Do Private Company Targets that Hire Big 4 Auditors Receive Higher Proceeds?*
研究美国私营企业出售控制权时，聘请四大审计师能否提高出售收益。发现四大审计师能显著降低私营企业折价，带来数百万美元的增值，对估值从业者和企业主有参考价值。
This study examines the impact of Big 4 auditor choice on sale proceeds of controlling interests in U.S. private firms.1 Prior research, such as Becker, DeFond, Jiambalvo, and Subramanyam 1998 and Francis, Maydew, and Sparks 1999, suggests that Big 4 auditors provide higher audit quality than non–Big 4 auditors for U.S. public companies. Further, Big 4 auditors reduce the cost of equity capital (Khurana and Raman 2004), the cost of debt capital (Mansi, Maxwell, and Miller 2004; Pittman and Fortin 2004), and increase initial public offering (IPO) proceeds (Willenborg 1999) for U.S. public companies. Prior research has not examined the impact of audit choice on perceived audit quality for U.S. private companies that sell all of their shares or assets. Relative to U.S. public companies, private companies in our sample are small and have poor information environments. Hence, the private company setting is an especially important one when examining the relation between Big 4 auditor choice and perceived audit quality. Recent empirical work supports the existence of a private company discount (PCD). Koeplin, Sarin, and Shapiro (2000) and Officer (2007) estimate the PCD to be in the 17 percent to 20 percent range for the valuation models used in our primary empirical analysis. Our corresponding estimates of 20 percent to 40 percent, based on a new multivariate estimation procedure that we introduce into the literature, are higher. The primary focus of our study, however, is to explore whether the PCD declines when the private seller engages a Big 4 auditor, consistent with higher quality audits. Such a finding would point to at least one explanation for the PCD, namely, the information quality facing the buyer. A lower PCD for private sellers engaging a Big 4 auditor implies higher sale proceeds for such firms. Using the PCD approach, the dollar value decrease in enterprise value for our representative firm due to not hiring a Big 4 auditor ranges from $2.0 to $3.2 million for stock-purchased private firms and from $1.9 to $2.9 million for asset-purchased private firms. These are substantial amounts, given that the representative private firm with a Big 4 auditor has a median enterprise value ranging from $14 to $18 million for stock purchases and from $10 to $12 million for asset purchases. We employ a second approach to explore the direct impact of Big 4 auditor choice on private-firm sale proceeds that excludes public firms in our estimation. Using this approach, for a representative private firm in a stock-purchase (asset-purchase) transaction, the dollar value decrease in enterprise value due to not hiring a Big 4 auditor ranges from $3.9 to $5.2 million (from $2.6 to $3.1 million). Again, the amounts are substantial. Either approach thus suggests a pronounced impact of Big 4 auditor choice on the sale proceeds of U.S. private companies. To further explore explanations for our Big 4 auditor "premium", we assembled leading valuation practitioners in a round-table session to solicit their opinions regarding Big 4 audit quality in a private company acquisition setting. The rationale for asking the experts is that they regularly appraise these deals and should be aware of audit quality differences, if they exist. As a caveat, the opinions are just that, rather than facts, but the discussions do identify important correlated omitted variables that future research can take into account when testing for and seeking to explain Big 4 effects. These discussions indicate that participants believe that Big 4 auditors do provide higher quality audits, and that the due diligence process is easier and less likely to result in downward price adjustments. It was also pointed out that private companies with a Big 4 auditor are more likely to have strong internal controls, appropriate levels of accounting personnel, other high-quality advisors (such as an investment bank), and a well-planned-in-advance sale. These companies are also less likely to have potential contingent tax liabilities. Of interest, because many buyers will not consider purchasing a company with poor-quality accounting systems and financial statements, the number of potential buyers is correlated with the hiring of a Big 4 auditor. Even if a private company were to successfully engage a Big 4 auditor, our results document the benefits in terms of proceeds, but do not capture the costs of hiring a Big 4 auditor. Examples of these costs include the higher price of a Big 4 auditor, more sophisticated information systems, as well as more and better-qualified accounting personnel. Participants also noted, in their opinion, that private company owners do not always act optimally, either because they are less sophisticated or, perhaps, because they are overconfident in their ability to manage the sale process. We highlight that these practitioner insights extend beyond explaining the Big 4 auditor impact on valuation multiples and could provide alternative explanations for the effects of a Big 4 auditor in other settings. For example, if a Big 4 auditor is positively correlated