Electronic copy available at: http://ssrn.com/abstract=999910
Beware of Venturing into Private Equity
Ludovic Phalippou is Assistant Professor of Finance, University of Amsterdam Business
School, Amsterdam, The Netherlands. His e-mail address is <firstname.lastname@example.org>.
Electronic copy available at: http://ssrn.com/abstract=999910
A large literature has pointed out that publicly owned companies may suffer from a
separation of their ownership by dispersed shareholders and their day-to-day control by
managers. This separation of ownership and control leads to a number of so-called agency
conflicts, in which management may not act in the best interests of shareholders.
Because a private equity fund buys 100 percent of the company and controls it, it has
often been argued that the arrangement will reduce these agency problems (Jensen, 1989;
Shleifer and Vishny, 1997).
But while private equity funds have full control of companies, the
fund itself is acting on behalf of outside investors. In a typical scenario, a private equity buyout
fund buys a company by borrowing money from banks and by using cash provided by a small
group of large investors such as university endowments or pension funds. The companies
targeted for buyout could be listed on a stock exchange (for example, Nabisco in 1989 or
Hospital Corporation of America in 2006) or be privately held (Hertz Corp. bought from Ford
in 2005 or Warner Music bought from Time Warner in 2004). This new private equity fund
governance structure may ameliorate some agency conflicts, but it may also introduce new
ones. As a step towards understanding whether private equity buyout funds reduce overall
agency conflicts, this paper describes the contracts between funds and investors and the return
earned by investors.
The paper sets the stage with a puzzle: the average performance of private equity funds
is above that of the Standard and Poor’s 500 – the main public stock market index – before fees
are charged, but below that benchmark after fees are charged. This fact leads naturally to a
discussion of the institutional background of buyout funds, beginning with the compensation
contracts between the fund and the investors. Next, it covers how buyout funds report their
returns, explaining terms like “multiples” and “internal rates of return.”
The average private equity buyout fund charges the equivalent of 7 percent fees per
year, despite a return below that of the Standard and Poor’s 500. Why are the payments to
private equity funds so large? Why does the marginal investor buy buyout funds? I explore one
potential – and probably the most controversial – answer; that is, some investors are fooled. I
show that the fee contracts are opaque. The compensation contracts for buyout funds typically
imply lower fees at first sight than actually occur. The larger fees are generated by what seem
This article will focus on private equity funds that operate buyout funds, but in some cases,
private equity funds are also involved in venture capital, or in various kinds of mezzanine
(debt), infrastructure or real estate funds.
like minor details in these contracts. Investors may thus underestimate the impact of fees. I also
show the different aspects of the fund raising prospectuses that can be misleading for investors.
I then discuss whether investors in private equity funds learn over time or whether the “low
performance – high fee” situation may be persistent. Finally, to further understand the potential
agency conflicts between buyout funds and their investors, I discuss a few features of buyout
contracts that exacerbate conflicts of interest, rather than mitigate them. For example, several
contract clauses provide incentives that can distort the optimal timing of investments, of their
leverage, and of their size.
The conclusion emphasizes that these problems with buyout funds should perhaps not
be considered too surprising, given that similar issues arise in other common investment
vehicles like many mutual funds and hedge funds.
The Puzzle: Low Average Returns on Private Equity Funds
The two studies with the most extensive and most representative samples of cash flows
going from and to investors of private equity buyout funds are Kaplan and Schoar (2005) and
Phalippou and Gottschalg (2008). Kaplan and Schoar have data on 169 funds that are U.S.-
based or U.S.-focused. Phalippou and Gottschalg include the same funds, plus non-U.S. funds
for a total of 314 funds. Both studies find that the average performance of buyout funds is
below that of the Standard and Poor’s 500, after fees are taken into account.
The findings of these studies contrast with the glowing reports of high returns from
private equity industry associations although the underlying data are the same. Phalippou and
Gottschalg (2008) show that three elements contribute to the discrepancy. First, industry
associations use a sample of funds that slightly over-represent good funds. Second, they treat
the self-reported accounting values of funds as market values. This approach imparts an upward
bias, because a number of poorly performing, old and inactive funds report surprisingly high
accounting values for their on-going investments. Third, they often use the internal rate of
return as a performance metric, which is particularly misleading in the buyout context for
reasons explained below.
An incompletely resolved issue here is whether the lower returns from private equity
might reflect a lower level of risk. Driessen, Lin and Phalippou (2008) devise an econometric
approach to estimate the risk of funds such as buyout funds and find that the buyout funds in
their sample have a lower risk than public equity. However, they have a relatively small sample
of buyout funds which makes this estimate imprecise. In any case, it seems unlikely that most
investors expect an average performance of private equity buyout funds, after fees have been
charged, below that of major stock indices.
The investors in buyout funds “commit” a certain amount of capital at fund’s inception
and cannot add capital during the fund’s life. Fund managers search out investments during the
first five years of the fund, known as the “investment period.” When an investment is
identified, the fund “calls” the necessary capital from the investors. The companies held by the
fund are called “portfolio companies.” Every two to five years, the private equity buyout firm
which runs the buyout funds may seek to raise a new fund. When raising a new fund, the firm
will send a fund-raising prospectus that contains the track record, contract, biographies, strategy
and case-studies on the past investments of the firm. The data I use in this study comes from a
proprietary collection of fund-raising prospectuses. It contains information on 6,000
investments made by buyout funds from 1971 to 2007 in 30 countries.
