Representations and warranties insurance in mergers and acquisitions

Information asymmetry between an acquirer and a target is an underlying issue in merger and acquisition (M&A) discussions and outcomes and derives from the acquirer’s incomplete knowledge of the target’s value. It is more pronounced for private targets, which, unlike public targets, are not subject to securities regulations, disclosure requirements, and public shareholder oversight. To mitigate the increased information asymmetry surrounding private targets, these sellers provide representations and warranties (referred to as “R&W” or “reps”) regarding the state of the seller’s business that are more extensive than those provided by public sellers. Through these reps, a seller promises that a set of facts about the target in the acquisition agreement are true. If the reps are found to have been breached following the closing of the deal, then the buyer’s remedies include pursuing damages under applicable law or including an indemnity provision in the merger agreement, usually accompanied by a holdback or escrow amount used to adjust the purchase price post closing. Alternatively, the parties may purchase an R&W insurance policy, and the value of damages can be recovered through an insurance claim. In this paper, we use a proprietary sample of R&W insurance policies to investigate the factors that affect the decision to purchase a policy (the decision to transfer risk to the insurer), the factors that determine the premium charged by the insurance company (the market measure of the risk), and the factors that are correlated with the likelihood of a claim (the realized risk).

Representation and warranties insurance transfers indemnification risk to a third-party insurer for an up-front premium payment.¹ These policies were first introduced in the 1990s but have only gained significant traction over the past ten years (Griffith 2020). These products arose because reps and indemnities introduce significant contracting costs for both buyers and sellers and are among the most difficult and time-consuming aspects of the acquisition agreement to negotiate (e.g., Freund 1975; Gallozi and Phillips 2002; Hill et al. 2016).² These costs arise because the seller may not have, or may not want to disclose, a complete and accurate picture of the state of the business, including the true value of its assets and liabilities. In this case, it may limit the types of warranties and the amount of indemnification it provides. In addition, risk aversion on the part of the buyer or the seller may preclude agreement on the amount or duration of an indemnity to guarantee the R&Ws. Lastly, if a seller maintains a relationship with the buyer post acquisition—e.g., the buyer retains the seller as a manager—any attempt to recover an indemnity from the seller/manager represent a direct principal-agent conflict.

In this paper, we obtain a proprietary sample of R&W insurance policies issued by a global insurance company for acquisitions of private targets between January 2011 and December 2016, representing 1,690 acquisitions and $470 billion in enterprise value.³ For each policy, our data from the insurance company includes, among other items, the premium and liability limit. For the acquisition parties, we observe deal size (enterprise value), buyer and target ownership type, target region, and target industry. We also observe claims reported for each policy and the type of reps breached, if any, as of September 2017.

Using this sample, we first investigate variables that are associated with the decision to purchase R&W insurance policies. The first set of variables we examine are industry metrics for valuation uncertainty, which follow along the lines of Datar et al. (2001) and Cain et al. (2011) who examine the role of earnouts in valuation uncertainty in private firms and who also lack the buyer and seller identities. The second set of variables are the acquirer and seller types—private equity (PE), strategic, or individual.⁴

We find that there is a higher likelihood of purchasing R&W insurance when acquisition targets belong to industries with higher target valuation uncertainty, such as greater return volatility and lower-quality financial reporting, as measured by the level of internal control weakness. We also provide evidence of lower demand for R&W insurance in industries where there are more growth opportunities, as captured by the target industry Tobin’s Q, consistent with the fact that the value of high-growth firms is predicated more on future prospects than on existing assets and liabilities, making R&W insurance less valuable. We also show that PE acquirers are more likely to purchase R&W insurance than are strategic buyers, consistent with the notion that an acquisition by a private-equity buyer exposes the buyer to more firm-specific financial risk than would a similar acquisition by a large diversified public acquirer. This is because PE buyers’ funds usually consist of a concentrated set of portfolio companies, held over a relatively short period of time. Demand for R&W insurance is also higher among PE sellers. PE sellers are more likely than strategic sellers to want to limit their post-closing indemnification liabilities on the sale of a portfolio company so as to be able to fully distribute sales proceeds to the limited partners of the PE fund at the time of the sale, thereby providing higher internal rates of return to investors (Griffith 2020). Last, we find that demand is significantly higher if the target is in a common law regime (as opposed to civil or other legal regimes), consistent with greater expected indemnity contracting costs in these legal environments.

We then examine variables associated with the premium charged by the insurance company. Our analysis reveals that the insurance company charges higher premiums for targets in industries with poorer financial statement quality, which are perceived to be associated with a greater risk of loss. We also find that a higher premium is charged when the target is in an industry with relatively high levels of research and development (R&D) to sales, indicating that the insurance company expects unrecognized intangible assets to have a greater risk of future claims. In contrast, we find that premiums are lower in targets in industries with more growth opportunities, consistent with our prior evidence of a lower likelihood of purchasing R&W insurance for high-growth targets, given their greater reliance on future prospects. We do not find differences between the premium that the insurance company charges to private equity buyers and strategic buyers. In contrast, we find that PE sellers and individual sellers pay a higher premium than do strategic sellers.

We conclude by investigating the factors that are correlated with the likelihood of a claim, which reflects the realized risk. Somewhat surprisingly, we find that most industry-related factors are not associated with the likelihood of a future claim (with the exception of R&D to sales), but that these industry factors are related to the demand for R&W insurance or the premium charged. We also find that private equity buyers are less likely to file claims. A finding that is consistent across our three analyses is that deals with private-equity sellers are more likely to involve the purchase of R&W insurance than are deals with other types of sellers, have higher premiums, and are more likely to result in a claim. Overall, the results suggest that the ex ante risks inherent in the demand and premium charges are, to a large extent, not consistent with the realized incidence of breaches.

Our study contributes to several strands of the accounting and finance literature. First, we add to research on target valuation uncertainty in acquisitions (Shleifer and Vishny 2003; Rhodes-Kropf and Viswanathan 2004; Officer 2004) and on the impact of contracting on reducing valuation costs (Cain et al. 2011). We extend this literature by demonstrating the prevalence of R&W insurance as a replacement for the seller indemnity, which is a contracting mechanism to allocate risk and reduce information asymmetry and valuation uncertainty between buyers and sellers. Second, we add to literature that examines contractual mechanisms used to allocate risk in acquisitions. Some studies examine the determinants and effects of earnout provisions—contractual mechanisms that allocate post-transaction performance risk and create value in mergers and acquisitions (Bates et al. 2018; Cadman et al. 2014; Cain et al. 2011; Datar et al. 2001; Kohers and Ang 2000). Other studies of risk allocation in acquisitions explore material adverse change or event clauses (Denis and Macias 2013; Gilson and Schwartz 2005; Talley 2009). Our study extends these papers by being the first empirical paper to explore R&W insurance, which has only been explored theoretically (e.g., Gallozi and Phillips 2002; Griffith 2020; Hill et al. 2016). Last, we contribute to the body of research on disclosures in mergers and acquisitions by extending the analysis to private targets, which constitute the majority of M&A targets (Cain et al. 2011); most studies to date have focused on the effect of disclosures on the valuation of publicly traded target companies (e.g., Raman et al. 2013; McNichols and Stubben 2015; Skaife and Wangerin 2013).

There is a rich literature on valuation uncertainty in acquisitions (Shleifer and Vishny 2003; Rhodes-Kropf and Viswanathan 2004; Officer 2004). In order to reduce this uncertainty, parties to the acquisition agreement apply different contractual mechanisms of risk allocation. One mechanism is the inclusion of material adverse change or event clauses, which are used to protect parties against the risk of unforeseen events or circumstances (Denis and Macias 2013; Gilson and Schwartz 2005; Talley 2009). Another mechanism is the inclusion of earnout provisions, which allocate post-transaction performance risk. Earnouts allow the seller to insure future growth and earnings prospects (e.g., Cain et al. 2011; Datar et al. 2001). A third, more significant mechanism is the inclusion of reps, which are a set of promises, made by the seller, warranting that statements made about the target are true, and of indemnities, which are financial guarantees to the buyer against breaches discovered in reps after the closing of the deal. Several theoretical studies have explored the role of reps and indemnities to allocate risk and to reduce information asymmetry and valuation uncertainty (e.g., Gallozi and Phillips 2002; Griffith 2020; Hill et al. 2016). However, despite reps and indemnities being more common and more fundamental to the acquisition process than other contractual features, there has not been any empirical research into their economic magnitude, most likely because information about reps, indemnities, and breaches is rarely, if ever, publicly disclosed.

We contribute to the existing literature by empirically examining R&W insurance policies, which effectively replace the seller indemnity and thereby decrease contracting costs associated with negotiating the indemnity and subsequently litigating indemnity disputes. Using a novel dataset, we investigate three key issues with respect to R&W insurance. First, we examine the factors that affect the demand for R&W insurance. Second, we explore the factors that determine the premium charged by the insurance company. Third, we examine R&W insurance claims data in order to identify the factors that are associated with the likelihood of a claim.

We begin our analysis by examining the determinants of the decision to purchase R&W insurance. As mentioned earlier, we explore two sets of variables to assess the demand for reps. Our first set of demand variables are industry metrics for valuation uncertainty and follow along the lines of Datar et al. (2001) and Cain et al. (2011), who examine the role of earnouts in valuation uncertainty in private firms. The first proxy reflects the target’s growth prospects using the target industry median Tobin’s Q ratio (target industry Tobin’s Q). The second is an indicator variable equal to one if the target’s industry has above-median standard deviation of daily returns over the prior year (target industry STD returns). The third is the industry ratio of R&D to sales (target industry R&D to Sales). We use an indicator variable equal to one if the target’s industry has an above-median ratio, which reflects both uncertainty about future prospects (Officer et al. 2009) and the extent to which there are unrecorded assets on the balance sheet (Lev and Sougiannis 1996). The financial statement reps signal the value of a firm’s recorded assets and liabilities, but most internal research-related assets are not recorded on financial statements. Therefore, these reps are a less likely source of breach for R&D-intensive firms. We add a fourth proxy, for target riskiness—the target’s industry mean rate of internal control weakness prevalence (target industry ICW). Inferences about internal control weakness are likely to signal other aspects of general management and target quality. This is consistent with prior findings that show a correlation between internal control deficiencies and information uncertainty, higher audit fees, lower financial reporting quality, and fraud (e.g., Amiram et al. 2018; Beneish et al. 2008; Doyle et al. 2007; Hogan and Wilkins 2008; Hoitash et al. 2009).

To examine the effects of target industry characteristics on R&W insurance demand, we use the following logit regression specification, where the dependent variable is an indicator variable equal to one if R&W insurance has been purchased:

We include two control variables. The first is deal size (Log(Deal Size)). We expect this variable to be positively associated with R&W insurance demand since the complexity and difficulty of conducting adequate due diligence during the acquisition timeline for larger deals may generate more breach claims. The second control variable is the target’s legal environment. We create an indicator variable (common law) that takes the value of one if the target is in a predominantly common law jurisdiction, and zero otherwise (La Porta et al. 1997). We predict that common law will be positively associated with the demand for R&W insurance, consistent with prior research that shows greater levels of investor protection and property rights in common law countries and that indicates that the enforcement of these rights often involves costly litigation efforts (e.g., La Porta et al. 2008). Since our R&W insurance data lists the target country for the US and the UK and the region for targets located in other countries, we consider a region to be common law if the majority of its acquisition activity by deal volume occurs in common law countries, using the SDC sample as a guide. In Appendix C Table 9, we provide further detail about each region’s common law classification.

Our second set of variables are the type of acquirer and seller, since these parties may assess target riskiness and the need for reps differently. While our sample includes an indicator variable if the seller is an individual, the SDC database does not record this category. Therefore, our purchase decision analysis is limited to examining PE and non-PE deal participants, wherein the non-PE participants consist of both individuals and corporate entities.¹⁰ To examine the relationship between PE firm involvement and R&W insurance purchase, we include an indicator variable equal to one if the acquirer is a PE firm (acquirer PE), and zero otherwise; and an indicator variable equal to one if the seller is a PE firm (seller PE), and zero otherwise. We examine the effects of seller and acquirer identity on R&W insurance demand by using the following logit regression specification, where the dependent variable is an indicator variable equal to 1 if R&W insurance has been acquired:

To account for industry-invariant factors, we also include industry fixed effects (based on the industry classification described in Appendix B Table 8). Table 4 presents the results of different specifications of the regressions described in Eqs. (1) and (2). For this test, we merge the SDC acquisition data used in Table 1 with our R&W insurance sample, and each observation, i, represents an acquisition in the combined sample. The SDC sample comprises 33,442 observations with the required control variables. Note that some of the acquisitions in the SDC sample may also have R&W insurance policy coverage; this will only serve to weaken our results.¹¹

As shown in Column (1) of Table 4, firms in industries with more growth opportunities (labeled by target industry Tobin’s Q) are less likely to purchase R&W insurance, consistent with the fact that the value of high-growth firms is predicated more on future prospects than on existing assets and liabilities, making R&W insurance less valuable. The coefficient on target industry STD returns is positive and significant for targets in industries with greater market volatility. The coefficients on R&D intensity (target industry R&D to Sales) and financial reporting risk (represented by target industry ICW) are not significantly associated with the purchase of R&W insurance.

