So You Think You Can Dance? Lessons from the U.S. Private Equity Bubble
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
So You Think You Can Dance? Lessons from the U.S. Private
Equity Bubble
Catherine J. Turco, Ezra W. Zuckerman
Massachusetts Institute of Technology
Abstract: This article develops a sociologically informed approach to market bubbles by integrating insights from
financial-economic theory with the concepts of voice and dissimulation from other cases of distorted valuation studied
by sociologists (e.g., witch hunts, unpopular norms, and support for authoritarian regimes). It draws on unique data—
longitudinal interviews with private equity market participants during and after that market’s mid-2000s bubble—to test
key implications of two existing theories of bubbles and to move beyond both. In doing so, the article suggests a crucial
revision to the behavioral finance agenda, wherein bubbles may pertain less to the cognitive errors individuals make when
estimating asset values and more to the sociological and institutionally driven challenge of how to interpret complex
social and competitive environments.
Keywords: financial markets; bubbles; valuation; private equity
Editor(s): Jesper Sørensen, Olav Sorenson; Received: September 16, 2013; Accepted: October 27, 2013; Published: March 24, 2014
Citation: Turco, Catherine J., and Ezra W. Zuckerman. 2014. “So You Think You Can Dance? Lessons from the U.S. Private Equity Bubble.” Sociological
Science 1: 81-101. DOI: 10.15195/v1.a7
Copyright: c 2014 Turco and Zuckerman. This open-access article has been published and distributed under a Creative Commons Attribution License,
which allows unrestricted use, distribution and reproduction, in any form, as long as the original author and source have been credited.
vexing question for both social scientists and ticular community, thus limiting their use of exit,
A policy makers is why public valuations (the
majority’s publicly enacted preferences) often be-
and voicing dissent may be socially risky, even if
formally permissible.
come wildly disjointed from objective conditions. From this perspective, though, financial mar-
Well-known examples include moral panics like ket bubbles are hard to explain. After all, finan-
witch hunts, in which fantastical beliefs are widely cial markets are defined by vehicles for exit and
endorsed; the cult of personality in authoritar- voice, with such vehicles eliminating distortions
ian regimes, where millions of people proclaim in prevailing public valuations (prices). Consider
allegiance to a cruel leader or bankrupt ideology; the stock market. If investors believe stocks are
norms that are broadly enforced despite question- overpriced, they can exit by selling their shares,
able social value; and financial market bubbles, thus lowering prices. Also, investors can publicly
when prices greatly exceed valuations justified by question the inflated prices with no attendant
the economic fundamentals. social risk, both informally and through various
A common account of such episodes is that forms of arbitrage. Because such arbitrage gen-
people have succumbed to mass hysteria or delu- erates high returns when skeptical investors are
sion. However, sociological and political science right, significant gaps between price and value
studies highlight mechanisms of exit and voice should never endure. Accordingly, the orthodox
(Hirschman 1970). That is, support for author- financial-economic view of bubbles—as summa-
itarian regimes often depends less on citizens’ rized in the efficient market hypothesis (EMH)—
delusion about how the regime governs and more argues that the mechanisms supporting exit and
on restrictions on speech and assembly that si- voice in markets are so powerful, and the incen-
lence their dissent, and on social pressure that tives for using them so high, that bubbles are
motivates them to dissimulate (publicly feign al- effectively impossible (Garber 2000).
legiance) (Wedeen 1999). Similar logic holds in Yet bubbles do occur. Examples include Dutch
less extreme cases like moral panics and ques- tulips in the 1630s, shares of the South Sea Com-
tionable norms (Centola, Willer, and Macy 2005). pany in 1720, the U.S. stock market of the 1920s,
There, actors may be highly committed to a par- Internet stocks in the late 1990s, U.S. housing
sociological science | www.sociologicalscience.com 81 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
in the 2000s, and the subject of this article— gies that dissenting investors adopt and that fuel
the mid-2000s U.S. private equity bubble, when bubbles.
asset prices rose far above historic levels. Echo-
ing accounts of other distorted public valuations, Silenced Voice through Rational
heterodox scholars have adopted one of two ap-
Sitting
proaches when explaining these phenomena.
During a bubble, short selling is the key way skep-
tical investors can dissent from prevailing public
Collective Delusion valuations and correct market distortions. It is
the market equivalent of standing in the common
The most prominent heterodox approach assumes
square and proclaiming that the authoritarian
that bubbles inflate when a majority of investors
regime is bankrupt or the witch hunt overblown.
become collectively deluded about asset values.1
Specifically, it involves an investor borrowing a
By some accounts, investors are driven to un-
security the investor thinks is overpriced and
reasonable estimates of value when they con-
selling it at that inflated price, then buying it
vince themselves that “this time is different”—
back later after the price has fallen to repay the
that inflated prices make sense because of fun-
original owner and keep the profit. When ortho-
damental economic changes invalidating “the old
dox theory assumes that bubbles are impossible
rules of financial valuation” (Reinhart and Rogoff
because price–value distortions are immediately
2009:xxxiv). Some argue that inflated valuations
eliminated, it explicitly assumes this sort of arbi-
are especially likely when investors can obtain
trage is unrestricted and riskless.
capital (in particular credit) cheaply because in-
In reality, short sellers risk massive losses if
vestors often ignore the cost of financing when
prices continue to rise before correcting (De Long
estimating asset values (Minsky 1975, 1992). In
et al. 1990a; Shleifer and Vishny 1997). Risk
general, behavioral economists focus on cognitive
aside, short selling opportunities are often quite
biases that lead investors to make mathemati-
limited too. For example, the small float of In-
cal errors or factual mistakes when estimating
ternet stocks prevented dissenters from shorting
asset values (Lamont and Thaler 2003; Rashes
the 1990s bubble (Ofek and Richardson 2003),
2001), whereas sociologists emphasize investors’
and before 2006, there were no institutional vehi-
reliance on flawed theories of value that render
cles whatsoever for shorting the U.S. real estate
them oblivious to major mispricings (MacKenzie
market (Gorton 2008; Zuckerman 2009).
