PE Fund Structures and Incentives David T. Robinson

(1)

PE Fund Structures and Incentives

David T. Robinson Professor of Finance

J. Rex Fuqua Distinguished Professor of International Management Duke University, Fuqua School of Business

National Bureau of Economic Research Per Strömberg Professor of Finance and Private Equity, SSE

SHoF, May 23, 2019

(2)

‹#› Page 2

The PE Value Proposition

• Idea: Increase value through active ownership.

• Financial, Governance, and Operational engineering

• Kaplan & Strömberg (2009)

• Not a zero-sum game!

• Fundamental conflict between active ownership and liquidity

•  Difficult to replicate in a public setting

• Research evidence: PE ownership adds value to portfolio companies

• Productivity, profitability, growth, business practices etc.

• Feeds back into higher investment returns to LPs if

• GP has “edge”

• Value-added are not fully priced in acquisition premia

• GPs do not capture all gains through fees

(3)

Conflicts of interest between LPs and GPs

Investors (LPs)

Private equity fund (GPs)

Firm

New governance problem

Traditional governance

problem SOLVED

???

(4)

Traditional fund structure (I)

marketing draw down/investment Realisation or returns and exit extension

marketing follow-on fund

10 years

1 year 2 years

Cash flows back to investors Indications of fund performance commitments by investors

multiple ‘closings’

most private equity is invested via

partnerships of a limited duration

(5)

PE FIRM

General Partner

THE FUND

Limited Partnership

INVESTORS

Limited Partners

Have all decision rights Receive annual fees, often 2% of committed capital Share a “Carried Interest”

of 20% of profits (above some hurdle rate)

Provide cash as needed Pay management fees

Receive cash back plus 80% of profits

Traditional fund structure (II)

(6)

Almost 5 trillion USD in these fund structures (June 2017)

11/15/2021 6

Fundtype DryPowder UnrealizedValue Total %oftotal

(USDBn) (USDBn) (USDBn)

Private Equity (incl. Distress) 1,077.6 2,053.8 3,131.4 64%

Real Estate 245.5 565.4 810.9 16%

Private Debt (excl. Distress) 126.9 208.2 335.0 7%

Infrastructure 149.3 268.4 417.7 8%

Natural Resources 70.9 158.2 229.1 5%

Total 1,670.2 3,254.0 4,924.2 100%

Source: Döskeland & Strömberg (2018), based on Preqin data.

(7)

Page 7

Why this structure?

• Incentive alignment between LP – GP – PC

• Carried interest and GP investment in fund

• Mirrors equity participation of PC management teams

• “Modest” management fee covers fund expenses, which are particularly high during investment period

• GP has discretion in making investment decisions  LPs need protection.

• Need to “show me the money” in 10 years

• Need to prove some progress before can raise more funds

• Leverage provides PC incentives and tax benefits, and economizes on

expensive fund equity

(8)

‹#› Page 8

PE net fund returns have exceeded the public index

BuyoutPMEs (701funds)

VCPMEs (1085 funds) Average

(S&P 500)

Median (S&P 500)

Weighted average (S&P 500)

Average (S&P 500)

Median (S&P 500)

Weighted average (S&P 500) Whole pd

Direct alpha 1.20

3.07% 1.14

2.40% 1.25

3.16% 1.35

2.07% 0.97

-2.93% 1.46 0.47%

2000s 1.23 1.19 1.28 0.96 0.81 0.99

1990s 1.23 1.16 1.25 2.05 1.26 2.26

1980s 1.16 1.09 1.25 0.89 0.76 0.98

Kaplan-Schoar Public Market Equivalents (KS-PME) Burgiss data as of June, 2014.

U.S.-focused funds, vintages 1984-2010

Source: Harris, et al, (2015)

(9)

‹#› Page 9

Some evidence of persistence of top PE firms

• Overstated? Previous fund performance not known at time of fundraising

• Understated? LPs have access to more info than just past performance (Hüther et al, 2017.)

• Harris et al (2014) show that persistence has decreased significantly for buyout funds in post- 2000 vintages (while it remains strong in VC).

97

Table 4.8: Evidence on performance persistence in PE funds

Source: Harris et al (2014).

