corporate values. J Financ Econ 27: McDonald J (2009) Handbook of biological statistics. Sparky House Publishing, Baltimore Miller D, LeBreton-Miller I, Lester R, Cannella A (2007) Are family firms really superior performers? J Corp Financ 13: Morck R, Shleifer A, Vishny R (1988) Management ownership and market valuation: an empirical analysis. J Financ Econ 20: Morck R, Shleifer A, Vishny R (1990) Do managerial objectives drive bad acquisitions? J Finance 45 (1):31 47 Mulherin J, Boone A (2000) Comparing acquisitions and divestitures. J Corp Financ 6: Prachyakom P (2010) Keeping it in the family: Proper succession plan prepared well in advance-essential for family businesses. McClatchy-tribune Business News, May 17, 2010 Roll R (1986) The hubris hypothesis of corporate takeovers. J Bus 59: Scholes M, Williams J (1977) Estimating betas from nonsynchronous data. J Financ Econ 5(3): Schwert G (1996) Markup pricing in mergers and acquisitions. J Financ Econ 41: Slovin M, Sushka M (1993) Ownership concentration, corporate control activity, and firm value: Evidence from the death of inside blockholders. J Finance 48(4): Stulz R (1988) Managerial control of voting rights: financing policies and the market for corporate control. J Financ Econ 20:25 54 Swanstrom M (2006) Corporate governance and the abnormal returns to acquisition announcements. J Bus Strateg 23: Travlos N (1987) Corporate takeover bids, methods of payment, and bidding firms stock returns. J Finance 42: Villalonga B, Amit R (2006) How do family ownership, control and management affect firm value? J Financ Econ 80: Weston J, Siu J, Johnson B (2001) Takeovers, restructuring, & corporate governance. Prentice-Hall, Saddle River Wiggenhorn J, Gleason K, Madura J (2007) Going public to pursue acquisitions. Q Rev Econ Financ 47 (2): Yermack D (1996) Higher market values of companies with a small board of directors. J Financ Econ 40 (2): Yook K (2010) Long-run stock performance following stock repurchases. Q Rev Econ Financ 50: