In my recent paper, I study whether and how private equity (PE) fund investors (hereafter general partners or GPs) manipulate their fund performance during fundraising periods. Recent studies (i) have found abnormally high PE fund valuations during fundraising periods and (ii) have debated (but have not settled) whether these valuations reflect manipulating the existing funds’ values (hereafter net asset values or NAVs) or timing their fundraising activities during periods of peak performance. As I elaborate below, a fund’s NAV can be decomposed into (i) valuation multiples (hereafter multiples or market multiples) and (ii) earnings of the portfolio firms. I examine the components of the NAVs and provide evidence that funds managed by low reputation GPs show inflated valuation multiples and inflated financial performance of their investments during fundraising, which is consistent with the manipulation hypothesis.
To study whether and how GPs inflate their current fund performance during fundraising, I exploit the fact that a fund’s NAV consists of valuation multiples and portfolio firm earnings. Specifically, because many of the PE investments are private and do not have quoted market prices, GPs provide fair values using a number of valuation techniques. One of the most common methods is to apply multiples to their portfolio firm earnings, such as EBITDA or sales. Therefore, to increase NAVs through this method, GPs can either choose to apply higher multiples or manage the earnings of their investments (i.e., actions that temporarily increase firm earnings, but would later reverse; these are not illegal.
I predict that GPs use aggressive multiples, inflate portfolio firm earnings, or do both to manipulate their NAVs during fundraising periods for the following reasons. First, theory provides a rationale for performance manipulation for at least a subset of PE funds. The intuition of these models is that low reputation GPs are “forced” to manage their current fund valuations because without them, low reputation GPs face even lower chance of raising a subsequent fund. Second, GPs have the ability to inflate both multiples and portfolio firm earnings. Inflating valuation multiples is possible because NAVs are calculated using GP’s discretionary assumptions and inputs. GPs can also manage portfolio firm earnings because they exert significant operational influence on their investments by (i) investing majority equity stakes in their portfolio firms, (ii) controlling the boards, and (iii) appointing managers of the portfolio firms.
Research Design
Testing my main hypotheses requires two steps. First, I partition the samples by GP reputation because prior studies show that low reputation/low performing funds have larger incentives to manipulate fund NAVs than do high reputation GPs. The intuition behind these findings is that the lack of reputation forces these GPs to rely much more on their interim fund performance for fundraising. Second, I compare the valuation metrics (either valuation multiple or portfolio firm earnings) of the funds managed by low reputation GPs to those managed by high reputation GPs.
To test whether valuation multiples increase during fundraising, for each low and high reputation GP sample, I regress the ratio of NAV/EBITDA (and NAV/Sales) on a dummy variable that indicates periods with or without fundraising. These ratios serve as proxies for valuation multiples. The key distinguishing feature of my research design from extant research is that I focus on the valuation multiples used, instead of the aggregate NAV of the fund. To control for time-invariant fund-level characteristics and time attributes, I add fund fixed effects and calendar year-quarter fixed effects, respectively.
To investigate whether portfolio firm earnings abnormally increase during fundraising, I transition to portfolio firm data and test whether PE investments owned by low versus high reputation GPs manipulate earnings during fundraising. Specifically, similar to the fund-level analysis, I regress portfolio firm earnings management (EM) on the fundraising indicator for funds with a low versus high reputation. To capture EM, I use performance-matched accruals earnings management (AEM) and real earnings management (REM). AEM is defined as managers shifting accruals to temporarily increase firm earnings; REM is defined as managers taking real actions (e.g., building up excess inventory to reduce cost of goods sold, or drastically reducing expenses that could have long-term payoffs, such as R&D expenses) to also increase earnings in the short term. I focus on measures of EM instead of conventional measures of financial performance (e.g., ROA or sales growth) variables, because EM proxies provide clearer evidence of aggressiveness than do financial performance metrics. To avoid measurement errors of EM variables noted in a previous study, I include both first-step and second-step regressors in the portfolio firm-level regressions, following their recommendations.