This section begins with a composite of the key compensation terms for a buyout fund.
It then offers a concrete, if simplified example, of how a buyout firm would compute fees and
report its performance.
A Composite Fee-Contract
There are four sets of fees in a typical, if simplified, fee agreement for a private equity
buyout fund. The technical vocabulary used here is the same as in the original contracts.
First, the annual management fee is 2 percent of capital commitments until the end of
the five-year investment period. Thereafter, the management fee is 2 percent of funded capital
commitments outstanding. The management fee is payable semi-annually in advance. In
addition, the investor bears all organizational expenses incurred in the formation of the fund
(for example, legal, travel, accounting, and filing expenses).
Second, carried interest is an incentive fee based on the returns earned by the buyout
fund. Of the capital divested by the fund, 100 percent goes to the investors until the cumulative
distribution to investors equals an “internal rate of return” (a term explained in more detail in
the next section) of 8 percent per year. This 8 percent rate is calculated annually based on the
sum of two components: i) the capital contribution used to acquire all realized investments, plus
the (proportional) write-downs of unrealized investments; and ii) all expenses including
management fees allocated to the realized portfolio investments. Once investors have received
8 percent annual return, 100 percent of additional returns goes to the private equity firm until it
has received 20 percent of the difference between total amount distributed and the sum of the
two components just mentioned (the “catch up provision”). At that point, 80 percent of returns
have gone to investors and 20 percent to the private equity firm, and any additional returns
above that level are also divided 80:20. Finally, most contracts have a “claw-back provision,”
which holds that on termination, if the final carried interest due is lower than what was
received, the excess amount is returned to investors.
Third, portfolio company fees are taken directly out of the portfolio companies and so
not directly visible for investors. These include a number of expenses: i) transaction fees when
purchasing and sometimes selling a portfolio company, ii) expenses related to proposed but
unconsummated investments, iii) taxes, expenses of accountants, litigation, counsel, annual
meetings, iv) advisory and monitoring fees, and v) director fees. These fees are quite opaque.
Contracts specify neither the amount nor when such fees will be charged. In this paper, I
concentrate on the two components that Metrick and Yasuda (2008) document with practitioner
interviews. They report that usually transaction fees are 2 percent of transaction value below
$100 million, 1 percent of transaction value for the next $900 million and 50 percent of these
expenses are used to offset management fees. These are charged at entry but can also be
charged when exiting an investment. Next, Metrick and Yasuda report that typical monitoring
fees are 0.40 percent per annum on the value of the firm, for five years, irrespective of the
actual length of the investment, and 80 percent of these expenses are used to offset management
fees. For the example that follows, I ignore other portfolio company fees and account only for
transaction fees and monitoring fees. Also, to be conservative, I assume that firm value stays at
cost. In practice, firm value may increase over time and so would monitoring fees.
Fourth, a number of extra fees or costs can be imposed. For instance, cash proceeds can
be kept up to three months before being distributed to investors. Also, distributions to investors
can be in kind and with restrictions rather than cash, which creates an extra cost for the
investors. Investors may also pay penalties for selling their stakes or missing a capital call. In
the calculation below, I also ignore these extra fees.
An Example – The representative fund
Now consider applying this standard contract to a representative buyout fund. The fund
has $250 million of capital committed and makes two investments of $110 million each. One is
made at the end of the second year and one at the end of the fourth year. This schedule captures
the fact that private equity funds typically invest the committed capital slowly over time
(Ljungqvist and Richardson, 2003). Each investment is leveraged three times—that is, for $1 of
equity invested, $3 is borrowed (this is the average debt-to-asset ratio according to the Standard
& Poor’s Leveraged Lending Review). Both investments return 10 percent per year for five
years. The cost of (risky) debt is 7 percent per year. These numbers are chosen to match the
representative fund performance described above. For simplicity, tax rate is set to zero and the
fund does not change the net asset value of its portfolio companies over time. Table I shows the
resulting stream of cash flows and fees.
The first column shows the management fees paid every six months. With $250 million
of capital committed, the 2 percent management fee is $5 million per year—or $2.5 million
every six months. This fee is then reduced starting June 1986 to 2 percent of equity invested –
or $2.5 million per semester until December 1987 and $1.1 million per semester until
December 1989. At inception (in December 1980) the organizational fee of $1 million is added
to the management fee. Note that after the first investment is made (in December 1982),
management fees are significantly smaller. This is because the fees taken directly on the
portfolio company partly offset the management fees due (as described above).
The second column shows the carried interest. This is taken directly out of the exit
proceeds. Out of the $110 million of capital called from investors, $104.6 million is invested
(after transaction fees) and $330 million is borrowed. The value of the company starts thus at
$104.6 million plus $330 million and grows at 10% per year while paying $0.88 million of
monitoring fees every semester. After 5 years, the company is worth $688.92 million and the
debt is worth $462.84 million, hence the equity is worth $226.08 million. The management fees
related to this investment sums to $20.35 million. The rate of return is above 8 percent (the
hurdle rate), hence a carried interest is received. It amounts to 0.20 x (226.08-110-20.35) =
The third and fourth columns are the portfolio company fees received by the fund
managers. The monitoring fees are 0.4 percent per year: thus, 0.002 x $440 million = $0.88
million every six months. 80% of this fee ($0.7 million per semester) offsets management fees.