2011; Zuckerman and Rao 2004). Yet whether
Accordingly, rational sitting models observe
ascribed to “animal spirits” (Akerlof and Shiller
that when prices rise, skeptical investors who
2009), “euphoria” (Kindelberger [1978] 2005:13),
would otherwise short sell are often unwilling to
a “fieldwide delusion” (Fligstein and Goldstein
(because of risk) or unable to (because of market
2010:34), or a “miasma of irrationality” (Pozner,
conditions), and so instead they sit on the side-
Stimmler, and Hirsch 2010:5), all such accounts
lines (Miller 1977; Zuckerman 2012b). This cedes
suggest that bubbles are driven by widespread
the market to more optimistic investors, and bub-
delusion about asset values.
bles can inflate unchallenged. A key implication—
and one that distinguishes this perspective from
Institutionalist Perspective collective delusion accounts—is that significant
private dissent can be contemporaneous with a
In contrast, a second approach suggests that bub- bubble: bubbles may be driven not by a deluded
bles can form without widespread delusion. In- majority but by a deluded minority and silenced
vestors may recognize a mispricing, but institu- majority.
tional conditions prevent them from voicing dis-
sent and correcting the distortion. Two variants
of this perspective propose two distinct strate- Dissimulation through Optimal
1 By “majority of investors,” we mean the majority of
Dancing
capital in the market. Sharing this expectation of widespread dissent,
sociological science | www.sociologicalscience.com 82 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
dancing models suggest that dissenters them- ist perspective seems compelling, it has received
selves may fuel a bubble by participating in—not little direct empirical validation. In fact, we find
sitting out of—the inflated market (Harrison and evidence of widespread investor recognition dur-
Kreps 1978). The logic comes originally from ing the PE bubble that asset prices far exceeded
Keynes’s ([1936] 1960) observation that in liquid any reasonable estimate of fundamental value,
markets—where investors can move in and out and this constitutes the first direct evidence that
of positions quickly and without affecting prices— a bubble can persist despite significant private
paying inflated prices may be rational if investors dissent from prevailing prices. While collective
expect to resell at even higher prices. This strat- delusion about value may drive other bubbles,
egy is the market equivalent of dissimulation in this test strongly suggests that bubbles do not
other cases of distorted valuation: investors’ be- depend on such delusion.
havior suggests they endorse the public valuation, Second, we test institutionalist theories’ com-
yet they privately dissent from it. mon prediction about investor behavior in illiq-
Investors often call such dissimulation “danc- uid markets. Both rational sitting and optimal
ing until the music stops,” and we label this set dancing models assume that investors who are
of related theories optimal dancing models be- sophisticated about value (who know prices are
cause they portray dancing as an optimal strategy, inflated) are also sophisticated in reading the
where dancers profit from their speculation, exit- market environment so as to know whether to sit
ing the market just before it crashes (e.g., Abreu or to dance—and, if they dance, that they know
and Brunnermeier 2003; Brunnermeier and Nagel how to time their exit before a crash. There is
2004; De Long et al. 1990b; Temin and Voth no precedent, then, for our finding that in the
2004). illiquid PE market, most dissenters danced. Post-
Crucially, though, dancing is only predicted bubble evidence we present indicates this was a
in liquid markets. Keynes ([1936] 1960:149–54) flawed investment strategy and the reason the
and all modern optimal dancing models assume PE bubble inflated to such heights.
this necessary condition, for without a high level These findings lay the groundwork for a pro-
of liquidity, dancers run the risk of holding over- posed theoretical revision. We argue that bub-
priced assets when the “music stops,” and this risk bles can inflate even when there is no collective
outweighs the potential benefits of riding the bub- delusion about value but, instead, when there is
ble. In illiquid markets, optimal dancing models collective error in the investment strategies pur-
reduce to rational sitting models: investors who sued. Such errors are unanticipated by existing
see that the market is in a bubble are expected theory, and we draw on in-depth qualitative data
to withdraw. from the PE market to probe their nature. The
data reveal how misperceptions about market
liquidity and peers’ strategies can make dancing
Agenda: Testing and seem viable even when it is not, and our analysis
Developing Heterodox Theory moves beyond institutional restrictions on voice
to identify the crucial importance of institutional
This article begins by testing key implications of restrictions on market visibility as well.
the heterodox approaches reviewed earlier against Taken together, this article’s findings suggest
unique data—longitudinal interviews with pri- that the errors fueling bubbles may pertain less to
vate equity (PE) participants during and after the cognitive and behavioral challenges of ascer-
that market’s mid-2000s bubble. First, we test taining asset values and more to the sociological
competing predictions about investor sentiment and institutionally driven challenge of how to
during the PE bubble. The collective delusion interpret one’s social and competitive landscape.
perspective expects optimistic investors to pre- This calls for a revision to the current behav-
dominate, whereas the institutionalist approach ioral finance agenda and offers a sociologically
expects significant private dissent. This test is informed path forward for research on financial
important because even though the institutional- markets.
sociological science | www.sociologicalscience.com 83 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
The Private Equity Market chase multiples to a company’s recent operating
profit, whereby a dollar of operating profit is
There are approximately 600 PE firms in the worth the same across time and within industry.
