University of Amsterdam
Bachelor Thesis
The Role of Acquisition Activity of Newly Traded Firms in The
Long-Run Underperformance of Initial Public Offerings:
Empirical Analysis of the US
Author:
Supervisor:
Wouter Abrahams (10365893)
Dr. Jan Lemmen
Faculty of Economics and Business
BSc, Economics and Business, track: Economics and Finance
June 29, 2016
Statement of Originality
This document is written by Wouter Abrahams who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Abstract
This study investigates the impact of the acquisition activity of newly traded firms on the long run performance. The impact of the acquisition activity on the long run performance is
predicted using a multiple regression model that controls for other determinants of the long run performance. The results indicate that the acquisition activity of the IPO firms is a negative determinant of the long run performance, although this result is not statistical significant. The sample exists of 82 firms that went public in the US with a time span from January 1 1999 through December 31 2009.
JEL Classification: G12; G24; G34
Keywords: IPO; Acquisition; Long Run Performance
Contents
Statement of Originality 2 Abstract 3 Contents 4 1 Introduction 5 2 Literature Review 7 3 Methodology 113.1 Long run performance measurement 11
3.2 The regression model 13
3.3 Statistical assumptions 17 3.4 Hypothesis 18 4 Data 19 4.1 Data description 19 4.2 Descriptive statistics 20 5 Results 22
5.1 Long run performance after the IPO 22
5.2 Results and discussion regression analysis 23
6 Conclusion 27 References 28 Appendix 31
1 Introduction
Why do newly traded firms underperform in the long run? An unambiguously answer is still not provided ever since Ritter (1991) discovered the long-run underperformance of newly traded firms. Therefore, this anomaly has been and still is a much-debated subject in the field of financial economics. The reason is that the determinants of the long-run underperformance are not fully understood, even though researchers have tried to solve this anomaly.
This paper tries to answer the previous raised question to give the increased acquisition activityof newly traded firms as a possible explanation for the long-run underperformance of IPO firms. This possible explanation is relevant to the existing literature because it is a relatively unexplored area in the long-run underperformance of IPO firms. This seems plausible, as several academic papers have concluded that newly traded firms are active acquirers in the post-IPO period. In addition, it has been concluded in academic literature that acquiring firms underperform in the long run. As a result, it is reasonable to argue that the increased acquisition activity in the period after the IPO is an explanation for the long-run underperformance. Based on the findings of several academic papers that newly traded firms are active acquirers and that acquirers underperform in the long run, this study tries to provide an answer on the following question:
What is the impact of the increased acquisition activity due to an initial public offering (IPO) on the long-run underperformance of newly publicly traded firms?
The central question in this paper will be answered by performing a multiple regression analysis. The necessity for adding these control variables is that these parameters might influence the performance. In order to calculate the long-run underperformance of the IPO firms, this study will make use of the cumulative abnormal returns(CARs) method. The cumulative abnormal returns will be calculated based on a time period up to 5 years after the IPO. In addition to the existing academic literature, this study uses a different data set and a different methodology to answer the question raised in this paper. In particular, a different regression model is used that is based on the existing literature.
The remainder of this paper is organized as follows. Section 2 provides an overview of the relevant academic literature about the long-run underperformance of newly traded firms, the long-run performance of acquiring firms, the increased acquisition activity in the post-IPO
period and the acquisition activity as a possible explanation for the long-run
underperformance of IPO firms. Section 3 provides and analyses the data used in this study. Section 4 presents and elaborates the methodology used in the paper. Section 5 shows the results of methodology. Section 6 concludes.
2 Literature Review
Ever since Ritter (1991) wrote first about the long-run underperformance of IPOs, the academic world of financial economics has tried to explain this phenomenon. Many years later, the necessity to explain this puzzle still exists because IPOs are still regarded as an important exit strategy and investment opportunity for respectively firms and investors. As a result, underperformance in the abnormal returns of an IPO firm is harmful for investors. This section will provide an overview of the literature related to the long-run underperformance of IPOs and acquiring firms. In addition, literature will be provided about acquisitions activity of publicly traded firms as an explanation for the long-run underperformance.
A variety of explanations have been given ever since Ritter (1991) wrote about an
overpricing anomaly in the IPO valuation process. In other words, Ritter stated that a portfolio of IPO firms underperformed compared to a comparable set of firms. In his article, the
following explanation is given: ‘many firms go public near the peak of industry-specific fads’ (p. 23). This explanation implies that many investors are too optimistic about firms becoming publicly traded. In accordance with the result found by Ritter about the long-run
underperformance of an initial public offering, Ritter and Loughran (1995) stated that firms do not underperform much during periods of low issuing activity, while the underperformance is much higher during periods of high issuing activity. An explanation for the
underperformance puzzle for IPOs is not given in the paper.
As a result of the underperformance puzzle of IPOs found by Ritter (1991) and Ritter and Loughran (1995) many researchers tried to explain this puzzle. A possible explanation lies in the increased M&A activity of newly traded firms. Brau and Fawcett (2006) stated, based on a survey among CFOs in the US, that the acquisition motive is the most important motivation for firms to go public. A traditional explanation for going public is lowering the cost of capital, but the survey invalidates this explanation. It could be argued that a firm would rather go public so that it can be acquired rather than to acquire. However, Brau and Fawcett (2006) documented that newly traded firms are active acquirers. An argument for this increased acquisitions activity is that the positive inflow of cash facilitates cash acquisitions. Besides an inflow of cash, the creation of shares could serve as a method of payments for stock-based acquisitions. In particular, Brau et al. (2003) argue that shares serve a role as ‘currency’ in acquisitions with stocks. With this in mind, Brau et al. (2010) did not investigate if the newly traded firms actually increased their acquisition activity compared to the period prior the IPO.