Transaction fees are 2 percent of value under $100 million and 1 percent of value above that
amount. So, transaction fees at entry are (0.02 x 100) + (0.01 x 340) = $5.4 million. Half of this
amount ($2.7 million) offsets management fees. This is why there are no management fees due
in December 1982 and only $1.6 million due in June 1983 ($2.5 million due minus $0.2 million
of transaction fees still to be refunded minus $0.7 million of monitoring fees refunded).
In a row near the bottom of the table, I report the net present value of each fee. I use 5
percent as a discount rate for all fees except carried interest for which I use 10 percent discount
rate. The different discount rates are ad-hoc correction for risk.
The portfolio company fees are the largest fees. Again, investors do not pay these fees
directly. They are taken directly from the investments. They are thus the least salient and, as we
saw, the most opaque of all as their amount is not even specified ex-ante. In Table 1, they
amount to $23 million of net present value, which is slightly more than a third of the net present
value of all fees. The other two fees are also about one third of the total. In their quantification
exercise, Metrick and Yasuda (2008) also find each fee type represent one-third of the total.
This example ignores the fact that some investors will be invited by the fund to co-
invest in some deals and that no fees are paid on these co-investments. If an investor gives $100
million to a fund and gets invited to co-invest in all the deals at the same proportion as the fund
for another $100 million, then the fee bill is essentially divided by two. Both the likelihood of
being invited and the quality of the co-investments increase with how useful an investor is for
the fund. For instance, investors such as the Yale and Harvard endowments may be useful
because, thanks to their alumni network, they may provide funds with proprietary investment
opportunities or with executives for the portfolio companies. They are also useful when
marketing a new fund; that is, if the Yale or Harvard endowment commits money to a fund,
other investors are more likely to invest. Some investors may also be useful for their political
connections. For example, if a U.S. private equity firm invests in another country in a
politically sensitive company, it may co-invest with a large local pension fund it already has as
an investor. Certain investors, therefore, pay lower fees than other investors and may also get
higher gross-of-fees returns if they get invited to co-invest in better deals.
An investor cannot know in advance the quality and quantity of investments it will get
invited to co-invest in. Yet that often has a large impact on the fee bill and final performance.
Table 1 also shows what investors receive--the cash flow stream—both net of fees and
gross of fees. Measuring the performance of an investment based on a stream of cash flows is
not obvious. Textbooks often recommend using a Net Present Value, but in practice two other
metrics are used: an “internal rate of return” and a “multiple”. The internal rate of return is the
discount rate that makes the Net Present Value equals to zero.
The “multiple” is the
undiscounted sum of the cash flows received divided by the sum of cash flows paid. Table 1
shows that investors receive a multiple of 1.68, an internal rate of return of 11 percent, and a net
present value close to zero. Gross of all fees, the multiple is 2.26, the internal rate of return is
18 percent, and the net present value is $75 million.
It is trivial to compute the net present value of fees but less obvious to give a more
intuitive measure of fees such as a percentage per year of amount invested. One way is to
compute how much a standard equity mutual fund would have to charge per year to collect the
same net present value of fees as the buyout fund. For our working example, we assume that
this mutual fund is exactly like the buyout fund. It invests the same amount at the same time,
use the same amount of debt, has the same return on asset but charges a yearly 1% of equity
value fee. The present value of the fees collected by this fund is $8.8 million. To compute such
a present value, I use 16% discount rate, which is the internal rate of return achieved by the
Hence, it is the solution to the equation
I am thankful to Jeremy Stein for this suggestion.
The total fees collected are above 7 percent per year. They appear an excessive
compensation given that investors are underperforming the Standard and Poor’s 500 index.
Cross-section of fees
The fee for the representative private equity buyout fund shown in Table 1 is in line
with what both Metrick and Yasuda (2008) and Phalippou and Gottschalg (2008) report for the
average fund in their sample, so the simple calibration exercise above satisfactorily captures the
essence of actual cash flow data. To get a sense of the actual empirical distribution of fees
across funds, I now use the sub-set of fund-raising prospectuses for which both internal rate of
return net-of-fees and gross-of-fees are reported. I observe 98 observations of pairs gross and
net internal rate of return. Results are shown in Figure 1.
The relation between fees and gross internal rate of return is close to perfectly linear—
specifically, a simple linear regression produces an r-squared of .7. The slope represents the
actual pay-for-performance contract. The slope of 0.25 is higher than the typical carried interest
of 20 percent. This difference is due to the way in which transaction fees rise with the
performance of the buyout fund, as mentioned below.
To preserve anonymity of the data, I group observations as a function of performance. I
define seven groups which all contain between ten and twenty observations. The first group
corresponds to gross internal rates of return below 20 percent, the second group corresponds to
gross internal rates of return between 20 and 30 percent, the third group corresponds to gross
internal rates of return between 30 and 40 percent, etc. The plot shows how regular the relation
Why Some Investors May Misunderstand
The average payment received by buyout funds appears excessive. How then have such
funds continued to attract investors? This section discusses some potential reasons behind
investor misperceptions. It begins with arguing that while compensation contracts for private
equity funds may appear similar, they often differ in crucial details in a way that often generates
larger fees than expected. Moreover, current standards for reporting the performance of buyout
funds tend to produce exaggerated performance figures, while the relevant information needed
to correct the potential exaggeration is typically “shrouded.”