United States.2 Together they possess more than PE prices typically reflect multiples of earnings
$1 trillion in capital and own and control compa- before interest, tax, depreciation, and amortiza-
nies that employ more than 8 million American tion (EBITDA)—that is, the price, or purchase
workers.3 Although the few largest firms employ multiple, for the investment is the total capital
100 or more investors, most have staffs of 20 or used to purchase the company (in equity and
fewer. debt) divided by the company’s operating cash
PE firms purchase companies with the goal flow.
of reselling them at a profit. Purchases are struc-
tured as leveraged buyouts (LBOs), meaning they
are financed using not only equity but also debt
The Mid-2000s Bubble
raised from third-party lenders (e.g., commercial
or investment banks).4 The debt is secured by The earliest PE firms were founded in the late
the assets of the acquired company. The equity is 1960s and 1970s, and the market experienced
drawn from PE funds—10-year limited partner- its first bubble in the 1980s. From $1 billion
ships for which the PE firm serves as general part- in transactions in 1980, the market soared to
ner and its capital providers serve as limited part- more than $60 billion in transactions by 1988,
ners (LPs). LPs are generally large institutions, then crashed, contracting to $4 billion by 1990
including public and corporate pension funds, (Kaplan and Stein 1993). During the 1990s and
university and foundation endowments, and in- early 2000s, the industry retrenched, then grew
surance companies. When a PE firm raises a fund, slowly. In the mid-2000s, an even larger bubble
LPs commit a set amount of capital. The PE emerged.
firm is then responsible for investing the fund’s Figure 1 shows the total value of all U.S. LBO
capital. As it finds companies to purchase over deals from 1990 to 2012. Between 2005 and 2007,
the ensuing years, it calls down LPs’ capital com- PE firms invested approximately $960 billion,
mitments to finance the purchases. Upon selling far more than they had invested over the entire
a company, the PE firm returns to LPs the orig- previous decade. This run-up was characterized
inal capital plus 80 percent of any profit, and by an unprecedented number and size of LBOs.
retains 20 percent (called carried interest). The median LBO deal averaged $51 million in
When evaluating prospective investments, PE the 10 years prior to 2005 but grew to $59 million
investors engage in months-long due diligence to in 2005 and $100 million by the bubble’s 2007
value a target company. Two common methods peak. Importantly, these larger deals reflected
of valuation are (1) a discounted cash flow analy- the fact that PE investors were paying higher
sis, whereby a company is worth its future cash prices for the same fundamentals than they had
flows discounted by time and risk, and (2) the historically. For the 5 and 10 years prior to 2005,
application of historical, industry-specific pur- median purchase multiples in PE averaged 6.6
2 Unless otherwise specified, quantitative data on the times EBITDA, but they rose considerably during
PE market concern U.S., leveraged-buyout-focused PE the bubble, reaching a record high of 11.4 times
firms and were obtained from Pitchbook, a commercial EBITDA in Q4 2007 and Q1 2008.
data set focused on this market. Complete, reliable
data on the PE market are notoriously difficult to ob-
With the broader economic collapse in 2008
tain (Stucke 2011), but comparison of alternative data and 2009, the PE bubble popped and the market
sources (based on Harris, Jenkinson, and Kaplan [2011]) crashed. By 2009, the industry had returned to
suggested that Pitchbook offered at least as complete its early-2000s size, with only $58 billion in LBOs,
data as alternative sources. We extensively cross-checked
Pitchbook’s data against other sources.
a median deal size of $40 million, and median
3 Employment data from Private Equity purchase multiples of 5.9 times EBITDA. Since
Growth Council website, accessed May 7, 2012: the depths of the downturn, multiples and deal
http://www.pegcc.org/education/pe-by-the-numbers/. activity have increased somewhat, but the PE
4 This section draws from Kaplan and Stromberg
(2009:123–30) and Tuck (2003); also industry reports, market remains far below its 2007 peak. Today
trade press, and our extensive interviews. there is widespread consensus among analysts and
sociological science | www.sociologicalscience.com 84 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
Figure 1: Aggregate LBO Deal Value by Year
market participants that this period constituted in the cost of capital. However, the qualitative
a bubble of historic proportions.5 data we present later cannot be reconciled with
Both to understand how this happened and this. For now, it is sufficient to note that such
to develop more robust general theory, we test an account is inconsistent with the fact that mid-
existing theories of bubbles against evidence from 2000s PE investments have performed consider-
PE. Several conditions make this an ideal setting ably worse than those of other times: estimated
to pursue this agenda. average returns for PE funds raised during the
bubble are 52 percent of the historical average.6
Conditions Suggesting Widespread In fact, over the industry’s history, prices
have risen and returns have suffered during pe-
Delusion
riods of cheap and abundant capital, suggest-
Collective delusion is a plausible explanation for ing bubbles are particularly likely at just these
this bubble given capital market dynamics at the moments (Kaplan and Stein 1993; Kaplan and
time. Just as the first PE bubble occurred during Schoar 2005; Axelson, Jenkinson, and Strömberg
the 1980s’ junk bond craze, this one occurred 2013; Robinson and Sensoy 2011a). The surging
during an unprecedented surge in the credit mar- capital markets, combined with poor investment
kets. In the early 2000s, lenders began syndi- returns, specifically suggest PE may have experi-
cating a portion of LBO debt, packaging it into enced the sort of capital-fueled collective delusion
collateralized loan obligations to trade in the sec- bubble Kindelberger (1978) described—that is,
ondary market. This shifted originating lenders’ perhaps PE investors unwittingly paid inflated
incentives and encouraged more aggressive lend- 6 At the time of this writing, current Pitchbook esti-
ing practices, meaning PE investors could ac- mates of the median internal rate of return (IRR) for PE
cess more debt at cheaper rates than ever before funds raised from 2005 to 2007 average 8.37 percent, com-
(Acharya, Franks, and Servaes 2007). Concur- pared to 16.07 percent for funds raised from 1980 to 2004.
This may overstate returns during the bubble because
rently, equity capital from LPs also spiked (Figure IRR calculations for those years are based, in part, on PE
2). investors’ own estimates of as-yet unrealized returns (i.e.,
Orthodox theory would argue that price in- companies bought during the bubble but not yet resold).
creases merely reflected these exogenous changes A method of calculating PE returns that avoids this issue
is the “realization multiple,” that is, the capital actually
5 In our interviews in 2010 and 2011, PE investors returned from investments thus far divided by the total
unanimously reported that the mid-2000s had been a initial capital invested. For funds raised from 1994 to
massive bubble. Numerous industry reports and press 2004, the median realization multiple five years after a
articles discuss the unprecedented run-up in PE during fund was raised averaged 0.56; for funds raised in 2005 to
those years (e.g., Pitchbook Data Inc. 2011). 2007, it is 0.23.
sociological science | www.sociologicalscience.com 85 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
Figure 2: Limited Partner Capital Commitments to Private Equity
prices because abundant debt and equity blinded For example, PE investors are required to
them to the actual cost of this capital. co-invest alongside LPs, with at least 1 percent
A contemporaneous market development may of any PE fund’s capital coming from its own
have reinforced such value delusion. Historically, personal capital (Kaplan and Strömberg 2009).