A paper that did research this is the paper of Celikyurt, Sevilir and Shivdasani (2010) who empirically support the finding of Brau and Fawcett (2006) that acquisitions are an important motive for firms to go public. The pre-IPO period is a period of low acquisition activity compared to the post-IPO period where most newly traded firms conduct significantly more acquisitions. In particular, Celikyurt et al. (2010) find that the average IPO firm conducts only 0.43 pre-IPO acquisitions compared with four post-IPO acquisitions. They found that the number of cash financed acquisitions is positively correlated with the capital raised through an IPO, which is in accordance with the early explanation of Brau and Fawcett (2006) that an IPO facilitates acquisitions. However, it is difficult to argue that the increased acquisition activity is due to the IPO or that the IPO simply improved the ability of the firm to conduct more acquisitions. Nevertheless, the increase in post-IPO acquisitions implies that many firms go public to conduct more acquisitions. Hovakimian and Hutton (2010) find similar results regarding the increased acquisition activity of the IPO firm. In their article, it is argued that the cash raised in the IPO and the ability to acquire firms with publicly traded stocks is the key determinant for acquisitions.
In addition, Ismail (2008) makes a distinction between the impact of single and multiple acquisitions on the performance of the acquiring firm, which is an addition to the existing literature about the long-run performance of acquiring firms. In his article, a significant result was found that single acquirers outperform multiple acquirers. This is a relevant finding because it implies that more acquisition activities are harmful for the performance of the IPO firm and as a result could be one of the explanations for the IPO underperformance puzzle. A subject of matter that has been examined often in economic literature is the long-run performance of acquiring firms. The findings of these researches could potentially relate the long-run underperformance of acquiring firms with the long-run underperformance of IPOs due to an increased activity of acquisitions. This seems reasonable to argue because several economic papers conclude negative performance of acquiring firms. However, the results about the long-run underperformance of IPOs are open to discussion due to contrasting results among economic papers. In particular, researchers typically find two results: (i) significant underperformance of acquiring firms or (ii) no significant underperformance. For instance, Franks, Harris and Titman (1991) did not find statistical underperformance of acquiring firms. Nonetheless, Rau and Vermaelen (1998) conclude that acquiring firms significantly
underperform in the long run. In order to obtain unbiased results about the performance of acquiring firms, Rau and Vermaelen (1998) adjust for both firm size and book-to-market ratio. This adjustment is done based on the paper of Fama and French (1993) who argue that
the finding of a negative long-run performance of acquiring firms is due to not adjusting for book-to-market ratios. Algrawal, Jaffe and Mandelker (1992) did not make this adjustment but their paper documented a significant underperformance. However, the scientific result is questionable due to a lack of the ‘right’ methodology proposed by Fama and French (1993). According to Rau and Vermaelen (1998), the hubris hypothesis of Roll (1986) is an
explanation for the underperformance of acquiring firms. In his article, Roll (1986) reasons that managers overestimate the benefits of acquisitions due to overconfidence. Wealth loss among shareholders of the acquiring firms is a consequence of this eagerness of managers to acquire firms.
Comparably, Loughran and Vijh (1997) find similar significant results as Rau and Vermaelen (1998) do about the underperformance of acquiring firms. In addition to existing literature about the underperformance of acquiring firms, Loughran and Vijh (1997) report about the overall wealth gains among shareholders. This means that the preacquisition and postacquisition returns are combined and examined. Besides the previous addition to existing literature, their research computes the excess returns differently. As a result of these additions to the existing literature, the importance of the papers of Loughran and Vijh (1997) and Rau and Vermaelen (1998) cannot be overstated, because these papers find significant
underperformance of acquiring firms with plausible and different methods then their
predecessors did. Although the papers of Rau and Vermaelen (1998) and Loughran and Vijh (1997) confirm the underperformance of acquiring firms in the long run, it does not
investigate if acquisition activity is an explanation for the IPO long-run underperformance. The study that combines all the previous academic literature is the paper of Brau, Couch and Sutton (2012). In this article, the increase in acquisitions activity after the IPO as an explanation for the long-run underperformance of IPO firms is investigated. Although the IPO long-run underperformance is not a new anomaly in financial academic literature, the
increased acquisition activity as an explanation for the IPO long-run underperformance is a relatively unexplored explanation for the IPO long-run underperformance. Therefore, the importance of the paper in question cannot be overstated since the long-run underperformance of IPO firms is still not fully understand. Based upon the findings that acquiring firms
underperform in the long run and increased acquisition activity of IPO firms, the study of Brau, Couch and Sutton (2012) relates these two findings with the IPO long-run
underperformance. Indeed, the paper shows results that confirm the finding of Celikyurt et al. (2010) that newly traded firms are active in acquiring firms. In addition, it is concluded that the acquisition activity of IPO firms is a plausible element in explaining the long-run
underperformance. However, the paper in question concludes that the performance of acquiring firms is good in the first year. This picture changes over a longer period that is consistent with the earlier literature about the long-run underperformance of acquiring firms. In addition, Brau, Couch and Sutton (2012) argue that the performance of firms that acquire within the first year after the IPO is worse than that of firms that do not acquire within the first year. The reasons why firms perform better that do not acquire a firm within the first year after an IPO is that the learning curve is high for managers. As a result, managers will make better decisions regarding acquisitions because they are less subject to overoptimism. The findings of Brau, Couch and Sutton (2012) are built on a regression model that controls for other possible explanations for the long-run underperformance. These control variables as possible explanations will be justified in the section methodology.
A more recent paper that shows similar results is the paper of Amor and Kooli (2016). The latter study addresses a slightly different research question, namely whether there is a
difference on the long-run underperformance between frequent and non-frequent acquirers. In accordance with Brau et al. (2010), Amor and Kooli (2016) argue that managers are
overconfident in the first year after the IPO. In their article, it is concluded that the frequency of acquisitions have a negative impact on the long-run underperformance of newly traded firms. In other words, firms that acquire more than one firm within the first year of the IPO perform less in the long run than firms that acquire only one firm. In addition, Brau et al. (2010) run a survival analysis to see the effect of being a first year acquirer on the survival of the firm. Besides the conclusions, the paper uses a regression model with different control variables then Brau et al. (2010) did. Therefore, the paper in question provides further insight about the acquisition explanation for the underperformance.