The results in columns (2) and (4) show that PE acquirers are more likely to purchase insurance (acquirer PE). This result is consistent with the notion that an acquisition by a private-equity buyer exposes the buyer to more firm-specific financial risk than would a similar acquisition by a large diversified public acquirer. This is because PE buyers’ fund structure usually consists of a concentrated set of portfolio companies, held over a relatively short period of time. The coefficient on seller PE is positive and significant, consistent with PE sellers’ greater desire, relative to strategic sellers, to limit their post-closing indemnification liabilities on the sale of a portfolio company. Doing so allows them to fully distribute sales proceeds to the PE funds’ limited partners and show higher internal rates of return to investors (Griffith 2020). In Column 5, we pool all variables of interest from Eqs. (1) and (2) into one regression and find results similar to those in columns (1) through (4). Moreover, when we add the type of acquirer and seller, the coefficient on target industry ICW becomes positively and significantly associated with R&W insurance purchase, which suggests that the addition of these independent variables is important and is consistent with the significantly higher adjusted R².

Finally, in all columns, we find positive and significant coefficients on Log(Deal Size) and Common Law, consistent with our expectations. Overall, we find that R&W insurance demand is correlated with both target industry characteristics and type of acquirer and seller.

In order to draw clearer inferences regarding the association between the R&W insurance premium and target industry metrics and the type of acquirers and sellers, we follow prior studies and also control for the abnormal amounts of insurance coverage purchased (Abnormal Limit), defined as the residual coverage limit (Lin et al. 2013; Cao and Narayanamoorthy 2014; Chalmers et al. 2002). The residual coverage limit is obtained from a limit prediction model similar to Core (2000), in which the dependent variable is the logged policy limit and the explanatory variables include target industry characteristics, acquirer and seller type, an indicator variable for acquirer policy, an indicator variable for seller individual, logged deal size, and legal environment.¹² We then estimate the following two regressions:

Table 5 reports the results of estimating Eqs. (3) and (4), where the dependent variable is the logarithm of the policy premium. Focusing on target industry characteristics in Column (1), we find that the coefficient on target industry Tobin’s Q is negative and marginally significant. This result is consistent with that of the previous section, where it was shown that there is a lower likelihood of purchasing R&W insurance for high-growth firms, given their reliance on future prospects. The positive and significant coefficient on target industry R&D to Sales indicates that the insurance company expects unrecognized intangible assets to have a greater risk of future claims and takes that into account in determining the premium.¹³ The coefficient on target industry ICW is positive and significantly associated with premiums, indicating that industries with poor financial statement quality are perceived to have a greater risk of loss. This inference is supported by Panel C of Fig. 1, which illustrates that financial statement breaches are the main source of R&W insurance claims, accounting for more than 21% of all claims. Focusing on acquirer and seller type in Columns (2) to (4), we find positive and significant coefficients on seller PE and seller individual, relative to a strategic seller. In comparison, the coefficient on the acquirer’s identity, represented by acquirer PE, is not significantly associated with the premium. Our results remain qualitatively unchanged when we pool all variables of interest from Eqs. (3) and (4) into one regression (Column 5).

Across all columns, we find that the coefficient on acquirer policy is positive and significant, illustrating the insurer’s perception of heightened risk of losses for buy-side policies, consistent with buyers’ information asymmetry and the exclusion of fraud claims if the seller is the policyholder. The coefficient on abnormal limit is also positive and significant in all regressions, indicating that the insurer recognizes and charges increased premiums when the insured demands abnormally high limits, thereby countering potential adverse selection. The coefficient on common law is positive and significant, consistent with those regions’ higher levels and costs of litigation, which affects both expected losses and R&W insurance contracting costs. In all columns, larger deals result in higher premiums, given their greater degree of overall complexity and valuation uncertainty. Overall, we conclude that the insurance underwriting process takes into consideration valuation uncertainty measures as well as the type of the seller.

