Permanent private equity: Market performance and transactions

• Published date: 01 June 2022
• Author: Maurice McCourt
• Date: 01 June 2022
• DOI: http://doi.org/10.1111/jfir.12277

Received: 28 May 2020

|

Accepted: 15 February 2022

DOI: 10.1111/jfir.12277

ORIGINAL ARTICLE

Permanent private equity: Market performance

and transactions

Maurice McCourt

Finance Department, University of

Melbourne, Melbourne, Australia

Correspondence

Maurice McCourt, Finance Department, Level

11, 198 Berkeley Street, University of

Melbourne, 3010 Victoria, Australia.

Email: maurice.mccourt@unimelb.edu.au

Funding information

Agence Nationale de la Recherche,

Grant/Award Number: ANR10‐JCJC‐

1810‐01

Abstract

Using market data for a comprehensive sample of publicly

listed permanent capital private equity funds, I confirm

existing results in the private equity literature and establish

new empirical facts for private equity investments. Over

2000–2019, only mezzanine funds outperformed public

markets, whereas buyout, growth, and funds of funds

underperformed somewhat, and venture capital funds

underperformed significantly. Buyout funds exhibit per-

formance persistence. Larger funds with higher asset

turnover have higher performance, and funds with higher

expenses have lower performance. Finally, contemporane-

ous and out‐of‐sample performance of buyout and mezza-

nine funds is associated with deal and exit transaction

characteristics.

JEL CLASSIFICATION

G02

1|INTRODUCTION

Although typical private equity (PE) funds have a fixed life (usually 10 years), after which they are liquidated and

capital is returned to investors, some PE fund managers are turning to a long‐term or even permanent funding

model. For example, in 2020, it was reported that one of the largest PE firms (Blackstone) raised $104 billion in

permanent capital to invest in private assets including PE, and that permanent capital constitutes one‐fifth of

J Financ Res. 2022;45:339–383. wileyonlinelibrary.com/journal/JFIR

|

339

This is an open access article under the terms of the Creative Commons Attribution‐NonCommercial‐NoDerivs License, which

permits use and distribution in any medium, provided the original work is properly cited, the use is non‐commercial and no

modifications or adaptations are made.

© 2022 The Authors. Journal of Financial Research published by Wiley Periodicals LLC on behalf of The Southern Finance

Association and the Southwestern Finance Association.

Blackstone's total funds under management, a 43% increase year over year. Apollo, another major PE firm,

announced that half of its $331 billion in assets were permanent capital (Vandevelde, 2020).

Permanent capital fundraising by traditional PE

1

fund managers (such as Blackstone) is a relatively recent

development, so little historical data are available on which an analysis of these funds can be based. In contrast,

permanent capital funds in the form of publicly listed closed‐end funds of PE investments have been in existence

for decades. Therefore, I examine a comprehensive sample of these listed private equity (LPE) vehicles. This setting

2

allows me to shed new light on topics that are actively debated in the PE literature, such as performance,

performance persistence, economies of scale, transaction‐based proxies for performance, and syndication.

Furthermore, I make several significant findings that are novel in their own right, including the association between

performance and a range of financial characteristics that are not typically observable for traditional funds, such as

asset–turnover ratio, expense ratio, debt ratio, and transaction volume. Finally, I leverage my unique data set to

examine the predictability of fund performance using contemporaneous and lagged transaction data.

LPE has several attractive features for PE researchers. First, as LPEs are publicly listed entities traded on

international stock markets, financial data are readily available. The stock prices and financial history of the LPEs

that I study are accessible via the standard databases used in financial research. Second, LPE benefits from the

assumption that markets are efficient. PE assets are illiquid and difficult to value; therefore, estimates of the net

asset values (NAVs) provided by PE fund managers can be subjective, and there is evidence that some traditional PE

firms may manipulate their NAVs, especially when trying to raise new funds (Brown et al., 2016). LPE share prices,

in contrast, are an unbiased reflection of the market's best estimates of the true value of the LPEs' assets at any

given time. Third, LPE is well suited to studies of performance and performance persistence, as the complete

unbroken performance history can be observed for each LPE. Studies of performance persistence using 10‐year PE

funds depend on having performance data for the entire sequence of funds raised by a GP. But LPs may choose not

to invest in every fund raised by a GP, or GPs may choose not to disclose the performance of every fund they raise.

Therefore, the performance data they provide may have gaps, which could lead to biased estimates of persistence.

This discontinuity issue potentially affects all data sets that rely on investment data sourced from LPs and GPs,

including Burgiss, Preqin, and private data sets provided directly to researchers by LPs.

My first main finding is that LPE performance varies considerably across investment styles and time. An

annually rebalanced, value‐weighted portfolio of mezzanine LPEs has the best raw performance, returning 510%

over the full sample period (2000–2019), 189 percentage points more than the 321% returned by the Russell 2000

index. Mezzanine LPEs also beat the S&P 500 by 295 points, and the MSCI Europe SMID by 242 points. All other

LPE styles underperform the three market indices: Buyout and growth LPEs each underperform the S&P 500 by

about 40 points, funds of funds (FoFs) underperform by 95 points, and VC LPEs underperform by 308 points.

Jegadeesh et al. (2015) estimate the benchmark‐adjusted returns for a sample of buyout, VC, and FoF LPEs, and

find that these LPEs earn benchmark‐adjusted returns that are not significantly different from zero and have betas

close to 1. My findings are in line with theirs for buyout and FoF LPEs, but for venture capital (VC), I find negative

and significant benchmark‐adjusted returns. Hochberg et al. (2014) argue that skill is relatively rare among 10‐year

VC fund managers as the majority of these fund managers fail to raise a follow‐on fund, which the authors attribute

to investors discovering that the manager has no skill. Thus, my finding of significant underperformance by VC LPEs

is consistent with Hochberg et al. (2014). Harford and Kolasinski (2014) test the hypothesis that the reputedly high

returns to leveraged buyout (LBO) sponsors are a result of wealth transfers from debt holders who finance PE deals,

and they conclude that this is not generally the case. My findings of outperformance by mezzanine LPEs (which

1

I use the terms “traditional fund”and “10‐year fund”interchangeably to refer to fixed‐life private equity partnership funds where the fund managers are

known as general partners (GPs) and the fund investors are known as limited partners (LPs).