The subsequent of the paper introduces the methodology used in this study to provide an answer on the research question.
3 Methodology
This chapter will introduce and explain the techniques and data used in this study in order to answer the research question. First, this study will explain why the CAR method is preferred above the BHAR method in order to measure long run underperformance. In addition, it is explained how the CARs are obtained. Secondly, the regression model is presented and justified based on the academic literature. Finally, a hypothesis will be constructed and the used data will be presented and described.
3.1 Long-run performance measurement
It has long been assumed that the buy-and-hold abnormal returns (BHAR) method was the proper technique to calculate the abnormal returns. Since the long-run underperformance of IPOs became known to the academic world, several studies have confirmed the
underperformance. In general, among these studies the BHAR method is the most frequent technique used to calculate the abnormal returns, especially in the years following Ritter (1991) who first wrote about the long-run underperformance anomaly among IPO firms. However, the latter method is not free of shortcomings. Fama (1998) argues that the cumulative abnormal returns (CAR) method should be preferred above the BHAR method becausethe CAR method shows less statistical problems. In particular, Fama (1998) argues that the BHAR method is subject to cross-correlation among abnormal returns. Further, Fama (1998) argues that the bad-model problems are more severe for the BHAR method. In
addition, Brav et al. (2000) report that the BHAR method is likely to overestimate the long-run underperformance. Consequently, the test statistics of the BHAR method are less reliable because of biasedness.
However, the CAR method is, just as the BHAR method, not free of limitations. Barber and Lyon (1997) conclude that a positive biasedness is present in the CAR method. Besides the positive biasedness, Fama (1998) argues that the average monthly return used in the CAR method is not an appropriate measure for investors with a long investment horizon. The motive behind this outline of methods is not necessarily to argue that one method is better than another. Rather, it is to show that every method has its own limitations and that the perfect method does not exist.To conclude, this study will make use of the cumulative
third, fourth and fifth year following the IPO. In order to prevent biasedness in the results, the first year will be excluded from the calculations because the first year contains returns that are obtained prior the acquisition. A year is defined as 252 trading days and the abnormal return is computed with the following formula:
𝐴𝑅!,! = 𝑅!,! – 𝑅!,!
where 𝑅! represent the realized return of the firm i for day t, and 𝑅!,! represents the
benchmark return of firm i for day t. A primary step in calculating the abnormal returns is to calculate the realized return that is computed with the following formula:
𝑅!!! = 𝑆!!!− 𝑆! 𝑆!
where Rt+1 represents the realized return, expressed as a percentage, on date t+1, where S is
the closing price of the asset on respectively date t. The second step in order to calculate the abnormal return is to calculate the benchmark. The stock prices are obtained from
Datastream. The benchmark return (𝑅!,!) is calculated with the empirical version of the capital asset pricing model (CAPM) and is given with the following formula:
(𝑅! − 𝑅!) = 𝛼!+ 𝛽! (𝑅!"# – 𝑅!) + 𝜀!
where (𝑅! − 𝑅!) represents the excess return of firm i and (𝑅!"# – 𝑅!) represents the market excess return of the market portfolio. This study assumes the risk-free interest rate to be equal to the 5-year USA Treasury bond rate. This study will use the 5-year USA Treasury bond rate becausethe maturity of the bond is equal the long-run performance horizon (5 years after the IPO) used in this study. The rates are obtained from Datastream. The sensitivity of the stock return to the market return is captured by 𝛽! , which is a common a measurement for market risk. Following the calculations of the abnormal returns, the CAR is computed according the following formula:
𝐶𝐴𝑅!,! = 𝐴𝑅!,! !
!!!
where 𝐶𝐴𝑅!,! represents the cumulative abnormal returns from day q to day s and is calculated by the taking sum of the abnormal returns of firm i for day t. The cumulative average abnormal return (CAR) is computed by dividing the 𝐶𝐴𝑅!,! by the number N of IPOs. To sum up, the CARs of the IPO firms will be used as the dependent variable, which is relevant for the regression analysis.
The next section will introduce the regression model with the CAR as the dependent variable previously described.
3.2 The regression model
The regression model used in this study is based on the models used in the papers of Brau, Couch and Sutton (2012) and Amor and Kooli (2016). The purpose of the regression model is to determine whether the long-run underperformance of IPOs is partially explained by the acquisition activity of IPO firms. Therefore, the dependent variable will be cumulative abnormal returns that will measure the long run performance for the corresponding firm.In order to determine the impact of acquisition activity on the long-run performance, several control variables are added to the model. The regression model is presented in the following equation:
𝐶𝐴𝑅!
= 𝛽! + 𝛽!𝐴𝐶𝑄𝑈𝐼𝑅𝐸! + 𝛽!𝑉𝐶!+ 𝛽!𝑃𝑅𝐸𝑆𝑇𝐼𝐺𝐸! + 𝛽!𝑈𝑁𝐷𝐸𝑅𝑃𝑅𝐼𝐶𝐼𝑁𝐺! + 𝛽!𝑃𝑅𝑂𝐶𝐸𝐸𝐷𝑆! + 𝛽!𝐻𝐼𝐺𝐻_𝑇𝐸𝐶𝐻!+ 𝛽!𝐵𝑈𝐵𝐵𝐿𝐸! + 𝛽!𝐻𝑂𝑇! + 𝛽!𝐹𝑖𝑟𝑚𝑆𝑖𝑧𝑒!+ 𝜀!
The dependent variable represents the abnormal returns for year t with a time span from the second year after the IPO to the fifth year after the IPO. The right hand side of the equation contains the independent variables and represent the determinants of the cumulative abnormal returns based on academic literature. In order to justify the used variables in the regression model, the variables will be elaborated separately.