with better information systems, more financial expertise, and a better managed company, then audit quality or auditor selection, the two most common explanations in the literature for a Big 4 effect, are not the only reasons why firms that hire a Big 4 auditor have lower cost of (debt or equity) capital. We contribute to the literature in a number of ways. Our primary contribution is to show that valuation multiples are higher for private firms with a Big 4 auditor. We demonstrate that the reason for the discount paid for private firms relative to public firms goes beyond simple differences in liquidity as many people believe (e.g., Sloan 2002; Officer 2007). We note that the IPO evidence on Big 4 equity or debt pricing effects may not extrapolate to our private-sale setting, where information environments are especially poor, relative to companies that go IPO. For example, there is no offering document subject to the scrutiny of the Securities and Exchange Commission (SEC) or being reviewed by analysts and media. In addition, an IPO company is far more likely to make significant investments in improving corporate governance and internal controls over financial reporting prior to going public, all with the anticipation of experiencing fundamentally higher levels of oversight. Our results provide private sellers with empirical evidence regarding the potential impact of auditor choice on deal proceeds and contribute to the literature examining the impact of auditor choice on the cost of capital. Second, this article presents a rigorous study of the PCD in the context of controlling interests, which is needed because, as Pratt (2001, 173) argues, the existing analysis is neither comprehensive nor thorough enough to answer once and for all the question of whether private-firm controlling interests sell for less. Our multivariate approach, which controls for other determinants of value and alternative data sources, complements and improves on the extant PCD literature. For example, the existing studies use a matched-pair approach. Although this approach has some advantages, one significant disadvantage is the additional restrictions in the sample selection process, which using our data results in a sample that is less than half that of our multivariate method. Our large sample helps to ensure that results better extrapolate to the population of private firms. Our results should be of particular interest to valuation practitioners. For example, The National Association of Certified Valuation Analysts (NACVA 2008), whose members specialize in the valuation of small and medium-sized enterprises, describe helpful research as follows: "Any research relating to the value of privately held enterprises or that helps reconcile the difference between public and private companies would be of particular interest." Section 2 contains our literature review and presents our hypotheses. Section 3 describes our sample, while section 4 discusses our research methods and results. Section 5 summarizes the discussion at our round-table session. Section 6 concludes. Private firms are different from public firms along a number of dimensions that potentially affect firm valuation. Recent empirical work supports the existence of a PCD. Koeplin, Sarin, and Shapiro (2000) employ 84 matched pairs of private and public acquisitions in the United States between 1984 and 1998 and estimate a PCD of 20 percent for EV/EBITDA but none for EV/SALES. EV is the enterprise value. It is defined as the sale price of the firms' equity plus total liabilities less current liabilities and hence it represents the entire firm value (i.e., the asset value) as opposed to just the equity value. SALES is total sales and EBITDA is earnings before interest, taxes, and depreciation and amortization. Using a matched-pairs approach and 1979–2003 data for unlisted targets, Officer (2007) estimates a PCD of 17 percent to 18 percent for the EV/EBITDA and EV/SALES models, respectively.2 While there is a growing consensus in the literature that a PCD exists on average, the explanation for the discount remains an open question. Less demand for financial information leads to less sophisticated accounting systems and weaker internal controls, all of which increases the unintentional errors in private-firm earnings. Private firms also potentially have innate characteristics that can lead to higher information risk. For example, consistent with private firms' smaller size, the product and market scope of private firms is more limited. This leads to private firms being less diversified and hence experiencing more variability in sales and operating cash flows (innate factors identified and used by Dechow and Dichev 2002 and Francis, LaFond, Olsson, and Schipper 2005). The empirical literature documents the lower quality of earnings for private firms in general, although there are some exceptions (e.g., Beatty, Ke, and Petroni 2002).3 In our setting, public and private sellers face similar income-increasing earnings management incentives. Ball and Shivakumar (2005) show that U.K. financial reporting for private firms is less conservative than for public firms due to different market demands, regulation notwithstanding. Burgstahler, Hail, and Leuz (2006) find earnings management is more pervasive in private firms across European countries. Katz (2006) finds U.S. firms that have publicly traded debt are less conservative if their equity is not publicly traded. Similar to these studies, our sample firms all prepare financial statements using the same standards (in our case, U.S. generally accepted accounting principles). The Ball and Shivakumar, Burgstahler et al., and Katz studies, however, are made possible by the fact that European private firms or U.S. firms with public debt submit their financial statements to country regulators and hence their financial statements are, in essence, public documents. In contrast, there is no regulatory oversight of our sample of U.S. private firms and hence regulatory monitoring and enforcement will be absent.4 The primary difference, however, between these studies and our setting is that our sample firms are being sold. In this setting, firm managers have incentives to take actions that increase their sale price. If management expects price to be a positive function of earnings, firms could manage accruals upwards. A large number of papers have studied earnings management around large transactions, such as the acquisition of target firms consistent with our study.5 The actual evidence of earnings management around large transactions is mixed. For example, Teoh et al. (1998a) show that IPO firms adopt income-increasing accruals. Ball and Shivakumar (2008), on the other hand, challenge these results and argue that the Teoh, Welsh, and Wong (1998a) results are unreliable for a variety of reasons. Ball and Shivakumar (2008) find no evidence of earnings management around the IPO for a sample of U.K. firms. Auditors are especially important for private firms that are being sold. Auditors provide independent verification of manager-prepared financial statements and express their opinions on the management's assertions. Audit quality is defined in terms of the level of assurances – the probability that financial statements contain no material omissions or misstatements (Palmrose 1988). Prior literature suggests that Big 4 auditors in the United States have incentives to provide high-quality audits. Big 4 auditors have large reputation capital and have more to lose in litigations because of their "deep pockets" (DeAngelo 1981). Big 4 auditors are more likely to be sued and suffer costly litigation damages. Thus, Big 4 auditors would provide higher-quality audits in order to reduce litigation risk and to protect their brand name reputation (DeAngelo 1981; Becker et al. 1998). Prior research generally suggests that Big 4 auditors are perceived to provide higher audit quality than non–Big 4 auditors. Khurana and Raman (2004) utilize the ex ante cost of equity capital and examine whether Big 4 auditors are perceived as providing higher-quality audits (relative to non–Big 4 auditors) in the United States and in the less litigious (but economically similar) countries. They find that a Big 4 audit is associated with a lower ex ante cost of equity capital for auditees in the United States but not in Australia, Canada, or the United Kingdom. Their results suggest that the estimated cost of equity capital decreases by 30 basis points, amounting to savings of $0.5 million annually, if a non–Big 4 auditee were to switch to using a Big 4 auditor.6 Other studies show that large auditors increase IPO proceeds. Beatty (1989) shows that private firms who undergo an IPO and are audited by a more reputable auditor have lower initial equity returns (consistent with receiving a higher IPO price) than IPO firms with a less reputable auditor. Balvers, McDonald, and Miller (1988) find that, relative to non–Big 8 auditors, Big 8 auditors on average reduce the level of underpricing by approximately four percent. Willenborg (1999) finds that, for IPO proceeds in excess of $6 million, underpricing is greater for non-national auditees. Researchers show that Big 4 auditees have a lower cost of debt capital for U.S. publicly traded companies, again suggesting higher perceived Big 4 audit quality. Mansi et al. (2004) examine the role of auditors from the perspective of bondholders and show that firms audited by a Big 4 auditor have a 63-basis-point lower cost of debt capital. They also find that the relation between auditor characteristics and the cost of debt is most pronounced for firms with debt that is noninvestment grade. This suggests that the role of auditors is more important as firm information risk increases. Pittman and Fortin (2004) focus on recent IPOs and examine the impact of auditor choice on debt pricing in the early public years after the IPO when firms are lesser known. They find that retaining a Big 6 auditor, which can reduce debt-monitoring costs by enhancing the credibility of financial statements, enables young firms to lower their borrowing costs. In particular, companies hiring a Big 6 auditor have an interest rate that is 143 basis points lower. They also provide evidence that choosing a Big 6 auditor affects firms' interest rates less over time and particularly benefits firms with short private histories, consistent with the information effect of auditor quality being important when information