PE firms found LBO targets through proprietary They also share in their investments’ long-term
networks. By the 2000s, however, target compa- profits through the 20 percent carried interest.
nies were increasingly retaining investment banks This carried interest is itself typically contingent
to run formalized auctions in which PE firms on factors that further motivate them to pay fair
bid against one another. Auctions often induce prices in pursuit of long-term returns—for ex-
participants to overpay (Thaler 1988), and be- ample, LPs are often guaranteed a “hurdle rate,”
cause investment banks receive a percentage of meaning PE investors receive no carried interest
the winning bid, they have an incentive to elicit until LPs reach a preset return on their capital;
the highest valuations possible regardless of fun- and PE firms face “clawback” provisions allowing
damentals (cf. Ho 2009; Perrow 2010). LPs to recall carried interest from earlier deals
if later investments perform poorly (Metrick and
Yasuda 2010). Given this set of incentives, insti-
Conditions Suggesting Widespread tutionalist theories that explain bubbles in the
Dissent absence of widespread value delusion seem like
But could a majority of PE investors really have promising candidates for explaining the PE bub-
been so deluded about asset values? While the ble.
untrained “noise” traders driving many collec-
tive delusion stories may fall prey to behavioral
biases and collective euphoria (see Akerlof and Conditions Leading to Testable
Shiller 2009; Barberis and Thaler 2005; Black Predictions
1986), PE is composed entirely of the type of
investors and incentives thought to rationalize Though compelling, institutionalist theories have
markets. Specifically, both PE firms and their received little direct empirical testing or valida-
LPs are sophisticated institutional investors, and tion, and collective delusion theories remain the
the market has been explicitly structured to mo- most prominent account of bubbles. Two condi-
tivate PE investors to overcome behavioral biases tions, however, make PE ideal for testing institu-
and value target assets accurately (Jensen 1989). tionalist predictions.
sociological science | www.sociologicalscience.com 86 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
Short selling constraints. Testing institution- holding periods, mean investors cannot expect to
alist theories requires a market with short selling sell assets readily, any time they wish.
constraints, the mechanism assumed to silence Consequently, we should not expect PE in-
even widespread dissent. PE certainly meets this vestors to adopt a speculative orientation wherein
condition: short selling is not just limited be- they determine what price to pay today based
cause of costliness or riskiness (as it is, say, in the on where they expect prices to move in the short
public stock market) but rather it is outright im- term. Because they cannot liquidate holdings
possible here. The assets being traded are private quickly, PE investors should be willing to pay
companies, so there are no freely traded securi- only what they estimate the asset’s intrinsic value
ties to short. Moreover, the entire investment to be. In short, given PE’s illiquidity, institution-
thesis is to go “long,” not short, and LPs contrac- alist theories carry a clear common prediction in
tually bind PE firms to limit their investment this market: dissenters should sit out of a bubble,
activities to LBO-like transactions. The strict not dance.
mandate is to use LP capital to purchase compa- This hypothesis is compelling, but given the
nies, increase their value by driving operational bubble and market conditions just described,
and strategic improvements, then resell them at a there is reason to doubt it will be supported in PE.
higher price. Finally, just as PE investors cannot If PE investors are sophisticated and motivated
short the market, neither can industry outsiders to recognize when prices have risen above funda-
short the equity of privately owned companies.7 mental value, and if illiquidity makes dancing a
Because few PE firms are publicly traded and poor strategy, then we should never see bubbles
only a portion of their capital comes from the in this market. Yet PE experienced a massive
public market, there are insufficient freely traded bubble in the mid-2000s. This implies the market
shares for outsiders to exercise an effective short was dominated by one or two kinds of investors
selling strategy. not easily reconciled with existing institutionalist
The complete absence of short selling means theory and intuition: (1) those who are actually
we can test theoretical predictions about investor deluded into overestimating value despite their
sentiment during the PE bubble: institutionalist sophistication and incentives and/or (2) those
theories predict considerable silent dissent on ac- who recognize that prices are inflated, yet opt to
count of the restrictions on voice, whereas collec- dance even though it is a poor strategy. Given
tive delusion theories expect widespread delusion. this, we will not only test the predictions of col-
Market illiquidity. Because PE is a highly lective delusion and institutionalist accounts for
illiquid market, we can also test institutional- PE but also use our data on investor sentiment
ist predictions about investor behavior during and strategies to move beyond these existing ac-
the bubble. Entire companies—not shares on counts.
an exchange—are bought and sold in this mar-
ket. PE firms hold investments for an average
of six years (Kaplan and Strömberg 2009), and, Methods
when they finally do exit an investment, they
This agenda necessitates capturing investors’ pri-
must sell the company to another company or
vate beliefs about value during a bubble as well
PE firm or via public offering on the stock market.
as their investment strategies. Our data and
These exits—like the initial LBO itself—are long,
methods—interviews with PE market partici-
complex transactions that take months and some-
pants (N =82) in 2005 and re-interviews with
times years to complete given financial, legal, and
those same participants in 2010 and 2011—are
regulatory issues. This, and the industry’s long
uniquely suited to obtaining such evidence.
