Private Equity and Portfolio Companies Lessons From The Global Financial Crisis

V O LU ME 3 2
N U MB ER 3
S U M M ER 2 0 2 0
Journal of
APPLIED
CORPORATE FINANCE
I N THIS I SSU E:
Private Equity 8 Private Equity: Accomplishments and Challenges

Greg Brown, University of North Carolina; Bob Harris, University of Virginia; Tim Jenkinson,
and Public
University of Oxford; Steve Kaplan, University of Chicago; and David Robinson, Duke University
21 Private Equity and Portfolio Companies: Lessons from the
Companies
Global Financial Crisis
Shai Bernstein and Josh Lerner, Harvard University; and Filippo Mezzanotti, Northwestern University
43 Board 3.0: What the Private-Equity Governance Model Can Offer

Public Companies
Ronald J. Gilson, Columbia University and Stanford University; and Jeffrey N. Gordon,
Columbia University
52 The Growing Blessing of Unicorns: The Changing Nature of the

Market for Privately Funded Companies
Keith C. Brown and Kenneth W. Wiles, University of Texas at Austin
73 EQT: Private Equity with a Purpose

Robert G. Eccles, University of Oxford; and Therése Lennehag and Nina Nornholm, EQT AB
87 Private Equity and the COVID-19 Economic Downturn:

Opportunity for Expansion?
David Haarmeyer
92 University of Texas Roundtable on LP Perspectives on the

State of Private Equity
Panelists: Chris Halaska, Memorial Hermann Health System; Tom Tull, Employees Retirement
System of Texas; Russell Valdez, Wafra; and Shelby Wanstrath, Texas Teachers Retirement System.
Moderator: Ken Wiles, University of Texas at Austin
100 Columbia Law School Roundtable on Public Aspects of Private Equity

Panelists: Emily Mendell, International Limited Partners Association; Chris Cozzone, Bain Capital
Double Impact; and Donna Hitscherich, Columbia Business School. Moderated by Aamir Rehman,
Columbia Business School
108 A CEO’s Playbook for Creating Long-Term Value: Ten Essential

Resource Allocation Practices
Harry M. Kraemer, Jr., Northwestern University; Michael J. Mauboussin, Counterpoint Global;
and Alfred Rappaport, Northwestern University
118 A Tale of Leadership in Value Creation

Greg Milano, Fortuna Advisors
128 What Public Companies Can Learn from Private Equity Pay Plans
Stephen O’Byrne, Shareholder Value Advisors
Private Equity and Portfolio Companies:
Lessons from the Global Financial Crisis
by Shai Bernstein and Josh Lerner, Harvard University and NBER, and Filippo Mezzanotti, Northwestern University*
oes private equity contribute to financial fragility during economic crises?

D As discussed at length in the article at the top of this issue, there is extensive
literature that explores the effects of private equity ownership on corporate productivity,
product quality, employment, and related dimensions during normal times. In general, the
picture painted by these studies is one in which PE sponsors have a mixed impact on the
companies in which they invest: while the firms do well on productivity and a variety of other
operational metrics, employee headcounts shrink and wages suffer.1
A particular concern around PE ownership, and the high lever- the underlying characteristics of the companies themselves.3
age that comes with it, is whether it makes companies more Periods of high leverage have been associated with higher
vulnerable during economic downturns. One clear feature transaction prices and lower returns, suggesting a clear
of the private equity market is its intense cyclicality, with the tendency of PE investors to overleverage and overpay when
volume of PE transactions moving in highly correlated fash- access to credit is readily available. Other work suggests lower
ion with equity valuations and economic cycles. Moreover, relative rates of productivity growth by PE-backed companies
the deals transacted during market peaks differ greatly from during such boom periods, suggesting an increased emphasis
those in other periods. Steve Kaplan and Jeremy Stein, in their on the kind of financial engineering that can weaken firms.4
analysis of PE’s first major boom and bust published in 1993, These cycles in the PE market may well have broader
reported signs of “overheating” in the U.S. buyout market in economic effects that extend well beyond the markets
the late 1980s, including higher valuations (as multiples of themselves. According to data from the Preqin database,
operating cash flow), transactions in riskier industries, higher during the three years (2006-2008) leading up to the finan-
leverage, and significant reductions in the net equity contrib- cial crisis, global PE groups raised almost $2 trillion in
uted by LBO sponsors.2 equity, with each dollar typically leveraged by more than
In a much more recent (2013) study of PE market cycles $2 of debt. The United Kingdom—the focus of the study
in several countries, Ulf Axelson and colleagues showed that discussed in this article—was the most PE-driven economy
the extent of leverage in buyouts has had far more to do in the world at the time of the crisis: PE-owned assets repre-
with interest rates and general credit conditions than with sented about 11% of gross domestic product (GDP).5 In
line with these numbers, the Bank of England estimated
that PE-backed companies had issued more than 10% of
*This article draws on, while summarizing the findings of, the authors’ earlier pub- all nonfinancial U.K. corporate debt before the crisis, and
lished article, “Private Equity and Financial Fragility during the Crisis,” The Review of
Financial Studies, Vol. 32 (Fall 2019), pp. 1309-1373. The authors thank the Harvard
Business School’s Division of Research for financial support. They also thank seminar
participants at Columbia, Duke, LBS, MIT, and Northwestern—especially Efraim Ben-
3 Axelson et al. (2013).
melech, David Matsa, Sabrina Howell, Steve Kaplan, David Robinson, and Morten So-
4 Davis et al. (2019).
rensen—for helpful comments. Lerner has advised institutional investors in private eq-
5 These numbers were obtained by dividing the total private equity fundraising
uity funds, private equity groups, and governments designing policies relevant to private
between 2004 and 2008, as estimated by the European Venture Capital Association and
equity.
PEREP Analytics (in the case of the United Kingdom) and Buyouts magazine (in the case
1 Davis et al. (2014, 2019). Eaton, and Howell, and Yannelis (2019) highlight these
of the United States) by GDP in 2008 (as reported by the World Bank). In both cases,
issues from another perspective.
we exclude venture capital funds.
2 Kaplan and Stein (1993).
Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 21

that as many as 20% of British private sector workers were about such firms than any outside investors. Finally, PE groups
employed by PE-backed firms in 2007.6 can—and generally do—redeploy their human capital from
Along with the poor selection and structuring of transac- undertaking new transactions to assisting with existing firms
tions during booms, the downturns that follow are likely to and generate operational improvements. As just one example,
be made worse by cutbacks in investment and employment by December 2008, the private equity fund Bain Capital had
by overleveraged PE companies.7 And in response to this shifted all of its 175 investment professionals from seeking
concern, both the European Central Bank (ECB) and the U.S. out new deals to developing and implementing action plans
Fed and other regulators have issued guidance capping the to help its portfolio companies survive the downturn.9
amount of leverage used in PE transactions, citing concerns In a study published in the Review of Financial Studies
about systematic risk. As the ECB noted when launching its in 2019, we analyzed the financial decisions and perfor-
regulations, mance of PE-backed and non-PE-backed companies in the
United Kingdom in the wake of the financial crisis to under-
The prolonged period of very low interest rates and the ensuing stand whether private equity exacerbates or instead works to
search for yield strategies have warranted specific monitoring of limit the sensitivity of its companies and hence the broader
credit quality by the ECB in general and of leveraged finance economy to cyclical downturns.10 The U.K. makes available
exposures in particular.... Both the appetite to underwrite a trans- detailed income statement and balance sheet information
action and the propensity to retain parts of the exposure have grown for almost every active company, whether public or private.
among the significant credit institutions supervised by the ECB.8 Similar financial data are not available in the United States.
The high penetration of PE in the U.K. and the severe credit
In a similar vein, the Bank of England has argued that market freeze during the 2008 crisis make it a natural setting
buyouts should be monitored for macro-prudential reasons to examine these issues.11 As illustrated in Figure 1, aggregate
because “the increased indebtedness of such companies poses investment declined by more than 20% during 2008 in the
risk to the stability of the financial system.” United Kingdom, which simultaneously experienced a sharp
Motivated by these concerns, in these pages we examine credit contraction (shown in Figure 2).
the fate of companies already financed by PE at the time of Most of the analysis in our study focused on a final sample
the Global Financial Crisis of 2008-09. We seek to understand of almost 500 companies that were backed by PE prior to the
whether these firms were more vulnerable than their peers, as financial crisis. Using a “difference-in-differences” approach,
the financial fragility hypothesis above suggests. we studied how the financial decisions and performance of the
As an alternative, we also explore the possibility that PE-backed companies were affected by the onset of the crisis
PE-backed companies are in fact resilient to downturns, and relative to those of a control group. The control group was
that the PE firms that oversee them actually play a stabilizing made up of companies operating in the same industry in 2007
role during bad times. In particular, PE-backed companies as the PE-backed companies and with similar size, leverage,
may be better positioned to obtain external funding when and profitability.12 These matching companies also had similar
financial markets are seizing up. For one thing, PE groups trends during the years leading up to the crisis in variables like
have strong ties with the banking industry that might be used investment, revenue, return on assets, equity contributions,
to expand their portfolio companies’ access to credit during and debt issuances. Such an approach was designed to enable
periods of crisis. And because PE groups raise funds from their us to explore differences attributable to PE ownership and
LPs in the form of capital commitments that are drawn down organizational structure alone, while holding the companies’
and invested over a series of years, they tend to have “deep balance sheet or investment characteristics fairly constant.
pockets” during downturns. These capital commitments allow We started by comparing the behavior of PE-backed
them to make equity investments in their own firms not only companies and their peers during the financial crisis—that is,
at times when accessing other sources of equity is generally in 2008-2009. We found that PE-backed companies reduced
challenging, but also when they are in a position to know more investments less, by spending between 5% and 6% more, than
9 Bernstein and Sheen (2016); Bernstein et al. (2016); Gompers, Mugford, and
6 Goergen, O’Sullivan, and Wood (2011). Kim (2012).
7 See Bernanke and Gertler (1990); Bernanke (1983). Giroud and Mueller (2015), 10 See the first footnote to this article.
after controlling for a broad array of other characteristics, show that more highly levered 11 Riley et al. (2014).
firms exhibited a significantly larger decline in employment during the crisis and that 12 This approach follows the methodology of Boucly, Sraer, and Thesmar (2011). As
these layoffs had important regional consequences. discussed in the RFS article, the main results are also confirmed when using a similar
8 ECB (2017), p. 2. matching procedure but excluding leverage as a matching variable.
22 Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020

Figure 1 Figure 2
Investment in United Kingdom Around the Financial Crisis Lending Growth in the United Kingdom
Around the Financial Crisis
This figure shows the quarterly business investment volume
in the United Kingdom (seasonally adjusted). Currency values This figure shows the growth rate in the stock of lending by
are as of 2013. The measure does not include expenditure on U.K. monetary financial institutions to private nonfinancial
dwellings, land and existing buildings and costs of ownership corporations (PNCF) or nonfinancial businesses. The stock of
transfer of non-produced assets. The data are available from lending is the total amount of outstanding net lending. Series
the Office of National Statistics in the United Kingdom (https:// included are PNFC M4Lx (seasonally adjusted), sterling loans
www.ons.gov.uk/economy/grossdomesticproductgdp/bulletins/ to PNFCs (seasonally adjusted), all currency loans to PNFCs
businessinvestment/quarter3julytosept2016revisedresults). (seasonally adjusted), all currency loans to nonfinancial
businesses (non-seasonally adjusted). PNFC M4Lx is the
UK levels of business investment
lending to PNFCs, which includes loans, securities, reverse
42
repos, overdrafts, and commercial paper. The other three

measures each include loans, reverse repos, and overdrafts.
40
The data are available at the official statistics of the Bank of