ACQUIRE represents the acquisition activity of an IPO firm and is the primary variable, because the regression result of this variable is an answer to the research question to this study. It is a dummy variable that equals 1 if a newly traded firm acquires a firm within the first year after the IPO and 0 otherwise. As previously stated, the first year after the IPO will
be excluded from calculations in order to prevent biasedness in the calculations of the
abnormal returns. The reason is that the first year contains returns that are obtained before the acquisition, which is not relevant for this study. The variable is expected to have a negative relation with the dependent variable because the acquisition activity of an IPO firm reduces shareholders value (Loughran and Vijh, 1997).
The first control variable refers to IPOs that are backed by venture capitalists and is represented in the regression model by VC. This control variable takes the form of a dummy variable and represent 1 if the IPO firm is backed by a venture capitalist. The relevance of this control variable is based on the paper of Brav and Grompers (1997). In this article, it is
concluded that venture-backed IPOs outperform IPOs of firm that are not venture-backed. In general, venture-backed firms tend to have better corporate governance compared to firms that are not venture-backed (Brav and Grompers, 1997). The reason is that the venture capitalists remain part of the board after the IPO is completed. As a result, non-venture backed firms perform worse than venture-backed companies. A second explanation is that venture-backed firms are less dependent on internal generated cash flows for growth
opportunities for the reason that venture capitalists have a better network with the commercial bankers (Brav and Grompers, 1997). As a result, asymmetric information is less and venture-backed companies will have easier access to loans. To sum up, the variable VC is expected to have a positive influence on the abnormal returns.
The second control variable is based on the prestige of the underwriter that is involved in the process of an IPO and takes the form of PRESTIGE in the regression. The variable is a dummy and equals 1 if the underwriter of IPO firm is prestigious and zero otherwise. The prestige of the underwriter is based on the paper of Corwin and Schultz (2005) who rank various underwriters. Therefore, the variable will equal 1 if the underwriter is ranked with at least a 7.5 on a scale from 1 to 9. Both the papers of Carter and Manaster (1990) and Carter, Dark and Singh (1998) conclude that the performance of an IPO firm is influenced by the prestige of an underwriter. In particular, the prestige of an underwriter is expected to have a positive relation with the performance of the IPO firm. The reason is that prestigious
underwriters are associated with lower risk (Carter and Manaster, 1990). Carter et al. (1998) argue that the underwriter protects its reputation by underwriting and promoting stocks that have a less severe long-run underperformance. As a result, investors have less incentive to perform a due diligence research on the firm that wants to become publicly traded.
UNDERPRICING is measured by taking the difference between the offer price and the first-day closing price as a percentage of the offer price. As previously stated, the first year
after the IPO is excluded from the calculations for the abnormal returns. However, the underpricing level might have an indirect impact on the abnormal returns. The relevance of this variable is based upon the article of Welch (1989) who argues that high-quality firms have a higher level underpricing in order to obtain a higher price at the seasoned equity offerings (SEO). This is a relevant given, because several academic papers have concluded that SEOs have a negative impact on the long-run performance of newly traded firms. For example, Spiess and Affleck-Graves (1995) state that firms that conduct SEOs significantly underperform. As a result, UNDERPRICING is expected to have a negative impact on the long-run underperformance.
PROCEEDS are captured by taking the natural logarithm of the capital raised with the IPO, based on the same definition that Amor and Kooli (2016) applied in their regression model. The importance of this variable might be based on the free cash flow hypothesis (Berk and DeMarzo, 2011), which states that a surplus of cash leads to inefficiency and agency costs. As a result, harm is done to the IPO firm that will lead to a decrease in operating-performance. Another explanation might be that the capital raised on the IPO date is used for acquisitions. As stated earlier in this study, an excess of capital might be used for acquisitions that might explain the underperformance of IPO firms in the long run.
HIGH_TECH is a dummy variable that equals one if the firm is a technology or Internet firm and zero otherwise. We control for high technology firms because these firms have different characteristics than firms that are active in the ‘old economy’. In particular, high tech firms have more growth opportunities. Therefore, this expectation of growth might be priced in in the stock price of the high tech firm. As a result, a distinction has to be made between high tech firms and non high-tech firms in order to estimate the abnormal returns properly. Loughran and Ritter (2004) argue that the level of underpricing, which might influence the abnormal returns, has changed in the past decades due to the appearance of high tech firms. In addition, Loughran and Ritter (1995) argue that young firms have a tendency to be overvaluated by investors who want to believe that they have discovered a new booming company. Therefore, the variable HIGH_TECH is expected to have a positive impact on the dependent variable.
Another control variable that is added to the regression model is the dummy variable BUBBLE. Coakley, Hadass and Wood (2007) conclude that IPOs underperform much during the years of the Internet bubble. This paper shows the relevance for the latter control variable as it controls for poor quality IPOs that were dominant during the Internet bubble years of 1998 to 2000. As a result, this study will also control for the bubble years prior the financial
crisis that started around 2008. To sum up, the BUBBLE variable is a dummy and takes the value of one if the IPO occurred from 1998 through 2000 and from 2005 through 2007 and zero otherwise. As a result of the findings of Coackley, Hadass and Wood (2007), the variable BUBBLE is expected to have a negative impact on the dependent variable.
HOTis the seventh variable in the regression model and controls for high issuing activities of IPOs. Once more, the variable is a dummy variable that equals one if the IPO occurred during a period of high issuing activity and zero if otherwise. The underpricing level will be used as an indicator for the ‘hotness’ or ‘coldness’ of the market. This is based on the paper of Ritter (1984), who uses the level underpricing as an indicator for economic segment. In accordance with the paper of Ritter (1984), the study of Helwege and Liang (2004) make use of an underpricing level of at least 25% as an indication for high issuing markets. In order to define ‘hot’ markets, this study will also use an underpricing level of at least 25%. In other words, the market is assumed to be ‘hot’ if the underpricing level of the IPO firm is at least 25%. To sum up, the control variable HOT is expected to have a negative impact on the abnormal returns. This expectation is based on the findings in the paper of Loughran and Ritter (1995) who conclude that firms do not underperform much during a period of low issuing activity. In their article, it is argued that the opposite is true when a firm issues equity during a period of high issuing activity. In this case, firms are associated with a higher degree of underperformance in the long run.