environments are poor. While a maintained assumption of audit quality studies is that Big 4 audit quality is in fact truly higher, there are contrary findings as well (e.g., Petroni and Beasley 1996).7 Researchers have provided direct evidence that large auditors provide higher quality audits than small auditors. Francis et al. (1999) report that Big 4 auditors constrain aggressive and opportunistic reporting by U.S. public companies. They find that both absolute discretionary accruals and income-increasing discretionary accruals are lower for Big 6 auditees than for non–Big 6 auditees. In addition, for small, non-venture-backed IPOs, Weber and Willenborg (2003) find that the pre-IPO opinions of larger auditors are more predictive of post-IPO stock delistings. Only two other studies have examined whether a Big 4 pricing differential exists for private companies, both involving the impact of Big 4 auditors on the cost of debt capital. For a sample of private U.S. companies issuing public bonds to qualified institutional buyers under SEC Rule 144A, Fortin and Pitman (2007) find no significant difference in the bond yield spread or credit ratings between Big 4 and non–Big 4 auditees. In contrast, for a sample of private Korean companies for which audits are voluntary, Kim, Simunic, Stein, and Yi (2010) demonstrate that, for the years prior to the 1997 Asian financial crisis, loan interest rates are lower for Big 4 compared to non–Big 4 auditees. Although not a pricing differential study, Chaney, Jeter, and Shivakumar (2004) find no audit fee premium for Big 4 auditors, once self-selection factors are controlled for, using a sample of private U.K companies. To the extent that audit fees reflect audit quality, their findings suggest no audit quality differentiation between Big 4 and non–Big 4 audited private companies. Thus, the Big 4 audit quality evidence for private companies is mixed. Further, no evidence exists for private sellers of equity or assets. Our study seeks to fill these voids. We highlight that the IPO evidence on Big 4 equity or debt pricing effects may not extrapolate to private sellers of equity or assets, a setting where information environments are especially poor, relative to companies doing an IPO. There is no offering document subject to the scrutiny of the SEC or being reviewed by analysts and media. Compared to a private company sale, an IPO company is far more likely to make significant investments in improving corporate governance and internal controls over financial reporting prior to going public, all with the anticipation of experiencing fundamentally higher levels of oversight. In addition, as documented later, private firms have innate earnings characteristics that imply a generally poorer information environment for private relative to public targets. Thus, the private seller setting is one where Big 4 choice potentially matters considerably, in terms of reducing information risk, suggesting the potential for large auditor-choice effects. This difference in information environments represents a fertile environment for examining Big 4 pricing differentials and may explain why we document Big 4 pricing premiums that are large relative to the extant Big 4 literature based on IPOs. Given the above discussion, our hypothesis is stated as follows: Hypothesis. Ceteris paribus, Big 4 auditors increase the proceeds of private firms. We note here that our hypothesis is based in part on findings in the extant literature. The discussion of our empirical results with practitioners, which we present in section 5, has led to additional reasons and insight into why this relation is expected to hold. Our study includes both stock- and asset-purchase private-firm transactions. Our motivation for including asset purchases is as follows. It is intuitive and widely accepted among both practitioners and academics that information risk is lower for an assets-only deal, compared to buying shares, because with the former the buyer is not responsible for any unrecorded liabilities. Thus, the costs of doing due diligence are lower for assets-only deals. It is empirically interesting to explore whether Big 4 price differential effects hold for asset purchases. Though information risk is lower for asset purchases, addressing these risks via a high-quality audit is still an important aspect of the deal because the buyer must be satisfied with the valuation assertions regarding acquired net assets. Thus, asset-purchase deals represent an interesting benchmark sample for comparison with stock purchases. The source of our private-firm valuation data is Pratt's Stats® published by Business Valuation Resources (BVR). It includes financial statement and transactional details on the sale of privately held firms for the period 1994–2005. This database is routinely used by intermediaries, such as accounting firms, investment and that represent buyers or sellers in these transactions. The data is from two the same that use the data in their valuation analysis contribute details of transactions to the Second, data is from regulatory in which public firms private firms these transactions in SEC such as The data in Pratt's Stats® include the (e.g., firm country information from the most