Billions of GBP
England (2014).
38
Lending to UK businesses
36
30
34
One-Year % Change
20
32
10
2007q1 2008q1 2009q1 2010q1

Date
0
Source: UK Office for National Statistics

−10
2004.01 2006.01 2008.01 2010.01

non-PE-backed companies did during the financial crisis—a Date
difference that was highly statistically significant. The invest- PNFC M4Lx Sterling loans to PNFCs
ment spending of the two groups was roughly the same in All currency loans to PNFCs All currency loans to non-financial businesses
the pre-crisis period, but the investment rate of the PE group Source: Bank of England–Trends in lending
diverged notably from the control group beginning in 2008.
And this divergence first becomes discernible exactly at the The idea that private equity firms can increase their
point when both aggregate investments and credit growth in portfolio companies’ access to capital during tough times
the United Kingdom began their sharp decline. is also consistent with two other findings. First, the positive
In the second part of our study, we showed that the effect on investment was particularly notable among compa-
higher investments by PE-backed companies were funded at nies that we deemed most likely to face binding financing
least in part by debt issuance that was 4% (of total assets) constraints during the crisis—companies that were smaller,
higher for PE-backed companies than their non-PE counter-
parts during the crisis—and by equity issuance that was 2%
RFS paper. Throughout the analysis, we control for firm fixed effects and thus remove
higher than their peers’. At the same time, PE-backed compa- time-invariant characteristics of the control and treatment firms. We also showed that
nies experienced a relative reduction in their cost of debt, the results are not driven by nonparallel trends in the pre-crisis period, and they are not
and a decline in their interest expense over total debt of 0.3%. affected by the addition of company controls. Second, our main results do not change
when we exclude companies whose private equity deals were management buyouts
As in the case of investment spending, these differences in (MBOs) or public-to-private transactions. Third, the results do not appear to simply re-
access to funding first showed up in 2008 and continued flect differences in attrition between PE and non-PE companies. Fourth, the results re-
main unchanged if we control for time-varying industry shocks around the crisis. Fifth
through the remainder of the period—though with varying and finally, we also confirm that the results are robust to alternative matching estimators.
levels of statistical significance.13 In particular, we find that removing leverage from the variables used to match companies
does not affect our results. Neither undertaking the matching approach in 2003, 2004,
or 2005 (well before the crisis) nor matching each PE-backed company in the year be-
13 The results are robust to a battery of checks, which are discussed in detail in the fore the PE buyout significantly changes the results.

subject to more volatile earnings, or more highly leveraged In the final part of the survey, we asked investors to
before the crisis. identify what they took to be the most important and distin-
Second, we found the increase in investment to be larger guishing features of PE investment that enabled them to assist
when the PE sponsor had more resources available at the portfolio companies during the crisis. The PE investors who
onset of the crisis to help its portfolio companies. Availabil- responded emphasized majority control, private ownership
ity of resources was measured in two ways: amount of “dry (and the associated lack of scrutiny of public equity markets),
powder”—that is, capital raised but not yet invested—at the and the long-term horizon of PE investors. Many PE investors
time of the financial crisis; and how recent the last fundraising also cited their better access to banks as providing them with
was (with more recent funds assumed to have more capital greater experience and an edge in managing high leverage and
capacity). Our study provides clear evidence that PE firms restructuring distressed debt. At the same time, most if not all
in the early years of their funds and with more dry powder our PE respondents highlighted the value and importance of
invested more resources—both financial and managerial—in their dry powder during the crisis as a source of liquidity for
their portfolio companies during the crisis. their portfolio companies.
In the third part of our study, we examined the perfor- In sum, our findings appear to provide evidence that
mance of PE-backed companies during the financial crisis. PE-backed companies, far from being more vulnerable during
We found, first of all, that PE-backed companies experienced the Global Financial Crisis, tended to be more resilient. In
greater growth in their stock of assets in the years after the what follows, we discuss our sample of U.K. PE-backed
crisis, consistent with the greater investment mentioned companies (and the control group), the design and findings
above. At the same time—and perhaps most telling—we of our empirical study of post-crisis investment and financ-
found that PE-backed companies increased their market share ing by PE-backed companies, and the consistency of those
in the industry during the crisis while maintaining the same findings with the results of our survey of partners in PE firms.
levels of profitability—measured both as EBITDA to revenue
and as net income to assets—as their peers. Finally, although Sample Companies
we found that PE-backed companies were not more likely to We collected our sample by identifying private equity deals
declare bankruptcy than their peers, they were more likely to in Capital IQ by searching for events such as “going private,”
be sold in normal (that is, non-distress) M&A deals. “leveraged buyout,” “management buyout,” and “platform.”
Fourth and last, to explore further and learn more about In so doing, we excluded “growth buyouts,” “venture capi-
the “mechanism” underlying our results, we conducted a large- tal,” and “expansion capital” investments, where investors
scale survey of more than 300 private equity investors. The generally buy a stake in the company using little or no
respondents were experienced investors, almost all of them leverage. Since our aim was to study the behavior of U.K.
at the partner level, with an average of 14 years of experience PE-backed companies around the financial crisis, we selected
in the industry. In the survey, we asked whether the operat- only firms that (1) were headquartered in the United King-
ing and financial activities of private equity investors changed dom at the time of the deal, (2) had received a PE investment
during the crisis and, if so, how. by the end of 2007, and (3) did not experience an exit by
Our survey responses suggest that PE investors were signif- the PE group by the end of 2008.14 The United Kingdom
icantly more likely to help portfolio companies with their is an ideal setting for studies of private companies since all
operating problems during the crisis and provide strategic registered limited companies are required to provide finan-
guidance. And given that PE investors were interacting more cial and income information annually to the public register,
frequently with portfolio companies during the crisis, it’s not known as the Companies House.
surprising to learn that they were spending much less time Nevertheless, the extent of the requirement to disclose
looking for new deals. financial information in the United Kingdom varies with
But, as the survey also makes clear, PE investors during the size of the company. Small (and some medium-sized)
the crisis were also more likely to get involved with the finan- companies are allowed to file abbreviated accounts.15 Since
cial restructuring of their portfolio companies, working with
bankers and lawyers to renegotiate debt obligations. As already
14 During 2008, 28 PE firms exited. The results remain unchanged if we include
noted, PE investors were also more likely to help their compa- them in the sample.
nies raise more debt financing, and perhaps to contribute 15 Since 2008, a small company is defined as meeting at least two of the three fol-
lowing criteria: total assets less than £3.26 million, annual turnover less than £6.5
more of their own equity, to maintain investment or work million, and an average number of employees fewer than 50. This group usually reports
themselves out of financial distress. only assets, revenue, and profits.

the amount of information small firms disclose to Compa- Are the U.K. Markets Representative of
nies House is quite limited, we excluded this group from our Other PE Markets?
analysis. We view the reliability of the source and its coverage How do we know that private equity investment in the United
of the remaining private firms as a key strength of our study.16 Kingdom is in any way similar or comparable to PE in other
Most of the companies in our sample thus consist of middle- countries? International comparisons of PE activity are not
sized private enterprises, for which similar financial data are straightforward. While the trade association of the European
not available in the United States.17 PE industry, Invest Europe (formerly known as the Euro-
After eliminating small firms, as defined by the Compa- pean Venture Capital and Private Equity Association) has
nies House, and financial companies and public utilities, we produced its own data since the 1980s, other national asso-
were left with an initial sample of 987 unique firms. And when ciations, including those in the United States, rely on private
we next excluded companies that did not meet minimum data companies. And these commercial data sources have often
data requirements for the matching process described below— produced inconsistent estimates using quite different meth-
namely, the firm’s 2007 industry, return on assets, capital odologies and definitions.
expenditures, asset, and leverage ratio—our sample included In the compilations reported in Figure 3, we focus on
722 companies. estimates of the activity by PE funds in general—that is,
For this set of companies, we then searched for indica- buyout, mezzanine, and growth, but not venture capital
tors of financial distress, including potential acquisitions and funds—organized by the country in which the fund is based.
bankruptcies during the crisis. To that end, we consigned to Thus, an investment by a London-based fund that is part of
an outcome called “Bad Exit” all sample companies that were a U.S. alternative investment group would be included in the
not active by 2011 and whose status, before disappearing from U.K. total.
our data set, implied the firm was in liquidation or insolvency As reported in Figure 3, during the 14-year period (from
proceedings. At the same time, we used Capital IQ to identify 2000-2013) that surrounds the financial crisis, the data for
potentially profitable exits by looking at firms involved in Europe as a whole, France, Germany, and the United States
M&A transactions from 2008 onwards. Since M&A transac- show a remarkably similar pattern in terms of four major
tions can also be responses to distress, we used an alternative variables: (1) equity value invested per year by funds based in
measure that excluded companies that were acquired but also a given country; (2) amounts invested as a share of GDP of
identified as distressed in the same period. the nation in that year (as reported by the World Bank); (3)
Finally, we collected information on the history of the PE average amounts invested in individual transactions; and (4)
investors for each portfolio company to identify when the PE the numbers of investments per year.19 While the volume of
investors raised their last fund before the crisis. As suggested activity varies across countries—with the United Kingdom
earlier, the more recent the last fundraising at the onset of the representing a disproportionate share of European activ-
crisis, the more capital and other resources were presumably ity—the patterns were, as noted, remarkably similar. More
available to the PE firm, since the PE firm had less time to specifically,
deploy capital to portfolio companies. And for each PE firm, • Virtually all the countries display run-ups in aggregate
we also constructed a measure of “dry powder,” a proxy for the transaction value until the inception of the financial crisis,
dollar amount of financial resources that the PE investors had followed by corrections afterward and recoveries in the final
available by the time of the crisis. To generate this measure, years of the sample.
we used ThomsonOne’s fundraising and investment history • In all countries, PE investments as a fraction of GDP
for PE investors for the period 2001-2007 to calculate the increased in the run-up to the crisis and declined afterward.
difference between each PE firm’s aggregate fundraisings and Following the crisis, the share of PE investments recovered,
investments.18 particularly in the United States and the United Kingdom.
• Relative to other European countries, average deal size
16 We used two data sources: Amadeus and Orbis, While both Amadeus and Orbis in the United Kingdom was significantly higher, although it
collect information from the Companies House, Amadeus generally removes firms from
the sample after a few years of inactivity. This is not the case for Orbis. Since the post- remained much smaller than transaction values in the United
financial crisis period was characterized by an increase in firm exit, using only Amadeus States. Deal sizes increased in all countries during the pre-crisis
would have generated selection concerns that could undermine the reliability of our re-
sults.
17 One limitation of this data set is that balance sheet items are always reported at tors. The results remain unchanged if we instead classify dry powder to be equal to one
the book value. if the investor that made the largest equity investments before the crisis.
18 If a portfolio company has more than one PE firm, we select the dry powder of 19 The source of the private equity data is Invest Europe, except for the U.S. data,
the investor with the highest level of dry powder to categorize the syndication of inves- which is Preqin (number of deals) and Cambridge Associates (dollar volume of deals).