There are some important caveats to these findings. First, these persistence results do not adjust for differences in systematic risk loadings across PE fund managers. Adjusting for different risk loadings has been shown to be very important when assessing the performance persistence of mutual funds and hedge funds (Wermers, 2011). This is understandable, given the lack of a well-established methodology to measure systematic risk in PE. Instead, analysts use relative percentile rankings across groups of PE managers that are believed to follow similar strategies, such as VC vs buyout, or different geographical focus.

Second, comparing performance persistence across time, Harris et al (2014) find that the performance persistence of buyout funds have gone down. Persistence is much lower in post-2000 buyout fund vintages, and is no longer economically significant (see Table 4.9). On the flip side, they show that the top three quartiles solidly outperforms the public stock market index post-2000, and even the fourth quartile breaks even on average, which indicates that the importance of fund-picking to beat the public market has

become less important for buyouts. Also, the strong performance persistence in VC remains, however, and they show that only the top 2 quartiles have beaten the public markets.

97

Table 4.8: Evidence on performance persistence in PE funds

Source: Harris et al (2014).

There are some important caveats to these findings. First, these persistence results do not adjust for differences in systematic risk loadings across PE fund managers. Adjusting for different risk loadings has been shown to be very important when assessing the performance persistence of mutual funds and hedge funds (Wermers, 2011). This is understandable, given the lack of a well-established methodology to measure systematic risk in PE. Instead, analysts use relative percentile rankings across groups of PE managers that are believed to follow similar strategies, such as VC vs buyout, or different geographical focus.

Second, comparing performance persistence across time, Harris et al (2014) find that the performance persistence of buyout funds have gone down. Persistence is much lower in post-2000 buyout fund vintages, and is no longer economically significant (see Table 4.9). On the flip side, they show that the top three quartiles solidly outperforms the public stock market index post-2000, and even the fourth quartile breaks even on average, which indicates that the importance of fund-picking to beat the public market has

become less important for buyouts. Also, the strong performance persistence in VC remains, however, and they show that only the top 2 quartiles have beaten the public markets.

Source: Harris, Jenkinson, Kaplan, and Stucke (2014)

(10)

marketing draw down/investment Realisation or returns and exit extension

Higher fees during investment period

10 years

1 year 2 years

Incentives to hang on to investments longer, esp.

for laggard funds Mgmt fees change

from committed to invested capital

Lower fees when less capital is managed

Conflicts of Interest (I)

Evidence of “forced”

buying and selling

Evidence of “forced”

selling

marketing High profit

margin on mgmt fee for larger funds

(11)

Conflicts of Interest (II)

$2

$40 $40

$80 $80

$12

“whole fund” carry timing ensures

that LPs get paid first, but the

catchup period can warp exit

incentives

(12)

$2

$40 $40

$80 $80

$6

Conflicts of Interest (II)

“deal-by-deal” carry seems too

generous to GPs, but is associated

with better performance

(13)

Gross Net Notes:

Raw 0.407 0.329 Table 3

History 0.406 0.357 Table 6, Column 1

Vintage 0.322 0.252 Table 6, Column 2

Contracts 0.277 0.219 Table 6, Column 6

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45

Raw History Vintage Contracts

Gross Net

Conflicts of Interest (II)

“whole fund” carry seems to induce

early exit behavior to “put points on

the board” and then a rush to exit

once the fund is in the carry

(14)

Conflicts of Interest (III)

0 50 100 150 200 250

3 4 5 6 7 8 9

Total value of deals ($bn)

Total debt/EBITDA

Purchase Price / EBITDA

0 50 100 150 200 250

3 4 5 6 7 8 9

Total value of deals ($bn)

Total debt/EBITDA

Purchase Price / EBITDA

.511.52

1985 1990 1995 2000 2005 2010

Vintage Year

Value-weighted vintage-year PME (Preqin universe) Median Log D / EBITDA in fund vintage (sample)

.511.52

1985 1990 1995 2000 2005 2010

Vintage Year

Value-weighted vintage-year PME (Preqin universe) Median Log D / EBITDA in fund vintage (sample)

• Carry based on ROE (not ROA)  incentives to take on leverage in PCs

• Higher use of leverage associated with lower (!) returns to LPs, especially in credit booms

Source: Axelson, Jenkinson, Strömberg, Weisbach (2013)

(15)

‹#› Page 15

PE fund fees are substantial

• All-in fee estimates for

primary commitments vary between 5-7% of invested assets

• Scope for increased LP returns by lowering fees

• even at expense of lower gross alpha

Source: McKinsey (2017), “Equity investments in unlisted

companies”

(16)

Page 16