FirmSize is the last control variable that has been added to the regression model with the primary purpose of controlling for the size of firms. The variable is measured by taking the natural logarithm of the total assets of the firm in question. The importance of this control variable is based upon the findings of Moeller, Schlingemann and Stulz (2004) who argue that the return for acquiring firm shareholders is higher for firms that are smaller compared to larger firms. The reason is that larger firms offer a larger acquisition premium in comparison with smaller firms (Moeller et al., 2004). As a result, the control variable FirmSize is expected to have a negative impact on the dependent variable.
Table 3.1 will give a summarization of the previously described variables.
Table 3.1 Summarization of the regression variables
Variable Expected sign Reason
ACQUIRE - Acquisitions destroy shareholders value
VC + More value creation due to a better corporate governance
PRESTIGE + Prestigious underwriters are associated with better performing IPOs
UNDERPRICING - A high degree of underpricing results in SEOs which will have a negative impact on the performance
PROCEEDS - A surplus of cash leads to unnecessary acquisitions HIGH_TECH + High tech firms have more growth opportunities BUBBLE - Equities are overpriced during periods with unrealistic
expectations
HOT - Hot issuing markets are associated with a higher degree of underperformance
FirmSize - Larger firms offer a larger acquisition premium
3.3 Statistical assumptions
One of the assumptions of the multiple regression model is that the error term is assumed to be homoscedastic. This assumption implies that the error term is unconditional constant, meaning that it is not correlated with the dependent variables (Stock and Watson, 2012). This study will use the Breusch-Pagan test in order to test whether the error term is homoscedastic. The outcome of the test indicates that the error term is heteroskedastic instead of
homoscedastic. Therefore, this study will use a regression with robust standard errors. Secondly, to achieve unbiased results, the independent variables should not be perfectly correlated. Perfect correlation arises if the correlation between the independent variables is equal to 1. Table 3.2 shows that the regression model does not suffer from perfect
Table 3.2 Correlation matrix
3.4 Hypothesis
This section will construct a hypothesis based on the academic literature regarding the subject in this study. The purpose of this study is to investigate the impact of acquisition activity in the post-IPO period on the long-run underperformance of newly traded firms. The primary variable (ACQUIRE) in this study is expected to have a negative impact on the dependent variable. The arguments behind this expected negative impact are described in section 3.2. The dependent variable in the regression model represents the abnormal returns, an indicator of the long-run performance, and the coefficient represents the impact of acquisition activity on the long-run performance. As a result, the hypothesis is constructed as follows:
H0: β1 = 0 H1: β1 < 0
In words, the long-run underperformance of newly traded firms is partly explained by the acquisition activity. This is based on the papers of Brau, Couch and Sutton (2012) and Amor and Kooli (2016) who conclude that acquisitions done by IPO firms are an important
determinant of the long-run underperformance. Both papers refer to the research done by Celikyurt et al. (2010), who conclude that newly traded firms are active acquirers. In addition, several studies have shown that acquiring firms underperform in the long run. Therefore, the acquisition activities of IPO firms are expected to have a negative impact on the long-run underperformance.
The following section will provide and describe the data set used in this study. In particular, the data about the previous described regression variables will be provided.
ACQUIRE VC PRESTIGE UNDERPRICING PROCEEDS HIGH_TECH BUBBLE HOT FirmSize
ACQUIRE 1 VC -0.0178 1 PRESTIGE -0.023 -0.328 1 UNDERPRICING 0.072 -0.047 0.079 1 PROCEEDS -0.085 -0.147 0.312 -0.165 1 HIGH_TECH -0.152 0.195 -0.057 0.347 -0.185 1 BUBBLE 0.073 0.111 0.005 0.185 -0.094 0.156 1 HOT -0.088 -0.031 -0.122 0.635 -0.099 0.273 0.096 1 FirmSize 0.075 -0.112 0.160 -0.156 0.642 -0.241 -0.123 -0.113 1
4 Data
4.1 Data description
The data set used in this study is obtained from the widely used ZEPHYR. The IPO sample has a time period from January 1 1997 through 31 December 2009 with a total of 88 firms. The time period for the IPO sample is limited to the year 2009 there is allows this study to calculate the 5-year returns after the IPO. The sample excludes REITs, closed-end funds and financial firms. Table 4.1 will provide an overview of the IPOs year by year.
Table 4.1 Frequency distribution
IPO year Frequency % Of Total sample Number of 1st year
acquirers % of 1st year acquirers 2000 17 19.3% 5 29.4% 2001 1 1.1% 0 0 2002 5 5.7% 2 40% 2003 4 4.5% 1 25% 2004 10 11.4% 2 20% 2005 16 18.2% 5 31.3% 2006 13 14.8% 5 38.5% 2007 14 15.9% 2 14.3% 2008 2 2.3% 0 0 2009 6 6.8% 1 16.7% Total 88 100% 23 26.1%
Table 4.1 provides a breakdown of the IPO sample used in this study. The IPO sample contains 88 firms and makes a distinguishing between firms that are acquirers within the first year after the IPO and non-acquiring firms within the first year after IPO. The fifth column shows the percentage of first-year acquirers. In total, 26.1% of the IPO firms acquire another firm within the first-year of the post-IPO period.
Noticeable, the frequency distribution shows that the majority of the IPOs in the IPO sample occurred during the years that can me defined as the bubble years. In particular, most of the IPOs occurred in years 2000, 2005, 2006 and 2007, which are described as bubble years according to this study.