Figure 3
Comparison of Private Equity Activity Across Different Markets
This figure depicts several international comparisons of private equity activity in the UK, France, Germany, Europe as a whole,
and the US. Panel A describes the aggregate equity value invested per year by funds based in a given country, Panel B describes
the amount invested as a share of GDP of the nation in that year, Panel C reports the average amount invested in a given year,
and Panel D illustrates the number of investments. The source of the private equity data is Invest Europe, except for the US data,
which is from Preqin (number of deals) and Cambridge Associates (dollar volume of deals).
Panel A: Aggregate Equity Value (€ billions) Panel B: Annual Aggregate Deal Value/GDP
Left Y-axis describe all markets, except the US that is described in the right Y-axis.
70 160 1.60%
60 140 1.40%
120 1.20%
50
100 1.00%
40
80 0.80%
30
60 0.60%
20 0.40%
40
10 20 0.20%
0 0 0.00%
02 03 04 05 06 07 08 09 10 11 12 13 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013
UK France Germany Europe US UK France Germany US Europe
Panel C: Average Deal Value (€ millions) Panel D: Number of Transactions

Left Y-axis describe all markets, except the US that is described
in the right Y-axis.
80.00 80.00
70.00 70.00
60.00 60.00
50.00 50.00
40.00 40.00
30.00 30.00
20.00 20.00
10.00 10.00
0.00 0.00
02 03 04 05 06 07 08 09 10 11 12 13 02 03 04 05 06 07 08 09 10 11 12 13
UK France Germany US Europe UK France Germany US Europe
period, while during the crisis average deal size declined signif- sales” (sales from one PE firm to another). Only 5% of the
icantly in all countries. transactions in our sample were public to private. Lending
• In all countries, there was an increase in transaction confidence to these findings, the study of French PE deals
volume up until the financial crisis, and then a decline in cited earlier reports a very similar distribution of transac-
number of investments. tion types. And a 2008 study of global PE deals suggests a
The lone exception to all this appears to have been late- similar distribution of private equity investments around the
blooming Germany, where private equity had little traction world.20 Specifically, roughly 27% of transactions in France
before the financial crisis. Although the number of invest- and 26% of the global sample were divisional buyouts.
ments increased, total PE transaction value declined in Secondary sales were slightly less frequent in France (15%)
Germany as well. and worldwide (13%). Finally, in both of these samples, the
Another interesting comparison between different PE share of public-to-private transactions was around 4.5%,
markets relates to the types of buyouts. As reported in whereas private-to-private transactions were roughly 50%
panel D of Table 1, roughly 40% of U.K. PE transactions of the sample.
were private-to-private (that is, purchases by PE firms of
privately held concerns); 30% of the deals were buyouts of
divisions from public companies; and 20% were “secondary 20 Boucly, Sraer, and Thesmar (2011); Stromberg (2008).

Table 1
Summary Statistics
Panel A reports the industry distribution at the broad industry level (1-digit SIC) for the PE sample and the whole universe of medium
and large U.K. firms, but excludes financial, insurance, regulated, or public administration. Panel B reports the summary statistics
of sample firms in 2007 across treated (PE-backed companies) and non-treated firms (non-PE companies). The last column reports
the mean difference across the two groups. Level variables are in millions of dollars. Panel C reports the 1- and the 2-year growth
as a percentage increase in the characteristics in 2007. The last column reports the mean difference across the two groups. Panel
D reports the split in terms of deal type for the final sample of PE deals. The appendix of the RFS paper provides more information
about the variable definitions. *** denotes significance at the 1% level, ** at the 5% level, and * at the 10% level.
A. Industry Distribution
Industry distribution PE sample (%) Full sample (%)
Mining 1 2
Construction 6 15
Manufacturing 32 17
Wholesale trade 12 11
Retail trade 7 6
Transportation 4 6
Services 38 44
B. Firms’ Characteristics in 2007

PE sample Matched sample
N Mean Median SD N Mean Median SD Mean diff.
Revenue (M$) 432 98.05 35.30 240.81 1527 77.64 29.86 184.49 20.41*
ROA 434 0.09 0.09 0.23 1550 0.09 0.09 0.22 0.01
Investment/asset 434 0.19 0.20 0.18 1550 0.20 0.20 0.18 -0.01
Equity contr./asset 415 -0.02 0.01 0.13 1513 -0.01 0.01 0.13 -0.01
Net debt iss./asset 415 0.09 0.10 0.23 1513 0.11 0.08 0.24 -0.01
Debt/asset 434 0.71 0.70 0.39 1550 0.69 0.67 0.39 0.02
Private ownership 434 1.00 1.00 0.00 1550 0.99 1.00 0.01 0.01
C. Firms’ Trends in 2007

PE sample Matched sample
N Mean Median SD N Mean Median SD Mean diff.
1-year growth
Revenue 423 0.37 0.18 1.34 1456 0.35 0.17 1.17 0.02
ROA 427 0.71 -0.03 5.21 1483 0.79 0.07 4.48 -0.07
Investment/asset 386 1.54 0.10 5.86 1434 1.37 0.05 5.20 0.17
Equity contr./asset 372 -0.59 0.39 15.96 1376 -0.93 0.09 13.73 0.34
Net debt iss./asset 376 2.95 0.32 15.09 1428 2.25 0.20 12.86 0.70
Debt/asset 418 0.02 -0.03 0.34 1516 0.02 -0.02 0.31 0.01
2-year growth
Revenue 393 0.56 0.33 2.08 1362 0.71 0.34 2.33 -0.15
ROA 400 1.10 0.05 8.33 1388 1.40 0.11 6.97 -0.29
Investment/asset 339 1.85 0.61 6.22 1333 2.39 0.94 6.06 -0.54
Equity contr./asset 330 0.43 1.09 23.44 1274 0.70 1.05 18.95 -0.28
Net debt iss./asset 343 3.45 0.65 18.73 1359 2.94 0.76 13.99 0.51
Debt/asset 382 0.01 -0.04 0.46 1442 0.04 -0.04 0.60 -0.03
D. Buyout type
Buyout type Percentage
Public to private buyouts 5.3
Private to private buyouts 42.8
Divisional sales 29.9
Secondary sales 20.0
Distressed buyouts 2.0

Designing Our Test Similarly, all PE-backed companies and control firms were
To understand how the crisis affected the financial and invest- under private ownership (with the exception of a few publicly
ment policies of PE-backed companies, the ideal experiment traded companies in the control group), and not in default or
would compare two identical firms during the crisis, with the insolvency proceedings at the onset of the financial crisis. What’s
only difference that one is backed by a PE firm and the other more, as discussed at considerable length in our RFS article, our
is not. As a substitute for such identical firms, we developed a estimates are consistent with parallel trends between treated
“difference-in-differences” research design that enabled us to and control groups during the period leading to the crisis,
compare PE-backed companies to a matched set of compa- which is as noted earlier a critical assumption underlying our
nies around the financial crisis. difference-in-differences design.22 Using regression analysis, we
Private equity-backed companies are clearly not a random estimated a model of the relationship between PE backing and
sample of the population: for instance, they are likely to be company investment and financing using a panel data set that,
larger and more leveraged than the average non-PE-backed by spanning the period 2004 to 2011, provided a “symmetric
firm. Therefore, the first step in the analysis was to identify window” onto the 2008 shock.23 As we saw earlier in Figure 1,
a proper control group for the set of PE-backed companies. aggregate investment in the United Kingdom declined by more
Following the study of French PE transactions just than 20% between the beginning of 2008 and mid-2009. At the
mentioned, we identified a suitable control group using a same time, credit availability experienced a sharp contraction
matching procedure for each of the 722 PE-backed companies that started in the first quarter of 2008.
in our sample. We identified a set of U.K.-based control firms
that, in 2007, were operating in the same industry and had Investment and Funding of PE-Backed Companies
similar size, leverage, and profitability. Using this procedure, We began our study by examining whether companies backed
we were able to provide matches for 434 of the 722 firms, by PE investors were more or less affected by the financial
generating a total sample of 1,984 firms. crisis than otherwise comparable companies. Noting that over-
Then, for each of these almost 2,000 companies, we all investment dropped significantly in the United Kingdom
measured “capital investment” for any given year by calcu- during the crisis period, our aim was to determine whether
lating the increase in the firm’s total assets plus reported PE-backed companies experienced even more severe declines
depreciation. We measured net equity injections as the change during the crisis.
in the firm’s book value of equity minus net income, and debt We started our analysis by looking at the changes in invest-
issuance as the increase in total liabilities, during the year. All ment by PE- and non-PE-backed companies. As reported in
of these variables were scaled by total assets. In addition, we Column 1 of Table 2, we found that PE-backed companies
measured a company’s leverage as its total liabilities to total reduced their investment less than non-PE-backed companies
assets, and its cost of debt as the ratio of its total interest around the financial crisis. As stated earlier, the PE firms saw
expenses to total debt. their investments increase by almost 6% as a percentage of
As reported in Table 1, having eliminated small businesses total assets relative to the non-PE companies in the post-crisis
and companies in the financial or regulated sectors, the major- period.24
ity of our sample companies were in either the services (38%)
or manufacturing (32%) industries. Other well-represented
22 In particular, we compare 1- and 2-year growth rates ending in 2007 for the main
industries were wholesale trade, construction, and retail. The firm characteristics considered so far (Table 1, panel C). As reported in our RFS paper (in
sample industry distribution is relatively close to that of the Figures 4, 5, and 6), we find that the differences in the growth rates between the two
groups are not significantly different from zero across all observables.
universe of U.K. private and public companies: the major 23 Our model takes the form of the following equation:
difference is that PE-backed companies tend to be more yit= αt + αi + β1 (PE firmi * Post) + θXit+ εit , (1)
where γit is an outcome variable measured for company i at time t, (αi, αt) are a set of
concentrated in manufacturing, and less represented in the company and year fixed effects PE firmi, is a dummy for the companies that are backed
construction industry and services.21 by PE investors, and Post is a dummy for the period from 2008 to 2011. For consis-
As can be seen in panel B of Table 1, the average firm in our tency, each PE-backed company and its corresponding control group enter in the sample
at the same time, which is 2004 or the year of the PE deal if after 2004.
sample was a mid-sized company with around $80 million in 24 As discussed in the RFS study, we found that the results are unchanged—in terms
revenue; and by design, the two groups of firms have very similar of both size and statistical significance—when we add the standard set of firm-level
controls (odd columns, Table 2). What’s more, since we focus on the PE treatment ef-
ROA, investment, leverage, and equity and debt issuances. fects around the crisis and therefore after the PE investments, our results do not account
for the potentially positive impact of the initial PE investment on operations and financ-
ing. Since the effects of the initial investment has been generally found to be positive
21 And, of course, both the treatment and control samples have the same industry (e.g., Kaplan 1989), our estimates may underestimate the overall effect of PE on the
distribution thanks to our matching process. portfolio companies.