Variations within Similar-looking Fees
Fees for private equity buyout funds are typically described as being 2 percent
management fees and 20 percent incentive fees, with a hurdle rate of 8 percent. As such, the
fees appear the same across buyout funds and roughly resemble the fees of other so-called
alternative asset classes. However, compensation contracts for private equity are long and
complex, and include details that often lead to fees higher than the basic structure of the
contract might suggest. These contractual details also vary across buyout funds, which creates
significant dispersion in fees across funds.
As a first example, consider the basic management fees for buyout funds of 2 percent.
Such fees are typically charged on the committed capital, not on capital invested by the fund.
Ljungqvist and Richardson (2003) and Phalippou and Gottschalg (2008) report that on average,
half of the capital committed to a private equity fund is actually invested. Hence a 2 percent fee
on capital committed is the same as a 4 percent fee of capital invested, which means that the
actual management fee bill is significantly higher than what may be thought at first sight.
Contracts do vary on this provision; for example, some firms charge the management fee on
some combination of capital committed and invested. As a result, although most buyout funds
charge a 2% management fees, the effective amount charged varies dramatically across funds.
As another example of the importance of details, consider incentive fees. The broad
outline of the incentive fees almost never varies from 20 percent of profits and an 8 percent
hurdle rate, but much variation arises in the details of their calculation. Some funds start
receiving carried interest when they have returned 8 percent per annum on capital committed;
for others, it is when they have returned 8 percent per annum on exited capital (like the example
above). In the past, some funds would receive carried interest separately for each investment
made, which is an expensive detail for investors. Some funds have a “claw-back provision” as
described above for returning what turns out to be excess carried interest payments to investors;
some do not. Some buyout funds pay accrued interests when refunding the carry; some do not.
Another related “detail” is the hurdle rate. It is almost always 8 percent but it can be soft or
hard. The soft version is the one described above and lead to a payment of 0.20 x (229-110-
20.35) = $19 million. The hard version would lead to less than half that payment: 0.20 x (229-
(110+20.35)*(1.08)^5) = $7.5 million. All these details thus make a large difference for the
incentive fees paid.
Details like the refunding rule is important for portfolio company fees. Given the
magnitude of these fees, different refunding rules generate very different fee bills. For
transaction fees, Metrick and Yasuda (2008) report that 80 percent of the contracts require fund
managers to share a portion of transactions fees with investors. One-third of the funds refunds
all transaction fees to investors; one-third refunds 50 percent; and the remaining refund an
amount in between these two numbers. For monitoring fees, Metrick and Yasuda (2008) report
that most funds refund 80 percent to investors.
Nonetheless, an important detail is missing in the contract. It is never specified how
much will be charged in the first place. To get a sense of how much portfolio company fees
may be charged, investors ask private equity firms for the list of previous investments and the
list of related portfolio company fees charged. Investors say that some firms refuse to provide
such information and when they look at those that provide it there is wide discrepancy in the
rates applied for portfolio company fees.
This fact is puzzling. How can investors accept such contracts? When asked, some
investors ignore (voluntarily or involuntarily) such details. The investors who do not ignore
them say that if a fund charges too much portfolio company fees its return is negatively affected
and it may upset its current investors. Hence, it would raise less money in the future, thus
collect less fees in the future. It obviously seems like a fragile equilibrium especially if there is
a permanent downturn in the buyout fund raising industry. Note also that this means that a good
performing fund will probably charge more portfolio company fees than a poorly performing
fund. Hence, the proportion of incentive fees in the total fee bill may be larger than at first
Most fees and costs imposed by private equity buyout firms on their investors are
complex and contain a multitude of dimensions. It is thus difficult for investors to compare
different contracts, to anticipate accurately the magnitude of fees.
Shrouded Negative Internal Rates of Return
When displaying the performance of past individual investments, private equity buyout
funds always report multiples, but not always internal rates of return. Table 2 shows that the
poorer the performance (according to the multiple) the more likely it is that the internal rate of
return is missing. When the multiple for an individual investment is less than 0.1 (that is, less
than 10 percent of the money is returned to investors), then the internal rate of return is missing
in more than 80 percent of the cases. As the multiples get better, internal rates of return are
more frequently reported. When the multiple is above 2, internal rates of return are missing
only in 10 percent of the cases (and these are from funds that do not report any internal rate of
return). If the multiple is exactly zero —in which case the internal rate of return should be -100
percent--then the internal rate of return is not reported in 98 percent of the cases.
Shrouded accounting information and keeping losers at cost
About half of the investments listed in the average fund-raising prospectus for a private
equity buyout fund are fully realized. The performance of the other half of the investments is
computed using valuation standards applied by the fund itself. Obviously, aggressive
revaluations of on-going investments could exaggerate past performance. However, the main
issue appears to lie with the valuations that are not updated; specifically, the performance of
poorly performing investments is rarely acknowledged by a write-down. The related contractual
provision typically states: “Investments will be valued at cost unless a material change justifies
a different valuation.”
In my sample of private equity buyout investments, about one-third of the investments
are losses and two-thirds are gains. However, at age 1, investments reported at a gain are five
times more numerous than investments reported at a loss; at ages 2 and 3, investments reported
at a gain are three times more numerous than investments reported at loss. Among investments
that are four years old or more, the proportion between gain and loss is steady and close to that
of the universe. This pattern implies that the performance of poorly performing investments in
fund-raising prospectuses is likely to be substantially exaggerated.