The data concerning the regression variables in the regression model is obtained from ZEPHYR and Thomson One. Table 4.2 provides, in addition to the comprehensively
description of the regression variables in section 3, an overview of the methods of measurement and the sources of the data regarding the regression variables.
Table 4.2 Data explanations measurement and source
Variable Measured Source
ACQUIRE 1 for 1st year acquisition
0 for no 1st year acquisition
ZEPHYR
VC 1 for venture capitalist backed IPO 0 for no venture capitalist backed IPO
ZEPHYR
PRESTIGE 1 for IPO firms with underwriters with at least a ranking of 7.5 on a scale from 1 to 9
0 for IPO firms with underwriters with a ranking of less than 7.5 on a scale from 1 to 9
ZEPHYR
UNDERPRICING The difference between the offer price and first-day closing price, as percentage of the offer price
ZEPHYR
PROCEEDS The natural logarithm of the proceedings of the IPO firm. ZEPHYR
HIGH_TECH 1 for high technology firms 0 for otherwise
Thomson One
BUBBLE 1 for bubble
0 for no bubble
ZEPHYR
HOT 1 for high issuing markets (‘hot’) 0 for no high issuing markets (‘cold’)
ZEPHYR
FirmSize The natural logarithm of the total assets of the IPO firm. Thomson One The regression variables all have a sample size of 82 that equals the number of IPO firms studied.
4.2 Descriptive statistics
Table 4.3 provides an overview of the descriptive statistics of the variables used in the regression model. In order to control for the skewness, the natural logarithm is taken of the variables PROCEEDS and FirmSize. The reason is that the values of PROCEEDS and FirmSize are too big compared to the other control variables and this will result in a high
skewness. The natural logarithm has the primary advantage of shrinking the values and therefore shrinking the skewness.
Table 4.3 Descriptive statistics
Variable Average Min Max Std. dev. Skewness Obs.
ACQUIRE 0.2674 0 1 0.4452 1.0696 82 VC 0.1744 0 1 0.3817 1.7466 82 PRESTIGE 0.7558 0 1 0.4321 -1.2122 82 UNDERPRICING 0.2462 -0.1265 3.7778 0.5107 4.6929 82 PROCEEDS 12.7549 11.0791 16.7936 0.9609 1.0186 82 HIGH_TECH 0.2442 0 1 0.4321 1.2122 82 BUBBLE 0.6395 0 1 0.4830 -0.5916 82 HOT 0.2326 0 1 0.4149 1.2887 82 FirmSize 13.4832 8.4114 17.9183 1.6953 -0.2017 82
The variable ACQUIRE is the primary variable in this study and the other variables are included in the regression model in order to control for other determinants of the dependent variable.
5 Results
This chapter introduces the cumulative average abnormal returns (CARs), which will give a clear picture about how the IPO firms performed on average in the long run. In particular, the CAR is computed for the first 5 years after the IPO, with exclusion from the first year. In addition, acquirers are distinguished from non-acquirers in order to determine whether acquirers underperform compared to non-acquirers. Finally, the regression results will be presented and analysed in order to answer the research question.
5.1 Long run performance after the IPO
Table 5.1 provides an overview of the CARs for the different years after the IPO and makes a distinction between acquirers and acquirers. This study finds that both acquirers and non-acquirers do not underperform in the long run with a period up to 5 years after the IPO.
Acquirers have an average positive performance of approximately 4.3% compared to 7.2% for non-acquirers as shown in table 3.1. This finding is in contradiction with the paper of Ritter (1991) and Loughran and Ritter (1995) who both document about underperformance among newly traded firms. As stated earlier in this study, the result of underperformance among several academic papers is still subject to debate and therefore remains controversial. In discussing these results, Brav and Gompers (1997) conclude that newly traded firms do not underperform. Although this study uses a different benchmark in order to determine the abnormal returns, the result of no underperformance is in accordance with the results of Brav and Gompers (1997). Noticeable, this study uses the CAR method to conclude whether firms underperform in the long run. Gompers and Lerner (2003) conclude that underperformance is present for newly traded firms with the BHAR method. However, in their article it is
concluded that the underperformance disappears if the CAR method is used. This might be an explanation why this study did not find underperformance in the long run for both acquirers and non-acquirers.
However, acquirers seem to underperform compared to non-acquirers by 2.86% on
average. Although this underperformance is not substantial, the results in table 5.1 show that the difference in performance between acquirers and non-acquirers is substantial in the second and third year after the IPO. Respectively, the difference is 13.47% in the second year after the IPO and 16.43% in the third year after the IPO. The underperformance among acquirers disappears in the fourth and fifth year.
Table 5.1 Performance long run acquirers vs. non-‐acquirers
Variable Acquirers Non-acquirers Difference
𝐶𝐴𝑅! -0.0737 0.0610 -0.1347
𝐶𝐴𝑅! -0.0439 0.1204 -0.1643
𝐶𝐴𝑅! 0.2408 0.1052 0.1356
𝐶𝐴𝑅! 0.0496 0.0006 0.0490
𝐶𝐴𝑅!,! 0.0432 0.0718 -0.0286
Table 5.1 presents the 𝐶𝐴𝑅𝑠 that is defined as the average of the cumulative abnormal returns per year. The 𝐶𝐴𝑅𝑠 are given for respectively the second, third, fourth and fifth year after the IPO. CAR!,! represent the
average of the cumulative abnormal return for the second, third, fourth and fifth year. The first year is excluded from calculations. The cumulative average abnormal returns (CARs) are computed by dividing the 𝐶𝐴𝑅!,! by the
number N of IPOs. The last column presents the difference between the returns of acquirers and non-acquirers. The difference found between acquirers and non-acquirers is not statistical significant. This significance is tested with the Welch t-test (written in the appendix).
The following section will present the results of the regression analysis in order to answer the research question.