Figure 4 Figure 5
Effect of PE-Backed Companies on Investment Over Time Effect of PE-Backed Companies on Equity Contributions
Over Time
This figure illustrates the change in investment separately for
both PE and non-PE companies in our sample. Specifically, the This figure illustrates the change in equity contributions sepa-
figure reports αt of the following equation: γit = αt + αi + εit. The rately for both PE and non-PE companies in our sample. Specifi-
equation is separately estimated for PE and non-PE compa- cally, the figure reports αt of the following equation: γit = αt + αi
nies. αt captures year fixed effects, and αi captures firm fixed + εit. The equation is separately estimated for PE and non-PE
effects. The year 2007 is used as the base period, and there- companies. αt captures year fixed effects, and αi captures
fore the corresponding coefficient is normalized to zero. The firm fixed effects. The year 2007 is used as base period, and
estimates are plotted with standard errors above and below the therefore the corresponding coefficient is normalized to zero. The
point estimates. Standard errors are clustered at the firm level. estimates are plotted with standard errors above and below the
point estimates. Standard errors are clustered at the firm level.
.1
.0 5
0
Equity Contribution
Investment
0
−.1
−.2
−.05
−.3
2004 2005 2006 2007 2008 2009 2010 2011

−.1
Year
2004 2005 2006 2007 2008 2009 2010 2011
PE Non-PE Year
PE Non−PE
In Figure 4, we plot the year effects estimates around the portfolio companies to access credit markets during periods
crisis—and the corresponding standard errors—separately for of turmoil.
the PE-backed companies and matched companies. As illus- As reported in Table 2, we found that increases in net
trated in the figure, both the PE-backed and the control firms equity were larger for PE-backed companies than for the
followed similar paths before the crisis: the estimates are not control group around the crisis.25 Equity contributions as
statistically different from one another and thus seem to satisfy a percentage of total assets during the financial crisis were
the parallel trends assumption. Once the crisis started, both 2% higher to PE-backed companies than for non-PE firms.
the PE-backed companies and the matched companies cut As can be seen in Figure 5, equity contributions for both
investments sharply during 2008 and 2009. Nevertheless, the classes of firms dropped significantly during the crisis. But
investment cutbacks by PE-backed companies during the crisis the fact that the decline was smaller for PE-backed compa-
years were significantly smaller than those of their non-PE nies suggests that PE funds were willing to support the
counterparts. This higher level of investment by PE-backed operations of their portfolio companies by injecting more
companies persisted in the years after the crisis. of their own equity.
What accounts for such differences in investment? One At the same time, as reported in Column 5 of Table 2,
possibility is that the PE firms help their portfolio companies PE-backed companies also experienced a relative increase in
to maintain high investment levels by expanding their access debt issuance.26 While on average, debt issuance over assets
to capital, particularly during periods of financial upheaval. declined during the financial crisis, this decline was 4% smaller
This can happen in two ways. First, PE firms have fund
commitments that are rarely abrogated and may therefore be 25 We define equity contribution by looking at the changes in equity that were not
in a better position to inject equity into the companies if explained by profit. Therefore, we cannot determine whether positive effects were due to
raising more capital or paying out fewer dividends.
access to financial markets is limited. Second, PE firms have 26 As discussed in the data section and in the appendix of the RFS paper, this is
strong ties with banks and that should make it easier for their measured as the change in total debt, scaled by assets.

Table 2
Investment and Funding Policies
This table reports the estimates of a difference-in-differences fixed effects model on the investment and funding variables. All
specifications include firm and year fixed effects. The main parameter of interest is the interaction between the Post dummy and
PE-backed company dummy variable. Odd-numbered columns contain the baseline regression, and even-numbered columns aug-
ment the baseline model with a set of firm-level controls measured before the crisis and interacted with the Post dummy. These
variables include firm size (log of revenue), growth in revenue, cash flow over assets, ROA, and leverage. In Columns 1 and 2 the
outcome is investment scaled by assets; in Columns 3 and 4 the outcome is net equity contribution over assets; in Columns 5
and 6 the outcome is the net debt issuance scaled by assets; in Columns 7 and 8 the outcome is the total leverage; and in Col-
umns 9 and 10 the outcome is average interest rate. The appendix of the RFS paper provides more information about the variable
definitions. Standard errors are clustered at the firm level. *** denotes significance at the 1% level, ** at the 5% level, and * at
the 10% level.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Investment/assets Net equity contr./assets Net debt iss./assets Leverage Interest rate
PE firm x Post 0.059*** 0.056*** 0.022*** 0.021*** 0.042*** 0.039*** 0.013 0.012 -0.003** -0.003**
(0.013) (0.013) (0.007) (0.007) (0.011) (0.011) (0.015) (0.014) (0.001) (0.001)
Year fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Firm fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Firm controls No Yes No Yes No Yes No Yes No Yes
Observations 12,456 11,910 12,469 12,003 12,903 12,274 13,205 12,553 10,222 9,831
Clusters 1,984 1,878 1,981 1,876 1,982 1,876 1,984 1,878 1,841 1,743
R-squared 0.160 0.161 0.040 0.059 0.090 0.104 0.011 0.029 0.016 0.022
for PE-backed companies. But if their overall debt issuance was Why are PE-backed companies able to raise more
greater, PE companies did not materially increase their leverage, external financing? There are several possible explanations.
as can be seen from Columns 7 and 8 in Table 2.27 Nevertheless, Since private equity firms typically own majority stakes
as reported in Columns 9 and 10 of Table 2, we found that the in their portfolio companies and control their boards, the
relative cost of debt, measured by the ratio of interest expense information and control problems that beset public company
of total debt, declined for the PE-backed companies by 0.3%.28 boards are likely to be much more manageable, thereby
Overall, then, our findings suggest that private equity reducing “frictions” that tend to limit equity investments in
groups expanded their portfolio companies’ access to capital companies. Moreover, the repeated interactions of PE firms
during the financial crisis, allowing them to invest more when with banks is likely to reduce information asymmetries,
credit markets were frozen and economic uncertainty high. establish trust, and provide additional cross-selling activities
In particular, PE groups appear to have taken advantage of for the banks. All these considerations may have eased
their fund structures and bank relationships to provide both the financial constraints on already highly leveraged
equity and debt financing to their portfolio companies, while PE-backed companies seeking more debt financing during
reducing the cost of debt.29 the crisis.
27 And the fact that the PE coefficient in this regression is positive, though insignifi-
What Are the Financial Constraints Faced by
cant and small in magnitude, is consistent with the joint increase in equity as well as PE-Backed Companies?
debt. Our findings presented thus far are consistent with the idea
28 One concern about this interpretation of our findings is that by matching on lever-
age (in addition to other variables), we may have captured (non-PE companies) that are that private equity can play an important role during finan-
somewhat unrepresentative due to their high leverage. To address this concern, we re- cial turmoil by expanding their portfolio companies’ access to
peated the main analyses using an alternative matching that does not rely on leverage,
but only on size, ROA, and industry. This matching estimator allows the two groups to capital and so relaxing the financial constraints they (and their
have different leverage ratios in the pre-crisis period. And as reported in Table 4 of the
RFS article, when we repeat our analysis with the alternative matching methodology, all
results remain unchanged. less efficiently than did banks. This interesting question poses challenging measurement
29 We are not making any claims about whether PE firms allocated capital more or issues, and, while fascinating, it is outside the scope of our paper.

Table 3
Heterogeneity Across Firms’ Financial Constraints
These tables estimate standard difference-in-differences fixed effects model and repeat the specification of Table 2 while exploring
various proxies of financing constraints in 2007. All specifications include firm and year fixed effects. In each table, the interaction
term in Columns 1 and 2 is based on firm size, and equal one if the firm is at the top quartile of firm employment versus the rest
of the sample. The interaction in Columns 3 and 4 is based on dependency on external finance, measured by RZ index (Rajan and
Zingales 1998). The interaction equals one if dependence on external finance is in the top quartile, and zero otherwise. In Columns
5 and 6, the interaction is based on firm leverage. The interaction equals one if firm leverage is at the top quartile, and zero other-
wise. Panel A reports the results using investment as an outcome; panel B uses instead debt issuance over assets as dependent vari-
able; and panel C reports the results with net equity contributions over assets. Even-numbered columns augment the baseline model
with a set of firm-level controls measured before the crisis and interacted with the Post dummy. These variables are firm size (log of
revenue), growth in revenue, cash flow over assets, ROA, and leverage. The appendix of the RFS paper provides more information
about the variable definitions. Standard errors are clustered at the firm level. *** denotes significance at the 1% level, ** at the 5%
level, and * at the 10% level.
A. Results Using Investment as an Outcome

competitors) would otherwise face. But how constrained are
(1) (2)
Investment/assets
(3) (4) (5) (6)
such companies, and are the major differences among them
PE firm x Post 0.014 0.011 0.039*** 0.032** 0.032** 0.031**
in terms of the constraints they face?
(0.020) (0.020) (0.014) (0.014) (0.013) (0.013) We used several measures as proxies for financing
Interaction. x Post -0.025* -0.016 -0.049*** -0.040*** -0.072*** -0.050*** constraints. First was firm size. Consistent with the idea
(0.013) (0.015) (0.013) (0.013) (0.016) (0.018)
that small companies are more likely to be financially
Interaction x Post x PE 0.053** 0.051** 0.056* 0.068** 0.108*** 0.101***
(0.026) (0.026) (0.029) (0.029) (0.033) (0.032) constrained, and so more sensitive to credit market shocks,
Interaction variable Small External dependence High leverage we identified the large firms in our sample as those having
Year fixed effects Yes Yes Yes Yes Yes Yes
Firm fixed effects Yes Yes No Yes Yes Yes the top quartile of employment in 2007, and classified all
Firm control
Observations
No
11,539
Yes
11,105
No
12,456
Yes
11,910
No
12,456
Yes
11,910 remaining firms as small.30 Using this measure, we found
Clusters
Adjusted R-squared
1,824
0.160
1,742
0.162
1,984
0.161
1,878
0.162
1,984
0.162
1,878
0.162
(as reported in Columns 1 and 2 of Table 3) that the
positive effect on investment was stronger for the smaller
B. Debt Issuance Over Assets As Dependent Variable
PE-controlled companies in our sample.
(1) (2)
New Debt Issuances/Assets
(3) (4) (5) (6)
Second, we found similar results when we looked at
PE firm x Post -0.004 0.003 0.021* 0.019 0.028** 0.026**
companies that operate in industries that rely more heavily
(0.019) (0.019) (0.012) (0.012) (0.011) (0.011) on external finance.31 As a rule, companies with greater
Interaction. x Post -0.015 -0.030** -0.055*** -0.037*** -0.183*** -0.110*** dependence on external finance were expected to be more
(0.014) (0.014) (0.014) (0.012) (0.017) (0.017)
affected by the financial crisis, given the dramatic decline
Interaction x Post x PE 0.055**
(0.024)
0.046*
(0.024)
0.055**
(0.026)
0.058**
(0.026)
0.072**
(0.029)
0.065**
(0.029) in credit availability, and benefited more from the financing
Interaction variable Small External dependence High leverage
support of PE. Consistent with this, our findings, reported
Year fixed effects
Firm fixed effects
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
in Columns 3 and 4 of panel A of Table 3, show that the
Firm control
Observations
No
11,891
Yes
11,400
No
12,903
Yes
12,274
No
12,903
Yes
12,274
positive effect of being backed by PE was larger for compa-
Clusters 1,823 1,741 1,982 1,876 1,982 1,876 nies in more financially dependent industries.
Adjusted R-squared 0.089 0.101 0.091 0.105 0.105 0.108
Third, we found similar results when comparing compa-
C. Results With Net Equity Contributions Over Assets
(1) (2) (3) (4) (5) (6) nies that were more leveraged when entering into the crisis.
PE firm x Post 0.035*** 0.026** 0.018** 0.015* 0.016* 0.016**
In general, such firms have less financial flexibility and higher
(0.012) (0.012) (0.009) (0.008) (0.008) (0.008)
interest payment burdens and so face greater risk of business
Interaction. x Post -0.006 0.014 0.001 -0.003 0.073*** 0.043***
(0.007) (0.008) (0.007) (0.007) (0.008) (0.007)
Interaction x Post x PE -0.016 -0.008 0.013 0.014 0.011 0.011 30 See Petersen and Rajan (1994), and Bottero, Lenzu, and Mezzanotti (2020).
(0.016) (0.015) (0.016) (0.015) (0.016) (0.016)
31 We identify these industries using the procedure (RZ index) discussed in Rajan
Interaction variable Small External dependence High leverage and Zingales (1998). In particular, we defined more financially dependent firms as those
Year fixed effects Yes Yes Yes Yes Yes Yes operating in two-digit SIC industries whose share of capital expenditure that are exter-
Firm fixed effects Yes Yes No Yes Yes Yes
nally financed was in the top quartile. In line with the literature, this measure is com-
Firm control No Yes No Yes No Yes
Observations 11,564 11,193 12,469 12,003 12,469 12,003 puted using data available from Compustat for U.S. corporations between 1980 and
Clusters 1,823 1,741 1,981 1,876 1,981 1,876 2008. In particular, for each two-digit SIC industry, we measure the RZ index as the
Adjusted R-squared 0.045 0.063 0.040 0.059 0.051 0.062 median of CAPEX minus cash flows from operations, scaled by CAPEX.