This pattern is only indicative, but it does coincide with some casual evidence. In 2008,
a regulation has been passed to enforce some common valuation standards. The Wall Street
Journal reported on May 5, 2008, that the stock price of a private equity firm called American
Capital had decreased substantially (Eavis, 2008). This decline was attributed to the claim that
poorly performing investments held by this firm had previously were valued at cost, but that the
new accounting rule was forcing it to place a market value on these investments—which would
cause the firm to report losses. Interestingly, one may interpret the decrease in stock price as the
benefits of loose accounting standards for the buyout firm. Usually, improved accounting
standards increase the value of companies as it decreases asymmetric information. One
interpretation is that private equity firms manage to trick investors thanks to lose accounting
rules and as a consequence better accounting standards reduce their value.
Using IRR flaws to improve performance
A private equity buyout firm that has raised several funds need not report the
performance of each fund separately, but instead can pool them arbitrarily. The arbitrary
grouping of funds can bias performance dramatically, especially when firms have had some
highly performing investments in their early days. By grouping all funds together, the
reinvestment assumption of the internal rate of return will kick in –that future payouts can be
reinvested and earn the same return as past payouts—and will hide bad recent investments (see
Phalippou, 2008). I find that about 50 percent of private equity buyout firms pool all their
investments when reporting results. A task for future research is to investigate the
characteristics of those pooling to see whether funds with early successes are more likely to
pool, but preliminary results are consistent with this assertion.
In addition, Phalippou (2008) describes in detail how computing average internal rate of
return – a common practice – produces an upward bias in the average performance. The basic
reason is that good fund have an internal rate of return of say 50 percent and bad funds have an
internal rate of return of 0 percent. If they have the same size, it is tempting and commonly
done to claim that the average fund performance is 25 percent. The problem is that it is
empirically clear that the good funds liquidate earlier than the old funds. Hence a good fund
may provide 50 percent internal rate of return for 5 years while the bad funds will provide 0
percent internal rate of return for 10 years. The representative investor will thus not receive 25
percent internal rate of return.
Using the internal rate of return also exaggerates the dispersion in performance of
buyout funds, because the difference between the internal rate of return and the effective rate of
return increases with the absolute level of the internal rate of return. The spread between top of
bottom funds is often an argument for investors to invest in this asset class despite poor average
performance. However, the large differences in performance across funds is largely due to the
use of internal rate of returns, which is an improper performance metric (see Phalippou, 2008).
Sample bias: good track records are shown more often
About 20 percent of the fund-raising prospectuses include the list of investments that
some partners were “responsible for” in funds they have worked for. These investments
performance is above-average (2.8 average multiple versus 2.1 average multiple for the rest of
the sample) and prospective investors see them several times (for all funds that will be raised by
the new employer and for all funds that will be raised by the old employer). This creates an
obvious sample bias. In addition, no rules specify who is “responsible for” an investment, and
hence fund managers have some discretion for the inclusion of past investments—which may
partly explain why these investments have better performance than average.
Certain fund-raising prospectuses even specify that they are selecting track records from
the past. For example, a prospectus can state that “the objective of the current fund is to invest
in U.S. firms with sales between $100 million and $500 million. The following track record
shows the performance realized in the past on the investments that correspond to our objective.”
Obviously, in retrospect, one can select future investment criteria that correspond to successful
past investments. Another observed case is a track record that is the one of “the current team,”
which implies that by reorganizing or firing, a fund can improve its track record. These
explicitly selected track records are relatively rare--less than 10 percent of the case--perhaps
because it is too obvious that some past investments have not gone well. However, such
selection is possible and done in an some audited fund-raising prospectus.
All these facts mean that prospective investors see more often the best returns, which
may contribute to a biased perception of the performance in the industry.
Key Shrouded Details: Duration, Leverage, Net-of-fees Performance and Fee Details
The prospectuses used by private equity buyout firms to inform their investors often
withhold important details. For example, prospectuses report multiples of cash flows received
over cash flows paid in, but a multiple without duration is of little use. A multiple of 2
generated over three years is a significantly better performance than a multiple of 2 generated
over eight years. In more than a third of prospectuses, information on month and year of entry
and exit of various investments is not provided. Even when investors are provided the dates of
entry and exit, little or no information is provided on cash flows in or out of the investment at
intermediate times. Less than 2 percent of the fund-raising prospectuses show a holding period
weighted by cash flow. Some funds (about 5 percent) also report the date of initial realization,
in which case the holding period shown is shorter than the actual one. Hence, accurate
information on duration is missing, often making multiples of little use.
In addition to duration, it is important to know leverage. Obtaining a multiple of two
with $1 borrowed for $1 invested is significantly better than obtaining a multiple of two with $2
borrowed for $1 invested. However, the extent of leverage seldom is mentioned on fund raising
Finally, the details of how much fees were charged in the past are never mentioned and
only 25 percent of the funds report overall past performance net of fees.
Economists are congenitally suspicious of any argument where people are consistently
fooled; after all, even if the prospectuses themselves are not transparent, even unsophisticated
investors should be able to request the data necessary to compute past performance net of fees,
and to learn from it.