7 Some hedge funds have considered shorting the debt
of PE-owned companies by buying credit default swap
protection on public companies that PE firms could con- The 2005 Sample: Capturing Market
ceivably take private in the future (ZeroHedge 2010). This Sentiment in a Rising Bubble
is highly indirect, and public company LBOs are a frac-
tion of all PE investments, so such trades have minimal In 2005, the lead author conducted 43 interviews
impact.
in the PE industry as part of a larger multi-year
sociological science | www.sociologicalscience.com 87 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
study on the nature and structure of PE investing. high (90 percent overall) regardless of method of
The author entered the industry with no previous access.
awareness of—nor intent to study—a PE bubble,
and no interview questions asked directly about Re-interviews in 2010 and 2011: Iden-
the issue. Rather, the interviews included several
tifying Investment Strategies during
open-ended questions, such as, “Is there anything
else I need to know about the current PE mar- a Bubble
ket as I begin to study it?” and “Are there any Following the crash, we re-interviewed the orig-
recent trends in the industry I should be aware inal sample, asking directly about respondents’
of?” In response to these and other questions, views of the mid-2000s market (which could be
respondents routinely volunteered their personal compared to their earlier statements) and about
opinions on valuation levels in the current mar- their firms’ specific investment strategies during
ket. These responses reveal market participants’ those years. We later used a range of sources (in-
beliefs about public valuations, but, crucially, dustry reports, coverage of PE investments in the
without priming the idea of a bubble. trade press, firms’ own press releases during the
PE firms were randomly selected from a strat- bubble) to validate these self-reported investment
ified sampling frame. First, the author compiled strategies.
a comprehensive list of all PE firms. Because in- The second wave contains 31 re-interviews (72
dustry participants and observers regularly iden- percent re-interview rate), where we interviewed
tify three distinct market segments—PE firms either the original respondent from 2005 or some-
with small funds (several hundred million dollars one from the original firm. During re-interviews,
or less), medium funds (several hundred million respondents consistently identified several PE
to a few billion dollars), and large funds (multi- firms and LPs known for particularly good or
billion-dollar funds, the largest of which are called bad performance during the bubble as well as a
mega-funds)—and because the dozen or so largest consulting firm that actively advised firms dur-
firms represent a sizable portion of all capital in ing the bubble. Because these institutions could
the market, it was important to stratify the list offer useful analytical traction and insight into
of firms by size to gain a representative sample dynamics during the bubble, we added them to
of both perspectives and capital in the market. the sample.
PE firms were then randomly selected from each In total, the 2010 and 2011 sample contains 39
stratum.8 interviews. A supervised research assistant con-
Interviews with PE investors (N =21 individ- ducted the majority; the lead author conducted
uals; N =18 firms)9 were then used to build a the remainder. Table 1 presents each wave’s sam-
theoretical sample of other key PE market par- ple structure.
ticipants, including LPs (N =6) that invest in
numerous PE firms; major investment banks
(N =7) that interact with hundreds of PE firms
as lenders, advisors, and auctioneers; consultants Test 1: 2005 Market Sentiment–
(N =6) who advise PE firms on due diligence; Collective Delusion or Unvoic-
and trade journalists (N =3) who cover the en-
tire industry. Given the industry’s reputation for ed Dissent?
intense secrecy, the author initially relied on key
informants and personal contacts to gain access Our first result is straightforward. Despite factors
to the randomly selected PE firms and other mar- in the capital markets and auctions that could
ket participants. However, response rates proved have fed delusion, during the bubble, there was
significant investor dissent from market prices,
8 The first author also interviewed two non-randomly
not widespread delusion that these prices re-
selected firms, where personal contacts enabled deep ac-
cess and early background discussions. flected assets’ true value.
9 Because the author conducted interviews with several In 2005, 16 of 18 (89 percent) PE investors in
investors at each of the non-randomly selected firms, the our sample reported that the market was cur-
N s for firms and PE investors differ slightly. rently in a bubble with prices inflated above
sociological science | www.sociologicalscience.com 88 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
Table 1: Qualitative Sample
2005 2010 2010
Interviews Reinterviews New
Private Equity Investorsa
Large (>$2bn) 8 6 4
Medium ($500m−$2bn) 7 5 −
Small (Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
“suspect of the market,” a consultant warned, “At as in PE, such investors should sit out until prices
some point, the market will have a correction.” correct (Miller 1977).
Another said, “We worry about what happens Our data are surprising, then. While 4 of the
when the music stops.” Note that trade journal- 16 PE investors who recognized the bubble in
ists cover hundreds of firms and cultivate sources 2005 curtailed investment activity over the next
across the entire industry, and any one invest- few years, the majority—9 of 16—participated
ment bank or consulting firm works with numer- actively in the inflating market.11 Three of the
ous PE firms. So, the fact that these respondents firms added in 2010 did the same. These investors
believed public valuations were inflated further purchased more companies, at a faster pace and
validates that skepticism was prevalent across the in larger transactions than ever, paying prices
entire market, not just our sample. Furthermore, they knew to be in excess of the companies’ un-
as early as 2005 and continuing throughout the derlying values. We classify these investors as
bubble, there was public speculation about the “dancers” because, as in optimal dancing models
existence of a PE bubble both in the national (e.g., Abreu and Brunnermeier 2003), they spoke
news and trade press as well as among industry in terms of Keynesian speculation, that is, paying
leaders at major PE conferences (e.g., Primack high prices today with the intention of selling at
2005; Dixon 2007). even higher prices tomorrow. An investor whose
The foregoing evidence constitutes the first firm danced during the bubble explained, “We’d
contribution of our article. Across all categories project out the price that we felt we could get
of respondents, there was widespread dissent from when we sold the business, and we felt we were
prevailing public valuations in 2005. Comments basing this [the price they paid] on appropriate
that prices were “crazy,” “outrageous,” “stupid,” assumptions of sales prices.”