Table 4
Heterogeneity Across Funds
This table reports the estimates from a difference-in-differences fixed effects model, while exploring heterogeneity across resource
availability of PE firms backing the company. The analysis is a cross-section estimated using only the set of PE-backed companies.
High dry powder is a dummy variable equal to one if PE investors are at the top quartile for amount of dry powder at 2007, defined
based on the amount of capital raised but not invested. Note that if a portfolio company has more than one PE firm, we select the
dry powder of the investor with the highest level of dry powder to categorize the syndication of investors. The variable 1(Fund 02-07)
is a dummy variable equal to one if the PE firm raised its latest fund between 2002 and 2007. All specifications contain firm and
year fixed effects. Even-numbered columns augment the baseline model with a set of firm-level controls measured before the crisis
and interacted with the Post dummy. These variables are firm size (log of revenue), growth in revenue, cash flow over assets, ROA,
and leverage. The appendix of the RFS paper provides more information about the variable definitions. Standard errors are clustered
at the firm level. *** denotes significance at the 1% level, ** at the 5% level, and * at the 10% level.
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Investment/assets Net debt iss./assets Net equity contr./assets
Post*High dry powder 0.105** 0.086** 0.053* 0.062** 0.070*** 0.055**

(0.048) (0.041) (0.031) (0.03) (0.025) (0.022)
Post*1(Fund 02-07) 0.075* 0.090** 0.064** 0.073*** 0.036* 0.030

(0.039) (0.036) (0.026) (0.026) (0.021) (0.021)
Year fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Firm fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
Firm controls No Yes No Yes No Yes No Yes No Yes No Yes
Observations 1,582 1,539 1,582 1,539 1,589 1,546 1,589 1,546 1,565 1,527 1,565 1,527
Adjusted R-squared 0.108 0.117 0.106 0.117 0.064 0.068 0.064 0.068 0.028 0.048 0.023 0.044
disruption when credit markets dry up. We found that for advantage for financially constrained firms that were backed
our entire sample of companies, PE-backed and otherwise, by PE firms was larger during the 2008-2009 period than
those in the top quartile of the leverage distribution at the during the next two years.
onset of the crisis in 2007 experienced lower investment
during and after the crisis. But even so, the high-leveraged Do Differences Among PE Groups Explain Differences
companies that were backed by PE investors reduced invest- in Their Companies’ Investment and Funding?
ments significantly less than their non-PE counterparts To further explore the underlying channel of the findings, we
(Table 3, panel A, Columns 5 and 6). In other words, the focused on the differences across PE firms in their financial
presence of a PE investor worked to counterbalance the and operational resources that were available in 2007, at the
negative effect of high leverage on investments.32 onset of the financial crisis.
At the same time, we found that the effect of PE on First, we compared PE groups based on the amount of
debt issuance to be stronger among financially constrained “dry powder” that they had available at the onset of the crisis.
companies (Table 3, panel B), whether arising from small PE firms with more available capital were presumably better
firm size, dependence on external finance, or high leverage. positioned to provide liquidity to their portfolio compa-
In the case of equity contributions, the benefits appeared nies and better able to commit more time and attention to
to be roughly comparable across all PE-backed companies, portfolio companies, since they had deployed less capital. We
from the most to the least financially constrained. And in divided the PE-backed companies into two groups, depend-
still another test, we found that the investment and financing ing on whether their PE investors had dry powder at the top
quartile levels in 2007.
As reported in Table 4, we found that companies whose PE
32 On the other hand, companies that expect to respond more successfully to a
negative credit shock should ex ante employ more debt. Therefore, it is reasonable to
investors had a considerable amount of dry powder at the begin-
think that the results are downward biased. ning of the crisis increased their investment level relatively more.

Table 5
Performance Analysis
This table reports several analyses that aim to explore the performance of PE deals. Panel A reports a standard difference-in-differ-
ences fixed effects model exploring various performance measures. All specifications include firm and year fixed effects. In Columns
1 and 2 the outcome is 1-year assets growth; in Columns 3 and 4 is total EBITDA scaled by revenue; in Columns 5 and 6 is ROA.
Standard errors are clustered at the firm level. In panel B, the dependent variable is firm market share, measured as the log of share
of firms’ revenue scaled by total revenue at the level of three-digit SIC industry. Columns 1 and 2 estimate the standard model, but
using only data from 2004 to 2009. Columns 3 and 4 instead uses the full sample period of 2004-2011. Lastly, Columns 5 and
6 report the coefficient from the time-varying regression. Standard errors are clustered at the firm level. In panel C, we report the
marginal value (at the mean) of a conditional logit model, where we study the effect of being a PE-backed company on various exit
outcomes. Even-numbered columns have firm-level controls at 2007. In Columns 1 and 2 the outcome is a dummy equal to one
if the company was the target of an M&A activity in the post-crisis period; in Columns 3 and 4 the outcome is instead a dummy
equal to one if the company was a target of an M&A activity and the company does not exit from the data in the same time frame;
in Columns 5 and 6 the outcome is the dummy equal to one if the company exits the data set in the post-crisis period; lastly in
Columns 7 and 8 the outcome is a dummy if the company exits the data and it reported some financial difficulties before the exit.
Standard errors are clustered by two-digit industry. In the first two panels, even-numbered columns augment the baseline model
with a set of firm-level controls measured before the crisis and interacted with the Post dummy. In the last panel, control variables
are not interacted. These variables are firm size (log of revenue), growth in revenue, cash flow over assets, ROA, and leverage. See
the appendix of the RFS and the paper for more information about the variables. *** denotes significance at the 1% level, ** at the
5% level, and * at the 10% level.
A. Accounting Performance
(1) (2) (3) (4) (5) (6)
Assets growth EBITDA/REV ROA
PE firm x Post 0.148*** 0.124*** -0.009 -0.010 -0.003 -0.004
(0.040) (0.038) (0.013) (0.014) (0.009) (0.008)
Firm fixed effects Yes Yes Yes Yes Yes Yes
Firm controls No Yes No Yes No Yes
Observations 13,180 12,528 12,507 12,137 12,865 12,364
Clusters 1,984 1,878 1,960 1,878 1,984 1,878
R-squared 0.026 0.042 0.001 0.015 0.005 0.041
B. Market Share
(1) (2) (3) (4) (5) (6)
PE firm x Crisis 0.081** 0.079** 0.050 0.055*
(0.035) (0.034) (0.034) (0.033)
PE firm x y2004 0.039 0.048
(0.057) (0.059)
PE firm x y2005 0.035 0.047
(0.050) (0.049)
PE firm x y2006 -0.035 -0.020
(0.036) (0.035)
PE firm x y2008 0.094*** 0.106***
(0.031) (0.034)
PE firm x y2009 0.072** 0.088***
(0.031) (0.033)
PE firm x y2010 0.039 0.052
(0.037) (0.039)
PE firm x y2011 -0.007 0.005
(0.053) (0.055)
Sample 2004-2009 Whole sample Whole sample

Firm fixed effects Yes Yes Yes Yes Yes Yes
Firm controls No Yes No Yes No Yes
Observations 9,090 8,847 12,697 12,326 12,697 12,326
Clusters 1,960 1,878 1,960 1,878 1,960 1,878
R-squared 0.035 0.087 0.021 0.064 0.021 0.064

Table 5
continued
C. Exit outcomes
(1) (2) (3) (4) (5) (6) (7) (8)

Marginal eff. 1{Exit} 1{Bankruptcy} 1{M&A} 1{M&A, no distress}
PE Firm 0.058 0.039 0.058 0.039 0.351*** 0.325*** 0.351*** 0.318***
(0.085) (0.087) (0.085) (0.087) (0.023) (0.101) (0.024) (0.100)
Industry (2-digit) FEs Yes Yes Yes Yes Yes Yes Yes Yes
Firm controls 2007 Yes Yes Yes Yes
Observations 1,368 1,368 1,368 1,368 1,635 1,635 1,635 1,635
The finding was both statistically and economically significant: their performance. Here we looked at various measures of
a company backed by a top-quartile dry powder PE firm experi- company performance during and after the crisis.
enced a 10% increase in investment over assets relative to the As reported in panel A in Table 5, we found that the
control group. Consistent with this result, we also found this total assets of PE-backed companies grew faster than those
group of PE firms to be more active in financing their portfolio of the matched firms—a pattern that is consistent with prior
companies. Companies financed by top quartile dry powder findings that PE-backed companies decreased their investment
groups had 5% greater debt issuances and, perhaps more impor- relatively less during the crisis.
tant, 7% larger equity injections. Next, we explored a number of operating measures of
As also reported in Table 4, we also found a larger firm performance around the crisis period. Here we found
increase in investment when PE investors had raised funds that PE-backed companies achieved roughly the same level
more recently, suggesting the importance of the availability of of operating margins, as reflected in EBITDA as a percent-
resources for the PE group. We find similarly strong patterns age of revenue (see Column 5) and ROA, defined as net
with respect to debt issuances, which have increased much income over assets (Column 6). We also looked at market
more for companies where the investors had raised a fund share, which earlier studies suggest are one way that compa-
more recently (Columns 7 and 8). Both these effects were nies could have created longer-run value during the crisis.34
both economically and statistically significant.33 To explore this possibility, we examined two other aspects of
Overall, our findings are consistent with the hypothesis the companies’ performance: increases in market share during
that portfolio companies backed by PE investors with more and soon after the crisis; and the patterns of firm exit—both
resources at the onset of the crisis were more likely to increase positive (non-distressed M&A acquisitions) and negative
investments in their portfolio companies during the crisis—a (bankruptcy)—in the post-crisis period.
finding we come back to when discussing our survey of PE
investors at the end. Effects on Market Share
The increase in investment by PE-backed companies has the
Performance and Company Outcomes Analysis potential to increase longer-run profits and value to the extent
In the second main phase of our study, we attempted to it enables the firms to capture larger market share without
understand the longer-term efforts of PE ownership during sacrificing profitability. As reported in panel B of Table 5,
the crisis by examining the operating performance of and after measuring each firm’s sales relative to the total operating
prospects for PE-backed companies. If the investments by revenue in its industry (using the three-digit SIC codes),35 we
PE-backed companies proved to be unwise or wasteful, we found that during the crisis period (2008-2009) PE-backed
would expect such decisions to have had negative effects on companies experienced an 8% increase in market share relative
34 For example, Gilchrist, et al. (2017) show that during the 2008 crisis, less finan-
33 In Table A.9 in the appendix of the RFS paper, we explore heterogeneity across cially constrained firms lowered their prices as an investment to build market share. In
portfolio companies that were backed by a single PE investor versus portfolio companies contrast, financially constrained firms could not pursue such a strategy since they needed
backed by a syndicate of PE investors. A syndicate of PE investors may, on the one hand, the liquidity in the short run to meet their financial obligations.
benefit portfolio companies by enabling access to “deeper pockets,” but, on the other 35 The total operating revenue of the industry is constructed using only medium and
hand, may generate “free riding” and coordination problems within the syndicate. We large firms in the Orbis/Amadeus data, as previously discussed. Results are also similar
find no statistically significant differences between the two types of portfolio companies. using the SIC two-digit industry classification.