Investors may indeed request additional information from buyout fund managers to
make the calculations for duration, leverage and net-of-fees performance of previous
investments--as well as to get additional details of the fee contract. Not many do so and, when
they do, the fund may refuse to provide this additional information, especially if the investor is
Nonetheless, even if investors have all the relevant data, simply measuring performance
and the impact of fees is difficult. This can be seen from the above discussion and further
details are provided by Phalippou (2008). To illustrate this point on a concrete example, I take
the statement on the website of CalPERS – one the largest investors in private equity (available
May 19, 2008). The headline says that “since inception in 1990 to September 30, 2007, the
Alternative Investment Management Program has generated $12.5 billion in profits for
CalPERS. Given the young, weighted-average age of the portfolio (3.5 years) this amount will
continue to grow as the portfolio matures.” In short, CalPERS seems pleased with its private
equity investments and continues to increase its allocation to private equity. However,
CalPERS has invested $25 billion and so far only $19 billion has been received back. The rest
of the gain represents accounting values not yet realized. If we look only at mature private
equity funds (those raised between 1990 and 2000) in order to avoid relying on accounting
valuations, the multiple for CalPERS is 1.5. Is this return a good one? If the investments are
held for three years on average, this multiple equals that of a U.S. stock-market index fund over
the same period. If the investments are held for five years – which is the average in the industry
– the rate of return for CalPERS would be a disappointing 8 percent. It is thus difficult to
understand why CalPERS would be so happy and the justification on their website is “short” to
say the least. So, is one of the largest investor in buyout funds learning from past return? Maybe
In more dramatic cases, learning has occurred. For example, several banks have
decreased (or stopped) their allocations to private equity and banks are identified in Lerner,
Schoar and Wong (2007) as consistently worst investors in private equity. Hence, when it is
obvious enough, investors may learn but when it is less obvious, the complexity of performance
measurement may seriously impair learning.
There may also be organizational reasons within financial markets that allow the
combination of low performance and high fees in private equity buyout funds to persist. One
possibility concerns agency conflicts that may exist within the investor organization. The
person in charge of private equity investments in an organization may be compensated based on
the reported (misleading) performance. Others in the investor organization may not be
acquainted enough with all the details of the private equity industry to monitor or to detect a
discrepancy between actual and reported performance, or may not have sufficient incentive to
Yet another possibility is an economy described by Gabaix and Laibson (2005) where
sellers choose to hide certain contractual details. As one example, banks do not compete on
costs such as usage fees for automatic teller machines, or fees for minimum bank balances.
Instead, banks “choose to shroud” these fees and customers do not learn about the “details of
the fee structure until long after they have opened their accounts.” Gabaix and Laibson also
show that the printer market operates in a similar way. The main part of the actual cost is
hidden from customers. They propose an explanation based on consumer myopia; that is, some
customers are just not paying attention. Those who do pay attention can avoid the higher costs,
and no firm has an incentive to be more transparent. Such reasoning may apply to private equity
buyout funds as well.
Contracts and Conflicts of Interest
To further isolate potential conflicts between the managers of private equity buyout
funds and their outside investors, I discuss a few features of buyout contracts that exacerbate
conflicts of interest, rather than mitigate them. First, managers have an incentive to time cash
flows in a way that will increase incentive fees. Second, certain contracts provide steep
incentives for shortening investment horizons. Third, transaction fees may distort choices of
buyout firms in terms of leverage, size of investment, and number of changes in capital
Carried Interest And Strategic Timing of Cash Flows
Assume a $500 buyout fund makes five investments of $100 each. After five years, the
market value of the first investment rises to $300 while the market value of the other
investments declines to $10. All investments are expected to grow at a rate of 10 percent per
year from that point onward. Still, for simplicity, assume that the fund manager can liquidate
investments either now or at the end of the fund’s life, in five years time. Finally, assume that
investors’ hurdle rate is below 10 percent, which means that investors would prefer that all
investments were liquidated at the end.
If the fund manager liquidates all the investments at the end, no carried interest will be
paid because the fund earned less than the 8 percent hurdle rate. If the fund manager liquidates
the first investment now, it receives 0.20 x (300-100) = $40 of carried interest which can be
invested for four years at say 5 percent per year. In five years, the manager needs to refund $40
to the investors (“claw-back” provision) minus the (say 50 percent) taxes paid on the $40. This
strategy leads to a final profit of 40 x 1.05^4 – 40 x 0.5 = $4.3. Hence, fund managers have an
incentive to liquidate the good investment immediately and to delay as much as possible the
exit of the poorly performing investments. Such incentives are rarely in line with the benefits of
Incentives to exit early
Private equity contracts often contain provisions that encourage early exit from
investments. For example, some contracts allow re-investment of capital coming from
investments that are shorter than 18 months. This provision provides a clear incentive for funds
to exit investments early because it offers them a chance to reinvest funds and thus effectively
increase the assets under management, hence fees. Distributing large payments early to
investors also often increase more the internal rate of return than the effective rate of return. As
funds are just on their internal rate of return instead of effective rate of return, they have an
incentive to distribute large dividend early in the fund’s life. Finally, exiting some investments
early may increase the collected carried interest as seen in the previous section.
On the other side, early exit from an investment can also decreases management fees
after the investment period, and may decrease total incentive fees. From the standpoint of
manager compensation, the question of whether benefits of early exit outweigh the costs
remains an open empirical question. However, evidence that suggests that private equity funds
sell their portfolio companies at a larger discount than other sellers (Lee and Wahal, 2004, and
Masulis and Nahata, 2008) may be interpreted as selling companies too early.
Out of the 2500 liquidated buyout investments in my dataset, 6 percent liquidated within
1 year, 23 percent liquidated within two years, and 44 percent liquidated within three years.
Hence almost half of the investments are held for less than three years. Interestingly, the short-
term investments have similar multiples as the long investments, but significantly higher
internal rate of return (69 percent for short-term investments versus 10 percent for long
investments), which hints that more successful investments are being exited more quickly.