and bound to “crash” cannot be reconciled with Such behavior was not just limited to our sam-
an orthodox account, which assumes prices accu- ple. The prevalence of dancing across PE actu-
rately reflect value conditional on the financing ally inspired what has become the most infamous
environment; nor can they be explained by a col- quote of the broader financial market excesses
lective delusion account. Instead, they represent of that time. In 2007, Chuck Prince, CEO of
the first, direct evidence in support of the key Citibank (a lender in many PE deals), told his
prediction of institutionalist accounts that mar- employees that so long as PE investors wanted
kets with restricted short selling can experience to keep dancing—as they were widely believed to
bubbles despite widespread recognition of the be doing—then Citibank should keep providing
market’s mispricing. them with debt to finance their deals: “When
the music stops in terms of liquidity, things will
get complicated. But as long as the music is
playing, you’ve got to get up and dance. We’re
Test 2: 2005 to 2007 Dissenter still dancing.” Moreover, given the prevalence of
Behavior–Rational Sitting or skepticism about price levels both in our sample
and in the broader PE market dialogue of the
Optimal Dancing? time, the sheer volume of PE investment activ-
ity in 2005 to 2007 implies considerable dancing
But what did these dissenters (investors who be-
(Figure 1).
lieved current prices were inflated) do during the
Our 2005 interviews reveal specific examples
ensuing bubble years? As noted, existing institu-
of investor dancing at the time, offering direct
tionalist theories predict that, in markets where
evidence of this strategy. For instance, the in-
there are limits to arbitrage, the degree of liq-
vestor who called prices “outrageous” and “crazy”
uidity in the market determines how dissenters
explained in the same interview that his firm was
behave. When markets are liquid, the optimal
11 Given data limitations, we were unable to classify
strategy is often to dance, buying assets at to-
day’s inflated prices to resell at higher prices prior three firms in our 2005 sample as either a sitter or a
dancer. Even if all three had sat out (which anecdotal
to the market’s correction (Abreu and Brunner- evidence suggests is not the case), dancing still would
meier 2003). However, when markets are illiquid, have been the majority strategy in our 2005 sample.
sociological science | www.sociologicalscience.com 90 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
nevertheless highly focused on “closing deals” (in this finding is consistent with the sociological
the homebuilding product sector, one of the most and political science literature that says “dissim-
overpriced). Other investors who said prices were ulation” can drive distorted valuations (Wedeen
inflated spoke similarly about being highly “pre- 1999), such behavior occurred in the absence of
occupied” (in the words of one) with finding ever the key condition existing institutionalist theory
more investments to make in the current market. assumes is necessary for observing it: liquidity.
Post-crash interviews provide further evidence As one PE investor remarked, “We are not public
of dancing and additional insight into the strat- market investors who can get in and out of the
egy. Dancers admitted knowingly paying high market by picking up the phone to a broker when
prices that were unjustified by the underlying we see things start to turn. We’re dealing with
fundamentals. An investor whose firm danced sizable companies, transformative transactions
articulated the logic, “We knew we were paying that take weeks and months and sometimes years
prices above average for historical LBOs. . . We to line up.” Given this, theory tells us that danc-
believed ‘we can justify paying x times [EBITDA] ing would be suboptimal and therefore avoided.
because we’ll get x plus times on the way out.” ’ The finding of prevalent suboptimal dancing in
Across all market segments, dancers made similar the absence of theory to explain it constitutes
admissions: our article’s second contribution.
It appears that the PE bubble would not have
We all recognized that we were get- occurred had all the investors who recognized that
ting to a point where—let’s just say prices were too high sat out of the market. We
to an unsustainable pace. . . [But] we learn from this that bubbles can be fueled by
were active. . . We made a lot of in- investors who are sophisticated in their approach
vestments then. It’s absolutely not to value but nonetheless err in the investment
the case that we said, “Let’s stay out strategies they pursue. This calls for development
of market.” (PE investor, mega firm) of new theory. To do so, we must first address
two key questions: Was dancing really suboptimal
for PE investors? If so, why did some investors
We were fully aware we were in a nevertheless adopt it?
bubble. . . We did some deals at the
peak. (PE investor, medium firm)
Was Dancing Suboptimal?
There was a collective belief that The puzzle of PE dancing would be resolved if
things were at relatively high dancing were suboptimal only for LPs, not PE
prices. . . The idea that people didn’t investors. Indeed, PE investors receive signifi-
know there was a bubble didn’t cant short-term economic benefits from dancing,
exist. People knew these were frothy irrespective of long-term returns. They collect
times. . . We definitely overpaid for transaction fees every time they buy or sell a
the [deals] we did then. (PE investor, company and annual monitoring fees from each
small firm) portfolio company (Metrick and Yasuda 2010).
For each fund, they also collect an annual man-
A trade journalist who had identified the agement fee of 2 percent of the fund’s committed
bubble in his 2005 interview reflected back in capital. Because they cannot raise a new fund un-
2011, “PE had a bubble. Prices were rising til their existing one is mostly invested, they may
unsustainably. . . I thought for the most part disregard prices and invest quickly to raise that
they [PE investors] agreed with me intellectu- next fund and collect more management fees.
ally [that prices were inflated] but ultimately it Given recent examples of financial market
didn’t change their behavior.” That it didn’t— malfeasance (cf. Perrow 2010), we must ask
that many investors continued to invest despite whether these short-term incentives alone can ex-
realizing the market was overpriced—leaves us plain PE dancing during the bubble. Careful con-
with a puzzle. Although we observe that dancing sideration suggests a more nuanced perspective,
occurred and fueled the PE bubble, and although though, and the data demonstrate that despite
sociological science | www.sociologicalscience.com 91 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
its short-term benefits, dancing was suboptimal chased during the bubble and were now valued
for PE investors, not just LPs. either below cost or below the hurdle-rate return,
First, these short-term incentives are always meaning dancers will receive no carried interest
present in PE, so they cannot alone explain why from them.
investors would knowingly pay inflated prices dur- Furthermore, investors appeared quite aware
ing some periods but not others. Also, if dancers that lasting financial success in PE is achieved
were simply exploiting LPs’ capital for short-term by raising a sequence of increasingly larger funds
personal gain during the bubble, why admit to (not by playing a one-shot game of maximizing
us in 2005 that prices were too high? They could a single fund’s fees) and that this depends upon
have obscured the malfeasance by claiming prices building a track record of strong returns for LPs.