to the control group—but that this effect became progressively ily observable. To provide further evidence on the underlying
smaller and statistically insignificant in 2010 and 2011. This mechanisms driving these findings, we conducted a large-scale
result reflects our earlier finding that the changes in invest- survey of private equity investors.
ment and funding policy were mostly concentrated in 2008 Surveying PE investors is not easy because these investors
and 2009, when the financial turmoil was at its height.36 tend to be not only short of time, but reluctant to share propri-
etary details about their operations. To increase the likelihood
Analysis of Exits of participation in the survey, we limited our survey popula-
Finally, we examined exit patterns in the post-crisis period, tion to alumni from our respective business schools—Harvard,
comparing the relative likelihood that PE-backed compa- Northwestern, and Stanford. In total, we identified roughly
nies became distressed, went bankrupt, or were acquired. As 3,100 alumni with current or past PE experience, through
mentioned earlier, our aim was to distinguish “bad exits”— both alumni offices and the Pitchbook database. We distrib-
cases where firms exit our sample after a status of financial uted the survey electronically to these alumni and obtained
distress or bankruptcy—from “potentially profitable exits,” 319 responses, which amounts to a response rate of 10.3%.39
which we identify as acquisitions of firms showing no signs of As is typical of surveys, our sample is unlikely to be
corporate distress. Our study was “cross-sectional” and exam- perfectly representative of the PE universe. Nonetheless, we
ined the exit status of our sample firms at a single point in see no reason to believe that the sample, and the willingness to
time, the end of 2011. Our matching methodology requires participate in the survey, should bias our results toward partic-
that both PE-backed companies and control firms did not exit ular types of activities that the investors undertook during the
before the matching period at the onset of the financial crisis. crisis. As can be seen in Table 6, 80% of the survey participants
As reported in panel C of Table 5, we found that PE-backed were private equity partners, with an average of 14 years of
companies were not more likely to go out of business or enter experience as PE investors. Their roles in the PE groups were
into distress in the post-crisis period (Columns 1-4). At the identified as “deal making” (80% of survey participants), “deal
same time, we found that PE-backed companies were more sourcing” (72%), “financial structuring” (64%), and “improv-
likely to experience a potentially profitable exit (Columns ing portfolio companies’ operations” (61%). There is a wide
5-8).37 Finally, when we explored the correlation between variation in the size of assets under management of survey
equity injections and various firm outcomes, such as invest- participants, with roughly 40% of participants representing
ment, ROA, and firm exit, we found suggestive evidence that funds with less than $1 billion of assets under management,
PE-backed companies that received equity financing were less and more than 20% of the participants working with funds
likely to declare bankruptcy during the crisis, invested more, with over $10 billion.
and had a higher ROA .38 Private equity activism during the crisis. In the first part
of our survey, we presented participants a list of operating and
Survey Evidence financial activities that are common to PE investors, such as
If our main finding—PE-backed companies’ greater ability improving portfolio company operations, restructuring debt
to maintain investment during the crisis—can be explained obligations, and injecting equity. We asked participants to
by their greater success in raising equity and debt financing, state the degree in which they believed that private equity
the particular channels through which PE investors interacted investors were more or less likely to engage in such activities
with their portfolio companies during the crisis were not read- during the financial crisis of 2008 (as compared to normal
circumstances). We used a standard 5-point Likert scale, where
36 Furthermore, this timing is consistent with the findings of Gilchrist et al. (2017),
potential responses ranged from significantly more likely (=1),
who identified changes in pricing strategies of financially less constrained firms concen- more likely (=2), same (=3), less likely (=4), and significantly
trated in the 2008 and 2009 crisis years.
37 However, it is worth noting that “potentially positive exit” is a very crude measure
less likely (=5).
of a successful exit. More generally, given the empirical limitations of these analyses, the As reported in Panel C of Table 6, the most striking
results should be interpreted cautiously. One important limitation of the exit analysis is result was that almost 90% of participants indicated that PE
that because of the timing of the matching estimator, we do not have pre-crisis bench-
mark exit rates. To address this, we repeat the analysis with an alternative matching investors were more likely to assist portfolio companies with
approach, in which we construct the matched sample in 2004 before the crisis. The
concern with this approach is that it does not construct appropriate counterfactuals at
the onset of the crisis. Nevertheless, Table A.11 in the appendix of the RFS paper reports 39 By way of comparison, see Gorman and Sahlman (1989), who obtained re-
the results. The analysis reaffirms the results of the bankruptcy exits, but suggests that sponses from 49 venture capitalists. Gompers et al. (2016) obtained responses from 79
the increased tendency to exit through M&A is not unique for PE-backed companies in buyout investors. Our response rate is quite similar to those of other large-scale surveys.
the crisis period. Graham and Harvey (2001) obtain a response rate of 8.9% from CFOs, and Da Rin and
38 In Table A.12 in the appendix of the RFS paper. Phalippou (2014) obtain a response rate of 13.8% from PE-limited partners.

Table 6
Survey Results
This table reports the survey answers of 319 participants. Panel A describes survey participants’ characteristics, and panel B re-
ports the size of assets under management of the funds of survey participants. Panels C and D summarize survey results. Survey
participants received a list of common private equity activities related to firm operations (panel C) and firm financials (panel D).
For each activity, participants answered whether this activity is more or less likely to take place during the 2008 financial crisis,
when compared to normal times. Participants’ answers ranged from significantly more likely (=1), more likely (=2), same (=3),
less likely (=4), and significantly less likely (=5). Column 2 of panels C and D report the average response and the significance
level in which the average response is different from the neutral answer (Same=3). Column 4 reports the fraction of survey par-
ticipants with answer less than 3 (more likely during the crisis); Column 5 reports the fraction of participant responding precisely
3 (same during the crisis); and Column 6 reports the fraction of participants reporting more than 3 (less during the crisis). Figure
A.3 of the appendix of the RFS paper shows the survey questions. *** denotes significance at the 1% level, ** at the 5% level,
and * at the 10% level.
A. Participants Characteristics
Mean Median SD
Partner 0.793 1 0.46
Years of experience (as PE investor) 14.09 13 7.5
Tasks within the fund
Deal making 0.807 1 0.395
Deal sourcing 0.729 1 0.44
Improving portfolio company operations 0.609 1 0.488
Financial structuring of deals 0.644 1 0.479
Fund raising 0.429 0 0.495
B. Fund Size (Assets Under Management)
Less than US$1b 43.24%

US$1b-US$5b 28.38%
US$5b-US$10b 8.78%
US$10b-US$50b 15.54%
Greater than US$50b 4.05%
C. Operational Activities
Question Observations Avg. response SD More during Same (%) Less during
crisis (%) crisis (%)
1 Assist portfolio companies with their operating problems 319 1.586*** 0.715 89.58 9.45 0.98
2 Provide strategic guidance to portfolio companies 319 1.918*** 0.806 76.55 21.82 1.63
3 Replace CEO or senior executives of portfolio companies 319 2.495*** 0.810 45.93 47.23 6.84
4 Interact frequently with the management of the portfolio 319 1.778*** 0.698 84.36 15.64 0.00
company
5 Connect companies with potential customers, suppliers, or 319 2.456*** 0.737 47.23 49.51 3.26
strategic partners
6 Connect companies with potential investors 319 2.830*** 0.924 32.25 47.23 20.52
7 Help companies hire managers 319 2.615*** 0.702 37.46 56.68 5.86
8 Provide stronger incentive-based compensation to manage- 319 2.690*** 0.708 31.27 61.89 6.84
ment in portfolio company
9 Search for a potential buyer 319 3.508 1.101 19.87 22.80 57.33
10 Increase frequency of board meetings per year 319 2.521*** 0.653 41.37 57.98 0.65

Table 6
continued
D. Financial Activities
Avg. More during Less during

Question Observations response SD crisis (%) Same (%) crisis (%)
1 Inject equity to alleviate financing constraints 319 2.254*** 1.132 77.85 0.65 21.50
2 Search and evaluate new deals 319 3.592 1.331 28.34 7.49 64.17
3 Assist portfolio companies with financial structure issues 319 1.801*** 0.887 89.90 0.00 10.10
4 Assist portfolio companies to raise debt financing 319 2.726*** 1.320 59.28 2.93 37.79
5 Assist portfolio companies to renegotiate loan terms and
debt obligations 319 1.827*** 0.922 88.60 0.00 11.40
6 Interact with bankers and lawyers regarding the financial
structure of companies 319 2.042*** 1.055 81.43 0.00 18.57
7 Buy back debt obligations of portfolio companies 319 2.664*** 1.387 58.31 0.65 41.04
operating problems during the crisis. And roughly 85% of What was the private equity advantage during the
the respondents reported that PE investors were more likely crisis? In the final part of our survey, we asked PE inves-
to interact frequently with portfolio companies during the tors which factors they believed were most instrumental in
crisis, and 77% said that such investors provided more strate- assisting portfolio companies during the crisis. Our partici-
gic guidance during that time period. There was less consensus pants’ responses emphasized the value of majority control and
on the particular forms such operating guidance assumed, private ownership that does not require scrutiny from public
which included hiring managers, connecting to investors, and equity markets. Both of these factors were seen as critical to
finding strategic partners. the ability of PE investors to engage in operational engineer-
Overall, these responses suggest that PE investors were ing and to assist the portfolio companies during the crisis. At
more involved and engaged with firm operations during the the same time, many respondents noted that the long invest-
crisis—which in turn sheds light on the underlying mecha- ment horizon of PE investors was perhaps equally important,
nisms that led to the increased investment and market share enabling PE investors to make significant changes.
of portfolio companies during the crisis. Survey participants also cited the importance of PE inves-
Financing Activities. As reported in panel D of Table tors’ access to banks in facilitating the restructuring of debt.
6, 90% of our respondents also viewed PE investors as The repeated interaction of banks and PE investors, besides
more likely to help their portfolio companies with financ- their value in limiting information asymmetries faced by
ing challenges. More specifically, nearly 90% of our survey investors, was said to provide additional cross-selling oppor-
participants highlighted the renegotiation of debt obligations tunities for banks, which in turn increases the incentives of
and more than 80% emphasized increased interactions with banks to provide liquidity to PE-backed companies during
bankers and lawyers regarding the financial structure of the the crisis and renegotiate debt obligations.
portfolio companies. Interestingly, almost 80% of our respon- Finally, PE firms’ dry powder during the crisis was identi-
dents reported that their own firms were more likely to inject fied as critical to its role in assisting portfolio companies
equity into their portfolio companies during the crisis, and during the crisis. At the same time, however, the possible
60% helped in raising debt financing—again, all completely roles of high leverage and the strong financial incentives of
consistent with our empirical findings. managers in encouraging the operating efficiency of portfolio
Also consistent with the above, our survey respondents companies were viewed as relatively unimportant.
reported that PE investors were less likely to search and evalu-
ate new investments during this period. They were mostly
focusing on operating and financing challenge in their existing crisis relative to smaller funds. We find no differences across almost all dimensions. The
only two exceptions were that investors in larger funds were more likely to increase the
companies.40 frequency of board meetings and buy back public debt. Both differences may arise be-
cause these investors were more likely to acquire larger companies, in which communi-
cation was more formal (through board meetings) and more likely to have issued public
40 Table A.13 in the appendix of the RFS paper explores whether survey participants debt. We repeat the exercise by dividing investors by experience. We do not find signifi-
of large funds (above $5B in assets under management) behaved differently during the cant or economically meaningful differences in their responses.