5.2 Results and discussion regression analysis
The purpose of the regression model is to determine whether the acquisition activity of newly traded firms is an explanation for the long-run underperformance, which is the main purpose of this study. The results of the regression analysis are presented in table 4.2.
Table 5.2 Regression results
Variable Estimation of coefficient Robust p-values
ACQUIRE -0.0171 (0.6630) 0.797 VC -0.0255 (0.0818) 0.756 PRESTIGE 0.0865 (0.0712) 0.226 UNDERPRICING -0.0664 (0.1597) 0.678 PROCEEDS -0.0264 (0.0337) 0.435
HIGHTECH 0.0849 (0.0781) 0.278 BUBBLE 0.0653 (0.0548) 0.234 HOT 0.1013 (0.1218) 0.406 FirmSize 0.0165 (0.0172) 0.339 Intercept 0.0533 Observations 82 R2 0.0177
Table 5.2 provides the results of the coefficient used to explain the cumulative abnormal returns for the 5-year period with the first year excluded from calculations. The sample exists of 82 firms and is applied on American IPO firms. The variable ACQUIRE equals 1 if the IPO firm performed an acquisition in the first year after the IPO. The variable VC equals 1 if the IPO is venture-backed. The variable PRESTIGE equals 1 if the
underwriting firm of the IPO is defined as prestigious. The variable UNDERPRICING is the natural logarithm of the difference between the first-day closing price and the offer price, as a percentage of the offer price. The variable PROCEEDS is the natural logarithm of the capital raised with the IPO. The variable HIGHTECH equals 1 if the IPO firm is a high technology firm. The variable BUBBLE equals 1 if the IPO occurred during a period with a surplus of irrational investors. The variable HOT equals 1 if the IPO occurred during hot IPO market. The variable FirmSize is the natural logarithm of the total assets of the firm, based on the first known yearly
statement after the IPO.
Although this study does not find underperformance among IPO firms over a period of 5 years after the IPO, it is concluded in the previous section that acquiring firms underperform compared to non-acquiring firms. This finding is in accordance with the regression result of the dummy variable ACQUIRE. This variable is the primary variable in this study, because it captures the impact of a first-year acquisition on the performance of newly traded firms. According to the regression results, the variable is associated with a 1.71% decrease in the performance of the IPO firm. This result is in expectation with the expected sign of the
variable described in section 3 and is in accordance with the findings of Brau et al. (2012) and Amor and Kooli (2016). However, the result is not significant. Therefore, the described impact might not describe the role of acquisitions properly.
VC and PRESTIGE are added to the model in order to control for venture-backed IPOs and IPOs that are executed by prestigious underwriters. The variables are both included in the model as a dummy variable. According to the academic literature, venture-backed IPOs are expected to perform better than non venture-backed IPOs. However, this study finds that
venture-backed IPOs, captured by the control variable VC, are expected to have a negative impact on the performance. The variable PRESTIGE is expected to have a positive effect on the performance, as expected based on academic literature described in section 3. This result is similar to the findings of Brau et al. (2012). However, the variable is not statistical
significant. Therefore, the variable should be interpreted with caution.
The variables UNDERPRICING and PROCEEDS control for respectively the underpricing level on the first trading day and the capital raised on the IPO date. The variable
UNDERPRICING is supposed to have a negative impact on the performance, as shown in regression results. The variable PROCEEDS is supposed to have a negative impact on the performance. This study finds no statistical significance for the variables. However, the findings of these two variables are as expected and are in accordance with the findings of Amor and Kooli (2016). Amor and Kooli (2016) also did not find a statistical significance for the previous two control variables.
In order to control for high technology firms and periods with unrealistic expectations, the variables HIGHTECH and BUBBLE are included in the regression model. The variable HIGHTECH is a dummy variable and is supposed to have a positive impact on the
performance. This regression result is analogous to the result of Amor and Kooli (2016). The regression result illustrate that the variable BUBBLE has a positive impact on the
performance of newly traded. This finding is partly in accordance with Amor and Kooli (2016) who do not find unambiguously results regarding the variable BUBBLE. In addition, their paper also does not find significant results.
The last two control variables that are included in the regression model are the variables HOT and FirmSize. The variable HOT is added to the regression model in order to control for economic segment and the results show a positive impact on the performance of the IPO firms. This positive relation is not as expected based, because this study expected a negative relation based on the academic literature outlined in section 3. However, the finding of a positive relation is in accordance with the results of Amor and Kooli (2016), who also find a positive relation. As a result, if an IPO occurred during a period of high issuing, the
performance of the IPO firms is expected to perform better for 10.13%. The variable FirmSize is added purely with the purpose to control for the size of the firm. The results indicate a positive impact on the performance of the IPO firm. However, the result is not significant. All the previous considering, the hypothesis in this study can be tested. In academic literature, it has been documented that newly traded firms are active acquirers. Therefore, the acquisition activity might influence the long-run performance. This study expected a negative
relation between the acquisition activities of newly traded firms on the long-run performance. The reason is that several academic papers reported poor long-run performance of acquiring firms. According to the regression results, this study shows a negative relation between the performance of IPO firms and the acquisition activity after the IPO. This results is as
expected, but the result is not statistical significant. Therefore, the result is not unambiguously and should be taken into consideration with caution.
6 Conclusion
Since the long run underperformance of newly traded firms became known the to the
academic world, scientists have tried to solve this interesting puzzle. This study tries to prove that the increased acquisition activity after the IPO is an important determinant of the long run underperformance. This seems plausible because several academic papers conclude poor performance among acquiring firms.
Considering the previous, this study expects a more severe performance in the long run for acquiring than non-acquiring firms. The causality between the underperformance in the long run and increased acquisition activity after the IPO is tested using a multiple regression analysis.
This study finds a negative impact between acquisitions done by newly traded firms and the long run performance. This finding is in accordance with the existing literature, although the result of this study is not statistical significant. In defence of the statistical insignificant results, the studied subject in this study is relatively unknown and an unexplored field in the field of finance. Therefore, the results of this study are an addition to the existing literature.