Transaction fee incentives
Managers of private equity buyout funds receive compensation for making a purchase
or changing the capital structure which are not always fully refunded to investors. These fees
may distort choices of the buyout fund in terms of leverage, size and number of transactions
Buyout fund managers may prefer to invest in many small deals, rather than a few large
ones, because they earn more per dollar invested when arranging small deals (usually 2 percent
of asset value) than when arranging large deals (usually 0.5 percent of asset value.) Similarly,
fund managers may earn higher fees by selling a company in ten different pieces rather than in
one piece. In addition, fees are charged on asset value and not on equity value. This provision
provides incentives to increase debt, because a larger deal means larger transaction fees.
Finally, all the pay-for-doing type of compensation pushes managers to make more changes
than would be optimal — like modifying capital structure.
In addition, contracts do not always fully regulate the relations between several funds of
the firm. As carried interest is perceived only if a fund is profitable, it may be fee-maximizing
to allocate anticipated high performance investments to profitable funds and not to unprofitable
funds or to sell investments at a discount from an unprofitable funds to a profitable funds.
All these distortions are however speculative in the sense that they are what a contract
allows or do not prevent but future research may hopefully measure empirically the actual
distortions if any.
The average private equity buyout fund has a low rate of return for its outside investors
after fees are subtracted, and charge more than 7 percent fees per year. This finding is
consistent with a statement by who is certainly the most knowledgeable private equity investor
(David Swenson, CFO of Yale endowment): “The large majority of buyout funds fail to add
sufficient value to overcome a grossly unreasonable fee structure.” This result in addition to the
observation that compensation contract buries in details costly provisions that cannot readily be
justified on the basis of proper incentive alignment means that it is premature to assert that the
agency conflicts are lower in private equity than in public equity. More research is needed
before we can make a clear judgment on the benefits and costs of private equity governance
structure compared to the alternative of public equity governance structure.
Also puzzling is the fact that investors keep buying aggressively buyout funds with such
contracts. One potential explanation is misleading information and insufficient expertise of
certain investors. Nonetheless, the complexity of the private equity fund arrangements make
difficult things as fundamental as measuring past performance, fees paid and risk. This
complexity may prevent learning and maintain the low-performance-high-fees situation.
Lest these judgments about buyout funds seem overly harsh, it is perhaps worth pointing
out that research has reached very similar conclusions about actively managed mutual funds
and hedge funds. An extensive literature holds that average fees for actively managed mutual
funds are too high given their performance (for example, Malkiel, 1995). However, investors
keep on purchasing actively managed mutual funds and part of the answer seems to lie in a too
optimist probability to choose top-performing mutual funds. In addition, a variety of conflicts
of interest between mutual funds and investors has been revealed.
A similar argument can be made about hedge funds as well. The average performance of
hedge funds appears close to that of public equity (Fung, Hsieh, Naik and Ramadorai, 2008). In
addition, research by Brown, Goetzmann and Liang (2006) points out some potential conflicts
of interest. Specifically, funds “either rely on portfolio managers to determine the final prices of
their own portfolios, conduct day-to-day accounting internally without a third party verification,
or lack a fraud-proof mechanism for authorizing money transfer.” Their results indicate that
“funds with deficient operating mechanisms suspiciously report higher than average risk-return
Of course, most of the findings in this paper concern the average private equity buyout
firm. Some buyout firms perform genuinely well, and some offer superior contracts. This paper
seeks only to extend a warning that those who consider investing in a private equity fund may
wish to examine the contract and the past performance of the fund in great detail.
To name but a few, researchers have documented the existence of “soft dollar” arrangements
where mutual fund managers does not choose the broker with lowest commission. Investors
indirectly pay the extra cost and the manager receives a kick-back from the broker (Conrad,
Johnson and Wahal, 2001), favoritism within fund families (Gaspar, Massa, and Matos, 2006),
late trading practices (Zitzewitz, 2006), and incubation practices which artificially generate
good track records (Evans, 2006).
I am especially grateful to the three editors (James Hines, Jeremy Stein, and Timothy Taylor)
for their kind guidance and extremely valuable comments. I am also thankful to Dion
Bongaards, Arnoud Boot, Jose Miguel Gaspar, Oliver Gottschalg, Raj Iyer, Pedro Matos,
Zacharias Sautner and especially Ayako Yasuda for their comments.
Brown, S.J., W.N. Goetzmann, and B. Liang, 2006, Hedge fund fraud: A due diligence
Conrad, J.S., Johnson, K.M., and S., Wahal, 2001, Institutional trading and soft dollars, Journal
of Finance 56, 397-416.
Driessen, J., T.C., Lin, and L. Phalippou, 2008, Measuring the risk of private equity funds: a
new methodology, NBER Working Paper #14144.
Eavis, Peter. 2008. “Rule Clouds American Capital; Revaluing of Loans Is Likely to Force Big
Write-Downs.” Wall Street Journal, May 5, C2.
Evans, R.B., 2006, Does alpha really matter? Evidence from mutual fund incubation,
termination and manager change, Working Paper, Boston College.
Fung, W., D.A. Hsieh, N.Y. Naik and T. Ramadorai, 2008, Hedge Funds: Performance, Risk,
and Capital Formation, Journal of Finance 63, 1777-1803.
Gabaix, X., and D. Laibson, 2005, Shrouded attributes, consumer myopia, and information
suppression in competitive markets, Quarterly Journal of Economics, 505-540.