To summarize, our survey results suggest that PE investors returns, the consequences for portfolio companies will be
are active investors who were more likely to engage in both modest.
operational and financial activities during the crisis. Specifically, There are reasons, however, to be more concerned about
the PE investors claimed to spend more time with portfolio the situation today. First, the Global Financial Crisis was
companies helping them address operating as well as financ- characterized, when examined using a number of critical
ing challenges. They described themselves as more likely to financial market metrics, by a sharp “V” shape. By July 2009,
renegotiate debt obligations, help raise additional debt, and for instance, the BofA Merrill Lynch BB US High Yield
provide more equity. These PE investors also maintained that Index had retreated from its peaks back to values very close
such activities during the crisis were encouraged, if not made to that seen immediately before the beginning of the recession
possible, by particular features of private equity investments— (December 2007).41 If all goes well, the same will be said in
notably majority control, private ownership, deep connections hindsight about this crisis. However, there is a material possi-
to banks, and the availability of “dry powder.” These survey bility of a more prolonged economic downturn, which may
results are completely consistent with and supportive of our place greater stresses on private equity portfolio companies.
empirical findings. Second, once the widespread adoption of EBITDA adjust-
ments and capital call lines are factored in, it seems likely that
Conclusion PE-backed firms were substantially more levered at the end of
In a study whose findings were published in the Review of 2019 than at the end of 2007.42 And since such leverage may
Financial Studies in 2019, we examined how PE-backed pose greater challenges for companies and their returns, our
companies responded to the turmoil caused by the 2008 confidence in projecting the future for PE-backed firms must
financial crisis by exploring their investments, financing, and be tempered with caution.
performance. One of the main objectives of this analysis was
to explore whether PE-backed companies contributed to the
fragility of the economy during this period, as both regula- Shai Bernstein is an Associate Professor in the Entrepreneurial
tors warned and the findings of academic studies suggested Management Group at Harvard Business School.
they might.
Our main finding was that PE-backed companies reduced Josh Lerner is the Jacob H. Schiff Professor and Chair of the Entre-
their investment less than the control group during the finan- preneurial Management Unit at Harvard Business School.
cial crisis. This result is explained by the ability of PE-backed
companies to draw on the resources and relationships of their Filippo Mezzanotti is an Assistant Professor of Finance at Kellogg
PE sponsors to raise equity and debt funding in this difficult School of Management, Northwestern University.
period, and to reduce their cost of debt capital, as reflected in
their lower interest expense. Furthermore, the positive invest-
ment effect of PE was particularly large in companies more
likely to be financially constrained at the time of the crisis
and when the PE firms had more resources in the form of
“dry powder.” The increase in investment during the crisis
led to increased asset growth and higher market shares. We
find evidence consistent with all these empirical findings
in our survey of over 300 PE practitioners, as well as in an
analysis of renegotiation of loans. Viewed in their entirety,
then, our results provide little support for the popular notion
that private equity compounded the fragility of the economy
during the Global Financial Crisis.
A natural follow-on question is whether we can draw
conclusions about the crisis we find ourselves in today. These
results of this work might lead us to be optimistic about the
implications for PE of the COVID-19 virus and the economic
41 See https://fred.stlouisfed.org/series/BAMLH0A1HYBB.
disruptions that have followed. We might anticipate based 42 Albertus and Denes (2019); Rasmussen and Chingono (2019) https://ftal-
on these findings that whatever the impact on private equity phaville.ft.com/2019/11/11/1573493232000/Private-equity-run-amok/.

Appendix 1: Robustness Figure 6
Our RFS study included a set of robustness analyses. First, Effect of PE-backed companies on
we drop management buyouts from the main sample. At debt issuances over time
least historically in the United Kingdom, MBOs were char-
acterized by lower engagement by PE firms. If their inclusion This figure illustrates the change in debt issuances separately
completely drove the results, the interpretation and generaliza- for both PE and non-PE companies in our sample. Specifically,
tion of the analysis might be subtler. To explore whether this is the figure reports αt of the following equation: γit = αt + αi +
the case, we eliminate MBOs from the sample and repeat the εit. The equation is separately estimated for PE and non-PE
main analysis. We find similar results: the exclusion of MBOs companies. αt captures year fixed effects, and αi captures
does not affect the results. firm fixed effects. The year 2007 is used as base period, and
We also explore whether our results are driven by public- therefore the corresponding coefficient is normalized to zero.
to-private PE transactions. We find that the results remain The estimates are plotted with standard errors above and
unchanged. This result is expected, given that such deals below the point estimates. Standard errors are clustered at the
compromise only 5% of the transactions in our sample. firm level.
Next, we address concerns that our estimates may be
.1
biased by attrition. As usual with panel data, endogenous exit
through acquisition or bankruptcy may bias the results. First,
we note that the shift in investment and financing policies
0
occurred already in 2008 (at the peak of the financial crisis),

Debt Issuance
while firm exit only took place later. We can also illustrate
−.1
this pattern more directly by estimating our standard model

using data from 2007 and 2008 only, in which we find similar
results. In other words, much of the shift in corporate policy
−.2
happened soon after the onset of the crisis.

An alternative robustness test to address attrition bias
−.3
concerns is to focus only on firms that did not exit the sample. 2004 2005 2006 2007 2008 2009 2010 2011
Year
We take this conservative approach and drop every firm that
exited the database before 2011. This approach leads to PE Non−PE
approximately 15% fewer observations in the sample.43 Even

with this reduced sample, the main results remain unchanged. in magnitude and statistical power to the one presented in
PE-backed companies experienced a smaller decline in invest- the text.
ment and a relative increase in equity contributions and debt Finally, we further explore the robustness of our matching
issuance. At the same time, the leverage ratio stayed constant, procedure. As discussed earlier, in the analysis we undertake
and interest expense declined. the matching at the time of the onset of the financial crisis, to
Our results are robust also to changes in industry dynam- construct proper counterfactuals for the PE-backed companies
ics. One concern is that PE-backed companies may be more during the crisis. An alternative approach would be to match
or less sensitive than the control group to any changes in PE-backed companies in an earlier year or just before the
demand that were contemporaneous with the crisis. In princi- private equity buyout. While these matching approaches will
ple, this should not be a problem, because the treatment and not construct the most appropriate counterfactual at the onset
control groups are matched within industries. Nonetheless, of the crisis, it will provide a treatment effect that combines the
we augment our analysis with a full set of (two-digit) industry joint effect of the buyout and the crisis. We explore construct-
fixed effects interacted with the Post dummy. This set of fixed ing alternative control groups by conducting the matching at
effects can control non-parametrically for changes in demand various points in time. Specifically, we conduct the match-
and other time-varying industry characteristics. Despite the ing in pre-crisis years (i.e., in 2003, 2004, or 2005), as well
large number of fixed effects that the model introduces, the as in the year before the buyout. Across all these different
main results remain unchanged. The estimates are still close approaches, we find that main results of the paper remain
similar and relatively unchanged.
43 In our sample, about 310 firms exit before 2011.

Figure 7
Survey Responses–Comparing Crisis and Normal Times
This figure reports the survey answers of 319 participants. Survey participants received a list of common private equity activities
related to firm operations (panel A) and firm financials (panel B). For each activity, participants answered whether this activity is more
or less likely to take place during the 2008 financial crisis, when compared to normal times. Participants’ answers ranged from sig-
nificantly more likely (=1), more likely (=2), same (=3), less likely (=4), and significantly less likely (=5). Both figures illustrate the
fraction of survey participants with answers less than 3 (more likely during the crisis), answers equal 3 (same during the crisis), and
answers greater than 3 (less during the crisis). Figure A.3 of the appendix of the RFS paper presents the survey questions.
Panel A: Operational Activities
Assist portfolio companies with operating problems

Provide strategic guidance to portfolio companies
Replace CEO or senior executives
Interact frequently with the management
Connect companies with potential customers, suppliers,
or strategic partners
Connect companies with potential investors
Help companies hire managers
Provide stronger incentive-based compensation
Search for a potential buyer
Increase frequency of board meetings
0% 20% 40% 60% 80% 100%
More during Crisis Same Less during Crisis
Panel B: Financial Activities
Inject equity to alleviate financing constraints
Search and evaluate potential new deals
Assist portfolio companies with financial structure issues
Assist portfolio companies to raise debt financing

Assist portfolio companies to renegotiate loan terms and
debt obligations
Interact with bankers and lawyers regarding financial structure
Buy back debt obligations of portfolio companies
0% 20% 40% 60% 80% 100%
More during Crisis Same Less during Crisis

Figure 8
Survey Responses–Potential Explanations for Behavior during Crisis
This figure reports the survey answers of 319 participants. Survey participants were presented with a list of various aspects of private
equity firms’ structure and investments, and were asked to answer which aspects were most useful for portfolio companies to weather
the crisis. The distribution of the responses is provided in the figure. The survey questions are presented in Figure A.2 of the Appendix.
Better access to banks and to the restructuring of debt obligations
High leverage disciplines portfolio companies to be efficient
Dry powder that can be used for follow-on investments

during the crisis
Private ownership that requires no scrutiny from public markets
Long investment horizon of Private Equity firms
Strong financial incentives provided to management of

portfolio companies
Operational expertise of Private Equity investors
Majority control of portfolio companies

0% 10% 20% 30% 40% 50% 60% 70% 80%

References Giroud, X., and, H. Mueller. 2015. Firm leverage and
Albertus, J., and M. Denes. 2020, Private equity fund unemployment during the great recession. Working Paper,
debt: Capital flows, performance, and agency costs, https:// National Bureau of Economic Research.
papers.ssrn.com/sol3/papers.cfm?abstract_id=3410076. Goergen, M., N. O’Sullivan, and G. Wood. 2011. Private
Axelson, U., T. Jenkinson, P. Strömberg, and M. equity takeovers and employment in the UK: Some empiri-
Weisbach. 2013. Borrow cheap, buy high? The determinants cal evidence. Corporate Governance: An International Review
of leverage and pricing in buyouts. Journal of Finance 68:2223- 19:259-75.
67. Gompers, P., S. Kaplan, and V. Mukharlyamov. 2016.
Bernanke, B. 1983. Nonmonetary effects of the financial What do private equity firms say they do? Journal of Financial
crisis in the propagation of the Great Depression. American Economics 121:449-76.
Economic Review 73:257-75. Gompers, P., K. Mugford, and D. Kim. 2012. Bain
Bernanke, B., and M. Gertler. 1990. Financial fragility Capital: Outback Steakhouse. Harvard Business School Case
and economic performance. Quarterly Journal of Economics No. 212-087.
105:87-114. Gorman, M., and W. Sahlman. 1989. What do venture
Bernstein, S., J. Lerner, M. Sorensen, and P. Stromberg. capitalists do? Journal of Business Venturing 4:231-48.
2016. Private equity and industry performance. Management Graham, J., and H. Campbell. 2001. The theory and
Science 63:1198-213. practice of corporate finance: Evidence from the field. Journal
Bernstein, S., and A. Sheen. 2016. The operational of Financial Economics 60:187-243.
consequences of private equity buyouts: Evidence from the Illig, R. 2012. The dog that didn’t bark: Private investment
restaurant industry. Review of Financial Studies 29:2387-418. funds and relational contracts in the wake of the Great Reces-
Bottero, M., S. Lenzu, and F. Mezzanotti. 2020. Sover- sion. Michigan Journal of Private Equity & Venture Capital
eign debt exposure and the bank lending channel: Impact on Law 2:49-95.
credit supply and the real economy. Journal of International Kaplan, S. 1989. The effects of management buyouts on
Economics, forthcoming. operating performance and value. Journal of Financial Econom-
Boucly, Q., D. Sraer, and D. Thesmar. 2011. Growth ics 24:217-54.
LBOs. Journal of Financial Economics 102:432-53. Kaplan, S., and J. Stein. 1993. The evolution of buyout
Da Rin, M., and L. Phalippou. 2014. There is something pricing and financial structure in the 1980s. Quarterly Journal
special about large investors: Evidence from a survey of of Economics 108:313-57.
private equity limited partners. Tilburg University, Center Petersen, M., and R. Rajan. 1994. The benefits of lending
for Economic Research, Discussion Paper. relationships: Evidence from small business data. Journal of
Davis, S., J. Haltiwanger, K. Handley, R. Jarmin, J. Lerner, Finance 49:3-37.
and J. Miranda. 2014. Private equity, jobs, and productivity. Rajan, R., and L. Zingales. 1998. Financial dependence
American Economic Review 104:3956-90. and growth. American Economic Review 88:559-86.
Davis, S., J. Haltiwanger, K. Handley, J. Lerner, B. Rasmussen, D., and Chingono, B. 2019. Private Equity
Lipsius, and J. Miranda. 2019. The economic effect of runs amok, Financial Times.
private equity buyouts, https://papers.ssrn.com/sol3/papers. Riley, R., C. Rosazza Bondibene, and G. Young. 2014.
cfm?abstract_id=3465723. The financial crisis, bank lending and UK productivity:
Eaton, C., S. Howell, and C. Yannelis. 2018. When inves- Sectoral and firm-level evidence. National Institute Economic
tor incentives and consumer interests diverge: Private equity Review, no. 228, R17-R34.
in higher education. National Bureau of Economic Research Stromberg, P. 2008. The new demography of private
Working Paper no. 24976. equity. In Globalization of alternative investment working
European Central Bank. 2017. Guidance on lever- papers: The global economic impact of private equity report, vol.
aged transactions. Technical Report. https://www. 1. Geneva: World Economic Forum.
bankingsupervision.europa.eu/ecb/pub/pdf/ssm.leveraged_
transactions_guidance_201705.en.pdf.
Gilchrist, S., R. Schoenle, J. Sim, and E. Zakrajšek.
2017. Inflation dynamics during the financial crisis. American
Economic Review 107:785-823.