The used methodology and data in this study might have its limitations. First, the regression model might be subject to omitted variable bias, because the performance of newly traded firms might be influenced by more factors. In addition, the used sample size is small compared to existing literature. This might be a reason why this study finds no significant result between increased acquisition activity and long run performance of IPO firms. The distinction between diversifying and non-diversifying acquisitions might be an addition to future research. None of the existing papers with the chosen subject in this study makes a distinction between these types of acquisitions. This might be relevant, because it could influence both the performance and the valuation of the IPO firm.
References
Agrawal, A., Jaffe, J. F. and G. N. Mandelker, G.N. (1992). The Post-Merger Performance of Acquiring Firms: A Re-Examination of an Anomaly. Journal of Finance, 47, 1605–1621. Amor, S.B., and Kooli, M. (2016). Do Acquisitions Affect IPO Long-Run Underperformance? Evidence from Single vs. Multiple Acquirers. Journal of International Financial Markets, Institutions and Money, 40 (2016), pp. 63-79.
Barber, B.M. and Lyon, J.D., (1997). Detecting long-run abnormal stock returns: the empirical power and specification of test statistics. Journal of Financial Economics, 43, 341 –372. Berk, J., and DeMarzo, P. (2011). Corporate Finance. Harlow: Pearson.
Brau, J. C., and Fawcett, S.E. (2006). Initial Public Offerings: An Analysis of Theory and Practice. Journal of Finance, 61, 399–436.
Brau, J. C., Francis, B. and Kohers, N. (2003). The Choice of IPO versus Takeover: Empirical Evidence. Journal of Business, 76, 583–612.
Brau, J., Couch, R. and Kohers, N. (2010). The long-run underperformance of acquiring IPO firms. Brigham Young University, Provo, UT. Unpublished working paper.
Brau, J.C., Couch, R.B. and Sutton, N. K. (2012). The desire to acquire and IPO long-run underperformance. Journal of Financial Quantitative Analysis, 47, pp. 493–510.
Brav, A., and Gompers, P. A. (1997). Myth or Reality? The Long-Run Underperformance of Initial Public Offerings: Evidence from Venture and Nonventure Capital-Backed Companies. Journal of Finance, 52, 1791–1821.
Brav, A., Geczy, C. and Gompers, P. (2000). Is the abnormal return following equity issuance anomalous? Journal of Financial Economics, 56, pp. 209–249.
Carter, R. and Manaster, S. (1990). Initial Public Offerings and Underwriter Reputation. Journal of Finance, 45, 1045-1067.
Celikyurt, U., Sevilir, M and Shivdasani, A. (2010). Going Public to Acquire: The Acquisition Motive for IPOs. Journal of Financial Economics, 96, 345–363.
Coakley, J., Hadass, L. And Wood, A. (2007). Post-IPO Operating Performance, Venture Capital and the Bubble Years. Journal of Business Finance & Accounting, 34, 1423–46. Corwin, S.A., and P. Schultz, P. (2005). The role of IPO underwriting syndicates: Pricing, Information Production, and Underwriter Competition. Journal of Finance, 60, 443-486.
Fama, E.F. and French, K.R. (1993). Common Risk Factors in the Returns on Stocks and Bonds. Journal of Financial Economics, 43, 3–56.
Fama, E.F. (1998). Market Efficiency, Long-Term Returns, and Behavioral Finance. Journal of Financial Economics, 49, pp. 283–306.
Franks, J., Harris, R. and Titman, S. (1991). The Postmerger Share-Price Performance of Acquiring Firms. Journal of Financial Economics, 29, 81–96.
Graves, J.A. and Spiess, D.K. (1995). Underperformance in Long-Run Stock Returns following Seasoned Equity Offerings. Journal of Financial Economics, 38, 243-267.
Helwege, J. And Liang, N. (2004). Initial public offerings in hot and cold markets. Journal of Financial and Quantitative Analysis, 39, 541-569.
Hensler, T., Springer, R. and Rutherford, R. (1997). The Survival of Initial Public Offerings in the Aftermarket. Journal of Financial Research, 1, 93–110.
Hovakimian, A., and Hutton, I. (2010). Merger-Motivated IPOs. Financial Management, 39, 1547– 1573.
Ismail, A. (2008). Which Acquirers Gain More, Single or Multiple? Recent Evidence from the USA Market. Global Finance Journal, 19 (2008), 72–84.
Loughran, T. and Ritter, J.R. (2004). Why Has IPO Underpricing Changed Over Time? Financial Management, 33, pp. 5–37.
Loughran, T., and Vijh, A.M. (1997). Do Long-Term Shareholders Benefit from Corporate Acquisitions? Journal of Finance, 52, 1765–1790.
Loughran, T., and Ritter, J.R. (1995). The New Issues Puzzle. Journal of Finance, 50, 23–51. Moeller, S.B., Schlingemann, F.P. and Stulz, R.M. (2004). Firm size and the gains from acquisitions. Journal of Financial Economics, 73, 201-228.
Rau, P. R., and Vermaelen, T. (1998). Glamour, Value and the Post-Acquisition Performance of Acquiring Firms. Journal of Financial Economics, 49 (1998), 223–253.
Ritter, J. R. (1984). The “hot issue” market of 1980. Journal of Business, 2, 215-240.
Ritter, J. R. (1991). The Long-Run Performance of Initial Public Offerings. Journal of Finance, 46, 3–27.
Roll, R. (1986). The Hubris Hypothesis of Corporate Takeovers. Journal of Business, 59, 197– 216.
Stock, J. and Watson, M.V. (2012. Introduction to Econometrics. New York: Prentice Hall Welch, I. (1989). Seasoned offerings, imitation costs, and the underpricing of initial public offerings. Journal of Finance, 2, 421-449.