2

My article proceeds under the assumption that inferences about LPE performance can yield useful insights about the private equity asset class in general,

and about permanent private capital funds in particular. The advantages and disadvantages of the data and approach are discussed in detail in the next

section.

340

|

JOURNAL OF FINANCIAL RESEARCH

provide debt for PE deals) seem to be consistent with Harford and Kolasinski's (2014) conclusions. Looking at

buyout funds, Phalippou (2020) argues that measures of performance (especially fund internal rates of return [IRRs])

prominently reported by traditional PE buyout fund managers may give the impression that net returns to PE are

higher than they actually are. Instead, he argues that other measures of PE performance (such as net fund multiple

and public market equivalent) show that actual net returns to PE are closer to those of relevant public indices. My

results are consistent with Phalippou (2020) in that the buyout LPEs in my sample do not significantly outperform

the European small‐midcap index (MSCI Europe SMID), which seems most relevant for these LPEs.

Next, I show that buyout LPE performance is persistent, echoing the results of several studies of 10‐year

buyout funds.

3

Sorting LPEs by benchmark‐adjusted return into portfolios using a 12‐month formation period and

12‐month holding period, the winning portfolio of buyout LPEs achieves benchmark‐adjusted returns that are

higher than the losing portfolio by 0.49% per month (about 6% per year). Buyout fund persistence was strongest

before 2010 and has declined in recent years. These findings are in line with prior results for traditional funds, with

the advantage that my LPE data set avoids the potential discontinuity issues discussed earlier.

Diseconomies of scale are a feature of open‐end equity mutual funds (e.g., Zhu, 2018), but little evidence of

diseconomies of scale for PE funds has been found (Harris, Jenkinson, Kaplan, 2014, 2014; Kaplan & Schoar, 2005;

Lopez‐de‐Silanes et al., 2015). For my LPE sample, I find positive economies of scale for buyout LPEs in that larger

LPEs earn higher benchmark‐adjusted returns. For the other LPE styles, economies of scale are positive when

controls are excluded but not significantly different from zero when controls are included. I also find that other LPE

financial characteristics are significantly associated with LPE performance. LPEs with higher asset–turnover ratios

have consistently higher performance, echoing the study of open‐end equity funds by P´astor et al. (2017) who find

that funds that trade more generate higher returns. On the other hand, higher expense ratios are consistently

associated with lower LPE performance. Following Berk and van Binsbergen (2015), I examine the after‐fee value

added generated by LPEs, and apart from VC LPEs, which have significantly negative value added, I find that LPEs

generate net value added that is not significantly different from zero, suggesting that most LPE managers capture in

fees most of the rents they generate. Only mezzanine and FoF LPEs have (weak) positive net value added.

Given the evidence that I find of outperformance by mezzanine LPEs and performance persistence by buyout

LPEs, I next examine the characteristics of deals and exits

4

sponsored by these LPEs. I first consider whether the

types of targets and transaction styles selected by LPEs are different from those of traditional funds. Using a probit

analysis of a large sample of transactions by LPEs and traditional funds, I find that LPEs are significantly more likely

than traditional funds to invest in secondary buyouts (SBOs) and make cross‐border

5

deals. Using imputed

transaction values, I find that LPE deal values are slightly lower than those of traditional funds, but I also find that

LPE exit values are not significantly different from those of traditional funds. Geographically, mezzanine LPEs are

significantly more likely to invest in targets located in the United States, and buyout LPEs are significantly less likely

to do so. This geographic split reflects that there are few US‐domiciled buyout LPEs due to US regulatory

restrictions and that most mezzanine LPEs in my sample are US‐domiciled business development companies (BDCs).

Buyout LPEs are more likely to invest in solo deals but mezzanine LPEs are more likely to invest in syndicate deals.

3

Kaplan and Schoar (2005) find evidence of significant heterogeneity in performance across PE funds and find that persistence was strong for VC and

buyout funds raised in the 1980s and 1990s. Robinson and Sensoy (2011) obtain similar results for a sample of buyout funds, again raised largely in the

1980s and 1990s. Chung (2012) studies buyout and VC funds raised through 2000 and finds less persistence than the other papers. Harris, Jenkinson,

et al. (2014) find that PE persistence for buyout and VC funds was strong pre‐2000, unchanged for VC funds post‐2000, and weaker for buyout funds

post‐2000 especially at the upper end of the performance spectrum. Braun et al. (2017) also show that buyout PE firm deal‐level performance is persistent

but that this persistence has declined post‐2000. They argue that this decline is due to increased competition for deals among PE firms. Ewens and

Rhodes‐Kropf (2015) find evidence of persistence at the VC firm partner level. Korteweg and Sorensen (2017) find a large amount of long‐term PE

persistence, which they believe reflects the average outperformance of more skilled PE firms, but they conclude it is difficult for investors to separate

skilled PE firms from PE firms that are lucky.

4

I refer to target acquisitions as deals, and target disposals as exits.

5

I use the term “cross‐border deal”to refer to a deal where at least one of the acquiring PE sponsors is domiciled in a different country from the target

company. The term “cross‐border exit”refers to an exit where the acquiror is domiciled in a different country from the target company.

PERMANENT PRIVATE EQUITY

|

341