ADVISORY BOARD EDITORIAL
Yakov Amihud Carl Ferenbach Donald Lessard Clifford Smith, Jr. Editor-in-Chief
New York University High Meadows Foundation Massachusetts Institute of University of Rochester Donald H. Chew, Jr.
Technology
Mary Barth Kenneth French Charles Smithson Associate Editor
Stanford University Dartmouth College John McConnell Rutter Associates John L. McCormack
Purdue University
Amar Bhidé Martin Fridson Laura Starks Design and Production
Tufts University Lehmann, Livian, Fridson Robert Merton University of Texas at Austin Mary McBride
Advisors LLC Massachusetts Institute of
Michael Bradley Technology Erik Stern Assistant Editor
Duke University Stuart L. Gillan Stern Value Management Michael E. Chew
University of Georgia Gregory V. Milano
Richard Brealey Fortuna Advisors LLC G. Bennett Stewart
London Business School Richard Greco Institutional Shareholder
Filangieri Capital Partners Stewart Myers Services
Michael Brennan Massachusetts Institute of
University of California, Trevor Harris Technology René Stulz
Los Angeles Columbia University The Ohio State University
Robert Parrino
Robert Bruner Glenn Hubbard University of Texas at Austin Sheridan Titman
University of Virginia Columbia University University of Texas at Austin
Richard Ruback
Charles Calomiris Michael Jensen Harvard Business School Alex Triantis
Columbia University Harvard University University of Maryland
G. William Schwert
Christopher Culp Steven Kaplan University of Rochester Laura D’Andrea Tyson
Johns Hopkins Institute for University of Chicago University of California,
Applied Economics Alan Shapiro Berkeley
David Larcker University of Southern
Howard Davies Stanford University California Ross Watts
Institut d’Études Politiques Massachusetts Institute
de Paris Martin Leibowitz Betty Simkins of Technology
Morgan Stanley Oklahoma State University
Robert Eccles Jerold Zimmerman
Harvard Business School University of Rochester
Journal of Applied Corporate Finance (ISSN 1078-1196 [print], ISSN 1745-6622 Statement on Research4Life
[online]) is published quarterly per year by Wiley Subscription Services, Inc., a Wiley Wiley is a founding member of the UN-backed HINARI, AGORA, and OARE initiatives.
Company, 111 River St., Hoboken, NJ 07030-5774 USA. They are now collectively known as Research4Life, making online scientific content
available free or at nominal cost to researchers in developing countries. Please visit
Postmaster: Send all address changes to JOURNAL OF APPLIED CORPORATE FI- Wiley’s Content Access – Corporate Citizenship site: http://www.wiley.com/WileyCDA/
NANCE, John Wiley & Sons Inc., c/o The Sheridan Press, PO Box 465, Hanover, PA Section/id-390082.html
17331 USA.
Journal of Applied Corporate Finance accepts articles for Open Access publication.
Information for Subscribers Please visit https://authorservices.wiley.com/author-resources/Journal-Authors/open-
Journal of Applied Corporate Finance is published quarterly per year. Institutional sub- access/onlineopen.html for further information about OnlineOpen.
scription prices for 2020 are:
Print & Online: US$844 (US), US$1007 (Rest of World), €656, (Europe), £516 Wiley’s Corporate Citizenship initiative seeks to address the environmental, social,
(UK). Commercial subscription prices for 2020 are: Print & Online: US$1123 (US), economic, and ethical challenges faced in our business and which are important to
US$1339 (Rest of World), €872 (Europe), £686 (UK). Individual subscription prices our diverse stakeholder groups. Since launching the initiative, we have focused on
for 2020 are: Print & Online: US$137 (US), $137 (Rest of World), €115 (Europe), sharing our content with those in need, enhancing community philanthropy, reducing
£79 (UK). Student subscription prices for 2020 are: Print & Online: US$49 (US), our carbon impact, creating global guidelines and best practices for paper use, estab-
$49 (Rest of World), €41 (Europe), £28 (UK). Prices are exclusive of tax. Asia-Pacific lishing a vendor code of ethics, and engaging our colleagues and other stakeholders
GST, Canadian GST/HST and European VAT will be applied at the appropriate rates. in our efforts.
For more information on current tax rates, please go to https://onlinelibrary.wiley.com/
library-info/products/price-lists/payment. The price includes online access to the cur- Follow our progress at www.wiley.com/go/citizenship.
rent and all online back files to January 1, 2016, where available. For other pricing
options, including access information and terms and conditions, please visit https://on- Abstracting and Indexing Services
linelibrary.wiley.com/library-info/products/price-lists. Terms of use can be found here: The Journal is indexed by Accounting and Tax Index, Emerald Management
https://onlinelibrary.wiley.com/terms-and-conditions. Reviews (Online Edition), Environmental Science and Pollution Management,
Risk Abstracts (Online Edition), and Banking Information Index.
Delivery Terms and Legal Title
Where the subscription price includes print issues and delivery is to the recipient’s Disclaimer
address, delivery terms are Delivered at Place (DAP); the recipient is responsible for The Publisher, Cantillon and Mann, its affiliates, and Editors cannot be held respon-
paying any import duty or taxes. Title to all issues transfers FOB our shipping point, sible for errors or any consequences arising from the use of information contained in
freight prepaid. We will endeavour to fulfil claims for missing or damaged copies within this journal; the views and opinions expressed do not necessarily reflect those of the
six months of publication, within our reasonable discretion and subject to availability. Publisher, Cantillon and Mann, its affiliates, and Editors, neither does the publication of
advertisements constitute any endorsement by the Publisher, Cantillon and Mann, its
Journal Customer Services: For ordering information, claims and any inquiry concern- affiliates, and Editors of the products advertised.
ing your journal subscription please go to https://hub.wiley.com/community/support/
onlinelibrary or contact your nearest office. Copyright and Copying
Americas: Email: [email protected]; Tel: +1 781 388 8598 or Copyright © 2020 Cantillon and Mann. All rights reserved. No part of this publication
+1 800 835 6770 (toll free in the USA and Canada). may be reproduced, stored or transmitted in any form or by any means without the
Europe, Middle East and Africa: Email: [email protected]; prior permission in writing from the copyright holder. Authorization to photocopy items
Tel: +44 (0) 1865 778315. for internal and personal use is granted by the copyright holder for libraries and other
Asia Pacific: Email: [email protected]; Tel: +65 6511 8000. users registered with their local Reproduction Rights Organization (RRO), e.g., Copy-
Japan: For Japanese speaking support, Email: [email protected] right Clearance Center (CCC), 222 Rosewood Drive, Danvers, MA 01923, USA (www.
Visit our Online Customer Help at https://hub.wiley.com/community/support/onlineli- copyright.com), provided the appropriate fee is paid directly to the RRO. This consent
brary does not extend to other kinds of copying such as copying for general distribution,
for advertising or promotional purposes, for republication, for creating new collective
Production Editor: Shalini Chawla (email: [email protected]). works or for resale. Permissions for such reuse can be obtained using the RightsLink
Back Issues: Single issues from current and recent volumes are available at the “Request Permissions” link on Wiley Online Library. Special requests should be ad-
current single issue price from [email protected]. Earlier issues may be dressed to: [email protected].
obtained from Periodicals Service Company, 351 Fairview Avenue – Ste 300,
Hudson, NY 12534, USA. Tel: +1 518 537 4700, Fax: +1 518 537 5899,
Email: [email protected]
View this journal online at wileyonlinelibrary.com/journal/jacf.

Private Equity and Portfolio Companies Lessons From The Global Financial Crisis

Uploaded by

Copyright:

Available Formats

Private Equity and Portfolio Companies Lessons From The Global Financial Crisis

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Private Equity and Portfolio Companies Lessons From The Global Financial Crisis

Uploaded by

Copyright:

Available Formats

V O LU ME 3 2

Private Equity 8 Private Equity: Accomplishments and Challenges

21 Private Equity and Portfolio Companies: Lessons from the

43 Board 3.0: What the Private-Equity Governance Model Can Offer

52 The Growing Blessing of Unicorns: The Changing Nature of the

73 EQT: Private Equity with a Purpose

87 Private Equity and the COVID-19 Economic Downturn:

92 University of Texas Roundtable on LP Perspectives on the

100 Columbia Law School Roundtable on Public Aspects of Private Equity

108 A CEO’s Playbook for Creating Long-Term Value: Ten Essential

118 A Tale of Leadership in Value Creation

oes private equity contribute to financial fragility during economic crises?

operational metrics, employee headcounts shrink and wages suffer.1

Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 21

22 Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020

repos, overdrafts, and commercial paper. The other three

The data are available at the official statistics of the Bank of

2007q1 2008q1 2009q1 2010q1

Source: UK Office for National Statistics

2004.01 2006.01 2008.01 2010.01

Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 23

24 Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020

Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 25

UK France Germany Europe US UK France Germany US Europe

Panel C: Average Deal Value (€ millions) Panel D: Number of Transactions

UK France Germany US Europe UK France Germany US Europe

26 Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020

B. Firms’ Characteristics in 2007

C. Firms’ Trends in 2007

Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 27

28 Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020

2004 2005 2006 2007 2008 2009 2010 2011

Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 29

30 Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020

A. Results Using Investment as an Outcome

Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 31

Post*High dry powder 0.105** 0.086** 0.053* 0.062** 0.070*** 0.055**

Post*1(Fund 02-07) 0.075* 0.090** 0.064** 0.073*** 0.036* 0.030

32 Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020

Sample 2004-2009 Whole sample Whole sample

Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 33

(1) (2) (3) (4) (5) (6) (7) (8)

34 Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020

Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 35

B. Fund Size (Assets Under Management)

Less than US$1b 43.24%

36 Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020

Avg. More during Less during

Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 37

38 Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020

occurred already in 2008 (at the peak of the financial crisis),

this pattern more directly by estimating our standard model

happened soon after the onset of the crisis.

approximately 15% fewer observations in the sample.43 Even

43 In our sample, about 310 firms exit before 2011.

Journal of Applied Corporate Finance • Volume 32 Number 3 Summer 2020 39

Panel A: Operational Activities

Assist portfolio companies with operating problems

0% 20% 40% 60% 80% 100%

PostHigh dry powder 0.105 0.086 0.053 0.062 0.070* 0.055**

Post1(Fund 02-07) 0.075 0.090 0.064 0.073*** 0.036* 0.030