The regression results suggest a significant increase in valuation multiples for funds with low reputation GPs but not for high reputation GPs. In economic terms, EBITDA and sales multiples increase by 18.2% and 22.7%, respectively, compared to nonfundraising periods. Entropy-balanced samples also demonstrate similar findings. In contrast, high reputation GPs reduce their valuation multiples during the same period. Collectively, the results are consistent with my hypotheses and prior findings that low reputation GPs have stronger incentives to overstate their NAVs via an increase in valuation multiples.
Next, the results of the EM regressions indicate that portfolio firms of low reputation buyout GPs engage in both AEM and REM to inflate their earnings. Specifically, holdings of low reputation buyout GPs exhibit higher abnormal accruals (3.8% of portfolio firm assets) and abnormal production costs (14.3% of portfolio firm assets), which are consistent with my findings at the fund level. Using entropy-balanced samples exhibit statistically stronger results than the main sample. On the contrary, investments from high reputation GPs do not show (if anything, reduces EM) any evidence of EM during their fundraising. SUR tests confirm that the coefficients are substantially different between low and high reputation GPs.
Consequences
I examine the consequences of the overstated valuation multiples and the financial performance of the underlying investments. While previous literature has documented NAV management strategies to be unsuccessful, strategies executed at the portfolio firm level could potentially increase the chances of low reputation GPs to succeed in fundraising. I test and find that low reputation GPs that use some forms of EM (specifically AEM) are associated with successful fundraising, which is in contrast to previous findings. On the other hand, while having higher valuation multiples during nonfundraising periods is effective for future fundraising, having them during fundraising periods is not, consistent with findings that the LPs can look through these efforts.
Implications
This paper complements the debate whether GPs manipulate their performance by demonstrating how they achieve manipulation during fundraising (compared to previous studies that only document whether they manipulate their valuations). My evidence supports the manipulation hypothesis by showing that GPs can inflate valuation multiples and the financial performance of their portfolio firms to increase NAVs. In addition, I demonstrate that some forms of manipulation can increase the GPs’ chances of successful fundraising, which is in contrast to studies who find GPs’ efforts to manipulate fund NAVs are largely unsuccessful. By doing so, this study extends the PE literature with respect to the fund reporting behavior during fundraising.
My findings also contribute to the literature on (i) valuation of illiquid (Level III) assets and (ii) transparency of private firms. Regarding the valuation of illiquid assets, most of the research in this area has focused on whether the financial performance of portfolio firms matter for Level III asset valuation and the cross-sectional determinants of its valuation accuracy. The contribution of this paper is to introduce valuation multiples as a potential determinant of NAVs and managerial motives (fundraising) and GP reputation as novel sources of determinants of valuation accuracy. The findings have implications not only for academics but also for the regulators of the PE industry, who are increasingly interested in this subject.
With respect to the literature on transparency of private firms, I complement the literature in three ways. First, by focusing on the effect of PE investors on the EM of their portfolio firms, I demonstrate a case where long-term institutional shareholders induce EM by portfolio firms because of their short-term incentives during fundraising periods. Past studies find that, while short-term or transient investors induce EM, long-term investors suppress it. In this paper, I provide a case where long-term institutional owners can also prompt EM when these investors face short-term incentives. Second, in a PE fund setting, I show that fund managers (i.e., GPs) can inflate valuation multiples in addition to managing earnings at the portfolio firm level. This is unique compared to public firm settings because public firm (fund) managers are unable to manipulate the multiples. Finally, these findings contribute to the understanding of EM in private firms, which is an integral part of the economy and have different ownership structures.
Brian K. Baik is an Assistant Professor in the Accounting and Management Unit at Harvard Business School
This paper is adapted from his paper, “Private Equity Valuation Management During Fundraising,” available on SSRN.