reflected robust economic fundamentals. They further recognized that dancing was incon-
Most important, the data overwhelmingly sistent with that. Numerous dancers echoed one
demonstrate that dancing was simply not op- investor’s sentiment that “I’d not have invested as
timal for PE investors. In our 2010 and 2011 actively and would have waited it out,” preserv-
interviews, we asked PE investors what had been ing capital to invest after prices corrected. (Also,
the best strategy for a PE firm to adopt during whereas sitters took advantage of the bubble’s
the bubble, and all but one (regardless of what inflated prices to sell existing companies, dancers
strategy they had adopted) acknowledged that tended to hold theirs, believing they could sell
dancing had been wrong and that the first-best at even higher prices later. Because transaction
strategy was to have remained disciplined and fees are paid when companies are bought or sold,
sat out of the market by refusing to overpay for this further suggests dancers were not just trying
deals.12 to maximize fees.)
The issue is that although there are short- In our 2010 and 2011 interviews, dancers
term incentives to dance, there remain powerful feared a “judgment day,” “shakeout,” and “catas-
countervailing incentives to care about long-term trophic consequences” for the most egregious
investment returns. Recall that PE investors dancers, predicting that such firms would die
invest their own capital—not just LPs’—to pur- when LPs refused to invest in their subsequent
chase assets, meaning they are not simply agents. funds. Indeed, LPs significantly reduced their
In economists’ principal–agent models, having commitments postbubble (Figure 2), and dancers
such “skin in the game” (as one respondent put have had considerable difficulty raising new capi-
it) is the key mechanism for counteracting short- tal relative to sitters.14
term temptations to exploit. Moreover, PE in- To be sure, just because dancing might jeopar-
vestors share in the long-term investment profits dize LP relationships and suboptimize long-term
through carried interest, and such profit shar- returns does not mean PE investors kept these
ing can dwarf fee-driven compensation.13 Yet factors in mind. Research on hyperbolic discount-
dancing did not optimize such returns. Returns ing demonstrates that actors often struggle to
from investments made during the bubble have avoid short-term temptations even in the pres-
been poor relative to historical returns. As of ence of powerful long-term incentives. However, a
mid-2013, PE firms still held $776 billion worth bedrock principle in that literature is that invest-
of unsold assets, the majority of which were pur- ment in illiquid investments is the key mechanism
12 By contrast, Brunnermeier and Nagel (2004) and for disciplining actors to discount the future ap-
Temin and Voth (2004) describe dancing as the optimal propriately (Laibson 1997). Because PE is an
strategy in the bubbles they study. illiquid market, PE investors should have been
13 In 2011, the three founders of the Carlyle Group, a
14 Four PE firms in our sample raised new funds at
large PE firm, each made $134 million from their share
of the firm’s investment profits as well as approximately approximately the same time in 2005. Two danced; two
$70 million from returns on each founder’s own personal sat. By 2011, one dancer had given up on fund-raising
investments in the firm’s funds, compared to $250,000 in and appeared to be winding down operations. After a
salary and $3.5 million in bonus (i.e., the two fee-driven difficult fund-raising process, the other ultimately raised
compensation components). That year, each also put a new fund that was considerably smaller than the firm’s
between $97 million and $164 million of their own capi- prior fund. The two sitters (who invested the majority of
tal back into Carlyle’s funds (Zuckerman and Dezember their 2005 funds postcrash) had both raised new, larger
2012). funds.
sociological science | www.sociologicalscience.com 92 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
highly attuned to the future consequences of danc- prices they knew to be above fundamental value
ing and recognized that this was a suboptimal because they believed “all the now obviously false
long-term strategy, both for themselves and LPs. arguments” that conditions in the market had
In fact, when we examine why respondents chose permanently changed and they would be able to
to dance, we are led directly to their beliefs about sell at even higher prices later. According to an
liquidity and their LPs’ likely response. LP, dancers like this one “fundamentally believed
the world had changed” and operated with “the
Misperceptions Supporting assumption the capital markets would be there”
to support dancing.
Suboptimal Dancing
Not all dancers in our sample believed liq-
uidity had permanently changed, however. Like
Our data suggest that two misperceptions about
sitters, another third of the dancers recognized
the market environment made dancing seem rea-
that the capital markets were in a cyclical peak
sonable at the time and that dancers held these
and any increased liquidity in PE was only tempo-
mistaken beliefs, whereas sitters did not. Given
rary. In 2005, an investor from a mega-fund that
the nature of interview data, a causal link be-
danced said, “I feel like this will last as long as
tween these misperceptions and suboptimal danc-
interest rates stay low, but once the cycle turns,
ing should be regarded as preliminary. Never-
it will be harder to get deals done. . . Interest
theless, the qualitative evidence highlights their
rates aren’t low forever.”
importance in investors’ decision to dance.
The mistake of these dancers was to overes-
False sense of liquidity. During the bubble,
timate their firms’ ability to navigate it success-
one-third of the dancers in our sample seemed
fully.15 The investor quoted earlier as saying
to believe PE’s liquidity environment had perma-
that “PE will blow up” reported in 2005 that the
nently changed to now support dancing. To be
success of dancing “all depends on the credit men-
sure, PE was experiencing relatively high levels
tality,” which he readily acknowledged was tem-
of liquidity, in that debt was cheap and plentiful.
porary. Yet in his 2010 interview, he recounted
In 2005, sitters and dancers alike mentioned this
meetings where he and his colleagues had decided
as a major environmental factor. A mega-fund in-
to dance for a while longer, believing there would
vestor commented, “Capital is a commodity now,”
still be time to sell their assets at higher prices.
whereas one from a medium-sized firm observed,
Another dancer summed up the basis for such
“Capital is one of the least difficult things in this
thinking: “Everyone stayed in. . . People are going
space.”