In this section, we investigate the factors that are correlated with the likelihood of a claim, which reflects the realized risk incurred by the insurance company. We do so by estimating the following logit regressions:

The variable Premium-to-Limit captures the insurance company’s ability to predict the likelihood of a claim being filed that is not controlled for by our other variables. Table 6 reports the results of estimating Eqs. (5) and (6), where the dependent variable is an indicator variable equal to 1 if an R&W insurance policy resulted in a breach claim.¹⁴ Focusing on target industry characteristics in Column (1), we find that only the coefficient on R&D intensity (target industry R&D to Sales) is positive and significant, a result that is in line with our earlier documentation that the insurance company charges higher premiums to firms in R&D-intensive industries (Table 5). Additionally, whereas the insurance company charges higher premiums to firms in industries with more internal control weaknesses and lower premiums to firms in industries with higher Tobin’s Q, we do not find any association between these factors and the probability of a claim for breach of reps.

Focusing on the type of acquirer (columns 2 and 4), we find that, compared to strategic buyers, private equity buyers are less likely to submit a breach-related claim. A reason offered by Griffith (2020) as to why the insurance company might not charge lower premiums to PE buyers even though they are less likely to submit a claim is that the insurance company may not have enough historical information to realize that this is the case.¹⁵ In columns 3 and 4, we document that the coefficient on Seller PE is positive and significant. In column (5), when we pool all measures into one regression, the coefficient on Seller PE remains positive and significant. This result is consistent with our previous finding that the demand for insurance and the premium charged are higher when the seller is a PE firm, compared to a strategic seller.

We also find that claims are more likely to be filed in larger deals, supporting our earlier finding that higher R&W insurance premiums are charged in larger deals. Moreover, we find that the coefficient on Premium-to-Limit is positive and significant, meaning that claims are more likely when the insurance company charges higher premiums. This finding suggests that the insurance company is able to identify other areas of potential risk not captured by the other independent variables. Last, we find that claims are more likely under common law, consistent with these regions’ higher levels and costs of litigation (e.g., La Porta et al. 2008), which affect both expected losses and contracting costs and are thus more associated with the frequency of breaches in the reps (Ramseyer and Rasmusen 2010; Browne et al. 2000). Overall, the results suggest that the ex ante risks perceived by the demand and premium charges are, to a large extent, not consistent with the realized incidence of breaches.

Drawing from a large and unique sample of R&W insurance policies, we document that industry measures of valuation uncertainty, as well as the type of acquirers and sellers, play a role in the demand for R&W insurance, the premium charged by the insurance company, and the likelihood of filing a claim. While the reps and indemnities are an effective mechanism to encourage information production to facilitate M&A, the significant contracting costs associated with seller indemnities have bolstered the purchase of R&W insurance policies in recent years. Our evidence shows that the demand for R&W insurance is increasing in proxies for indemnity contracting costs, including those associated with valuation uncertainty, risk aversion, litigation costs, and institutional features that do not favor seller indemnities. Given the rising trend in transferring indemnity risk to an insurer, there is concern that R&W insurance reduces the incentives for acquirers to perform due diligence and for sellers to produce efficient information about the state of the target, which should be a topic for future research.

Acquisitions involve many trillions of dollars. Our results attest to the demand for R&W insurance and extend the theoretical literature by documenting the use of R&W insurance in risk allocation, reduction of information asymmetry, and promotion of valuation certainty. Understanding the effect of private target disclosures on the demand and pricing of R&W insurance policies is an interesting topic for future research. Another potential future research question is whether there is a relation between the purchase of insurance and the success of acquisitions or the purchase price of targets.

A limitation of our empirical analysis is that we do not know the identities of the insured parties and do not have access to detailed financial information about them. We only have access to a limited set of buyer and target characteristics. These include geographic region, industry sector, type of entity (private equity, strategic, or individual), and acquisition value. We use industry-based proxies based on these observable characteristics, as reflected in prior literature (e.g., Cain et al. 2011; Datar et al. 2001), to make inferences about the acquisition participants. An additional limitation of our study is that we have a single source of R&W insurance data for our analysis. Without being able to observe the entire population of R&W insurance policies, the extent to which our results are generalizable is not clear. Even with these limitations, our paper provides a contribution to the nascent literature on the value of disclosures in private M&A transactions.