recent financial statements (e.g., net cost of interest taxes, net assets, total assets, current total statement and and details the acquisition (e.g., buyer firm equity enterprise sale sale stock asset For given that only a small number of are for firms, we our sample to U.S. private firms. Although a comprehensive of financial information to the of practitioners, we private firms' financial statement data from the SEC of the U.S. because not all the information for this For example, no information from the cash but we operating cash (in particular, to the accruals for our quality and pricing As example, we the name of the auditor to auditor statement information is for the most recent but information is provided for the most recent prior to the which may not be the To and information in our we the at the of the our to in and we not only the most recent but also the prior financial U.S. public or with the SEC of an the of the and the buyer which is the number that the to that documents with the not provide these data to but to given our research The along with the enables to and the or of the public on the for private-firm For a small number of firms we the appropriate SEC or the not contain the financial statement data we We these firms from the A number of variables we are provided in the Pratt's Stats® data Given the of the Pratt's Stats® data and the of research using the we this same information from the SEC of the in order to the of process. all data in Pratt's Stats® with our If there is a difference, we the source SEC and reconcile the difference with from There is no to the Given this and process, we are that both Pratt's Stats® SEC data and our data is of quality. This process our sample of private firms to by public U.S. firms, and with financial statement data for the two years prior to the Our sample of public transactions is from the database of and public transactions are as stock purchases. The sample includes U.S. acquired by other U.S. public firms over the corresponding time period of our private-firm We our analysis to firms in which the buyer percent controlling interest after the This data is to the private firm transactions, which also represent the of percent controlling Our focus on controlling interests to from the of While the database includes financial statement it not contain all the data we for our We use a of and to the data with using firms' and We use financial statement data in to the with We all financial statement data from firms with data are from the analysis. we the private firms of Pratt's Stats® and the public firms of into one data We our sample to firms with positive value of equity and positive We financial from our sample to the effects of these We also enough data to estimate the variables used in our all the restrictions our sample of firms, of which are private stock are private asset and are public stock of our analysis a smaller sample and private stock and asset and public stock because we the and percent of variables in our contains for our sample of private stock-purchase to and private asset-purchase 4 to relative to public to Francis et al. (2005) argue that smaller firms, and firms with greater cash and sales operating and greater of have poor innate accruals quality that imply higher information risk. We focus on private stock-purchase with public targets. These private firms demonstrate innate characteristics associated with lower accruals quality, relative to public stock purchases. For example, such firms are Private firms' median total is median total sales is and median EBITDA is The of and cash from by total for private firms are that for public firms. It should be pointed out that private stock have a median operating and a lower of consistent with higher innate accruals quality for private targets. The median private firm is 5 percent of total assets. Less debt for private firms implies weaker monitoring by In addition, percent of private firms, are audited by a Big 4 auditor. A comparison of private asset-purchase (in 4 to to public similar we the evidence as consistent with a poorer information environment for private relative to public targets. This important for our in which we document Big 4 pricing premiums that are large relative to the extant Big 4 literature based on public companies. A comparison of private transactions as stock purchases (in to asset purchases (in 4 to that there are no large differences in private company characteristics such as size, and For example, the median and EBITDA for stock purchases are and compared to and million for asset purchases. 2 a similar of when all of firms stock private asset and public are into Big and non–Big We that the Big 4 auditor used in our valuation is for, among other higher innate accruals quality, consistent with the Big 4 auditor lower firm risk as by and To on this A of 2 innate quality characteristics identified in Francis et al. across Big 4 and non–Big 4 private stock-purchase targets. the one hand, private stock-purchase a Big 4 auditor are smaller asset is million, compared to million for Big a which points to lower innate earnings quality. the other hand, private stock a Big 4 auditor have a lower of operating cash flows by total compared to