to dance as long as the music plays. . . Everyone
Sitters were appropriately skeptical of this sit-
was suffering from the same illusions. It was a
uation, noting the debt markets were in a “crazy”
bubble mentality. . . People’s elevated view of self,
and “cyclical” moment that would not last. They
‘I know what I’m doing, I can lead us through
linked the current availability of capital to in-
the treacherous waters.” ’
creased liquidity in PE but were clear that any
False sense of safety in numbers. An addi-
increased liquidity was only as long-lived as this
tional misperception supported dancing by re-
credit cycle. Some dancers, however, interpreted
assuring investors that, even if dancing proved
the increased availability of capital as “sustain-
suboptimal, LPs were unlikely to punish them.
able,” reflecting the long-term “institutionaliza-
15 A particular strategy that facilitated such thinking
tion” of PE investing. A banker who advised
many dancers during the bubble captured this was the increased use of dividend recapitalizations (“re-
caps”) whereby a PE firm refinances the debt of an ac-
sentiment in 2005, saying, “This [liquidity] will quired company based on a higher valuation and extracts
never dry up. . . [PE] has gone from an illiquid a dividend. This strategy may have made it appear that
cowboy-type environment where a small group of it was easier to exit investments quickly at a profit, but
really bright people made money, to a very liquid recaps do not provide a level of liquidity sufficient to jus-
tify dancing (see Grant 2008). Two reasons are that (1)
environment.” Respondents’ postcrash comments recaps take at least several months to complete and (2)
provide further insight into such thinking. A they depend on an increase in the company’s valuation
dancer recalled internal meetings during the bub- from the time of acquisition, and so they depend on ei-
ther value-enhancing operational improvements or further
ble in which he and his colleagues decided to pay
price increases, both of which take time to transpire.
sociological science | www.sociologicalscience.com 93 March 2014 | Volume 1Turco and Zuckerman Lessons from the U.S. Private Equity Bubble
Specifically, dancers assumed they had “safety regardless of investments’ long-term prospects.
in numbers” (Zwiebel 1995; cf. Keynes [1936] So, although we argued earlier that fee-based
1960:158). In 2005, all of the respondents who incentives alone were insufficient for explaining
went on to dance through the bubble indicated dancing, once investors mistakenly believed that
that they believed everyone else was also partici- they had safety in numbers and/or that the liq-
pating in the bubble. Most believed everyone was uidity environment might support such a strategy,
knowingly dancing like themselves, though some the temptation to exploit short-term fees likely
suspected there were a few “fools” in the market added fuel to the inflating bubble. One sitter
too, unwittingly paying high prices. Post-crash, explained, “I guess it’s the Chuck Prince phe-
respondents observed that a “pack mentality” had nomenon. You’ve gotta keep dancing. Plus they
operated during the bubble. Such herding is rele- were rewarded handsomely to keep playing. The
vant in PE because when LPs decide whether to optimal strategy [in their minds] was to raise
invest in a new fund, they typically focus less on as much money as you could think of, spend it
a firm’s overall past performance and more on its as quickly as possible and take whacking great
relative performance.16 fees. . . But the strategy was flawed. It all came
By this logic, if there was a chance dancing crashing down.”
might be successful (because one believed the Our data suggest that the reason it “all came
liquidity environment could possibly support it), crashing down” was because dancers’ perceptions
then not dancing was risky. A dancer explained, of liquidity and safety in numbers were wrong.
“You definitely knew you were in a boom, but When the market crashed, dancers had overpriced
you’re worried about being left out.” Moreover, if investments they could not unload and dissatis-
dancing failed, the negative consequences would fied LPs, whereas sitters had preserved both cap-
be minimal and mistakes easy to justify because ital and LP favor. Had dancers read their market
everyone else’s returns would also be poor. Char- environment correctly, they would have antici-
acterizing the logic he believed some dancers used, pated this outcome and sat out, and the bubble
an LP said, “It was absolutely clear. Everyone would never have occurred. Thus, although our
knew it was insanity but you participate. If you’re data do not allow us to speculate about why sit-
going to fail, do it with a lot of company. . . Peo- ters avoided these misperceptions when dancers
ple [who danced can] say ‘We all drank the Kool- did not, these data strongly suggest that the PE
Aid. . . We all made the same mistakes. Now bubble was directly fueled by suboptimal dancing
we’re getting back to business.” ’ Consistent with and that this dancing was predicated on investors
this, an investor in our sample justified his own having misread key environmental conditions. In
firm’s dancing, saying, “We were in the same fish the next two sections, we develop the general
tank as everyone else. . . Most funds were hurt, implications of this finding for theory.
right?”
Note that with a sense of safety in numbers,
it then becomes reasonable to take advantage I. Theory Development: Mis-
of the short-term incentives to dance (i.e., fees), takes about Market Liquidity
16 Two reasons for this are as follows: (1) It is unclear
whether average PE returns exceed those of the public To appreciate this finding’s novelty, contrast it
markets and thus whether LPs should allocate as much with two well-known collective delusion theories.
capital as they do to this risky, illiquid asset class (see Reinhart and Rogoff (2009) attribute bubbles to
Harris et al. 2011; Kaplan and Schoar 2005; Robinson and
Sensoy 2011b). Nevertheless, there is general consensus
the widespread belief that “this time is different”
that returns within PE vary considerably and that, at and fundamental valuations have permanently
any given time, the best-performing funds outperform increased. But in PE, we saw investors invoke
other asset classes. Consequently, LPs try to identify “top- such language with regard to liquidity to jus-
quartile” funds, a strategy that makes relative performance
central. (2) Individual LP managers are evaluated on a tify dancing, even while recognizing that this
relative basis themselves. Overall returns often matter time was not different with regard to valuations.
less for their job security than do returns relative to “peer” Akerlof and Shiller (2009:152) contend that in-
institutions. Managers thus focus on investing in better vestors’ animal spirits make them overconfident
PE funds than their peers, again making relative PE
performance salient. in both their own estimates of value and their
sociological science | www.sociologicalscience.com 94 March 2014 | Volume 1You can also read