Gaspar, M., Massa, M., and P. Matos, 2006, Favoritism in mutual fund families? Evidence on
strategic cross-fund subsidization, Journal of Finance 61, 73-104.
Jensen, M.C., 1989, Eclipse of the Public Corporation, Harvard Business Review (September-
Kaplan, S.N. and A. Schoar, 2005, Private equity performance: returns, persistence, and capital
flows, Journal of Finance 60, 1791-1823.
Lee, P.M. and S. Wahal, 2004, Grandstanding, certification and the underpricing of venture
capital backed IPOs, Journal of Financial Economics 73, 375-407.
Lerner, J., A., Schoar, and W., Wong, 2007, Smart institutions, foolish choices? The limited
partner performance puzzle, Journal of Finance 62, 731-764.
Ljungqvist, A. and M. Richardson, 2003, The cash flow, return, and risk characteristics of
private equity, NBER Working Paper #9495.
Malkiel, B.G., 1995, Returns from investing in equity mutual funds 1971 to 1991, Journal of
Finance 50, 549-572.
Masulis R. and R. Nahata, 2008, Venture Capital Conflicts of Interest: Evidence from
Acquisitions of Venture Backed Firms, manuscript.
Metrick, A. and A. Yasuda, 2008, Economics of private equity funds, manuscript.
Phalippou, L., 2008, The hazard of measuring performance with IRR: The case of private
Phalippou, L. and O. Gottschalg, 2008, Performance of private equity funds, Review of
Financial Studies, forthcoming.
Shleifer, A. and R.W. Vishny, 1997, A survey of corporate governance, Journal of Finance 52,
Swenson, D., 2005, Unconventional Success: A Fundamental Approach to Personal Investment.
Zitzewitz, E., 2006, How widespread was late trading in mutual funds?, American Economic
Review (Papers & Proceedings), 284-289.
This table shows the fees paid by investors for a representative fund with $250 million of
capital committed. It makes two five-years investments of $110 million each. Each investment
is leveraged three times ($1 of equity, $3 borrowed). The cost of debt is 7 percent per year.
Investments return 10 percent per year (on asset). The first column shows management fees, the
second column shows carried interest, the third and fourth columns are the portfolio company
fees. The last two columns show the cash flows received by investors (net-of-fees) and the cash
flows gross of all fees (net-of-fees cash flows plus the four fees). The bottom of the table shows
the net present value of all cash flow streams, yearly fee equivalents and practitioner
performance measures (multiple and internal rate of return).
Fees Paid by Investors to Fund Managers
Management Carried Portfolio Company fees Cash flows
Date Fee Interest Transaction Monitoring Net-of-fees Gross-of-fees
Dec-80 3.50 -3.50
Jun-81 2.50 -2.50
Dec-81 2.50 -2.50
Jun-82 2.50 -2.50
Dec-82 5.40 -110.00 -104.60
Jun-83 1.60 0.88 -1.60 0.88
Dec-83 1.80 0.88 -1.80 0.88
Jun-84 1.80 0.88 -1.80 0.88
Dec-84 5.40 0.88 -110.00 -103.72
Jun-85 0.20 1.76 -0.20 1.76
Dec-85 1.10 1.76 -1.10 1.76
Jun-86 0.80 1.76 -0.80 1.76
Dec-86 0.80 1.76 -0.80 1.76
Jun-87 0.80 1.76 -0.80 1.76
Dec-87 0.80 19.15 1.76 202.86 224.57
Jun-88 0.40 0.88 -0.40 0.88
Dec-88 0.40 0.88 -0.40 0.88
Jun-89 0.40 0.88 -0.40 0.88
Dec-89 0.40 19.15 0.88 202.86 223.29
Net present value 19.59 17.95 9.34 13.34 7.14 61.09
Yearly fee 2.23% 2.04% 1.06% 1.52%
Multiple 1.68 2.26
Internal rate of return 10.83% 17.64%
Missing poor performance figures
This table shows the fraction of internal rates of return that are not reported as a function of the
multiple of the investments (fraction missing). Statistics are based on the sub-sample containing
buyout investments that are either fully realized (liquidated) or partially realized. Investments
made in the United States and those liquidated are shown separately. The total number of
observations that fall in each multiple range is displayed below each fraction, in italics between
parentheses. The data comes from a proprietary dataset of buyout fund raising prospectuses.
Multiple range All US Liquidated
(0.0 – 0.1] 0.81 0.81 0.81
(166) (95) (98)
(0.1 – 0.5] 0.76 0.70 0.78
(367) (195) (178)
(0.5 – 1.0] 0.60 0.59 0.65
(402) (200) (120)
(1.0 – 2.0] 0.21 0.19 0.11
(1458) (641) (603)
(2.0 – 5.0] 0.10 0.10 0.08
(1539) (794) (1029)
(5.0 – … 0.10 0.10 0.08
(715) (476) (536)
All 0.26 0.24 0.19
(4647) (2401) (2564)
Cross-section of Fees
Annual fees are proxied by the spread between gross and net internal rates of return. There are
98 observations and equally weighted average of fees is 13 percent. The reported R-square and
slope refer to the results of an ordinary least squares estimation. Observations are pooled as a
function of performance (the first bar corresponds to gross internal rates of return below 20%,
the second bar corresponds to gross internal rates of return between 20% and 30%, etc.) There
are at least ten observations for each bar.