Phalippou & Gottschalg (2009)

Citation: Phalippou, L., & Gottschalg, O. (2009). “The Performance of Private Equity Funds.” Review of Financial Studies, 22(4), 1747-1776.

This paper challenges the widely held belief that private equity funds deliver superior returns. After correcting for several biases in how PE performance is typically measured, Phalippou and Gottschalg find that the average PE fund underperforms the S&P 500 by 3% per year net of fees, and 6% per year after risk adjustment.

The conventional wisdom problem:

Industry benchmarks and previous academic research (including Kaplan & Schoar 2005) suggested PE funds delivered returns comparable to or exceeding public markets. These findings supported the high fees charged by GPs (typically 2% management fee + 20% carried interest). Phalippou and Gottschalg show this performance is overstated due to three systematic biases.

Bias 1: Inflated NAVs

Funds self-report Net Asset Values (NAVs) for ongoing investments. The authors find that funds past their normal 10-year liquidation date still carry substantial NAVs for “living dead” investments—portfolio companies that have no realistic exit prospects but haven’t been formally written off.

• 46% of funds with positive NAVs had no change in NAV for at least 5 years and no cash flow activity
• Another 31% of NAVs were from funds inactive for 3+ years
• Writing off these obviously worthless NAVs reduces average performance by 7 percentage points

Bias 2: Sample selection

The commonly used Thomson Venture Economics dataset is based on voluntary reporting by LPs. Comparing funds in this dataset to funds observable only through investment-level data reveals:

• Funds in the “base sample” have 50% successful exits (IPO or M&A)
• Funds not in the sample have only 45% successful exits
• This 5% spread is economically significant and translates to 4 percentage points of PI

Bias 3: Improper weighting

Standard practice weights funds by capital committed, but what matters is the present value of capital actually invested. Poorly performing funds invest more slowly, so committed-capital weighting understates their drag on aggregate returns. Correcting this reduces performance by 2 percentage points.

flowchart TD
  subgraph Standard Performance
    S[PI = 1.01]
    S --> OUT[Slight outperformance vs S&P 500]
  end
  
  subgraph After Corrections
    C1[Write off NAVs: -0.07]
    C2[Sample bias: -0.04]
    C3[Proper weighting: -0.02]
    C1 --> F[PI = 0.88]
    C2 --> F
    C3 --> F
    F --> UNDER[12% value destruction]
  end

The fee picture:

The paper provides one of the first comprehensive estimates of total GP compensation using actual cash flow data (rather than simulations):

• Total fees represent more than 25% of invested capital (6% per year)
• Two-thirds of fees come from management fees, only one-third from carried interest
• This is because management fees are charged on committed capital, while only 16% of committed capital is invested after year one

Fee Structure	Impact on Alpha
2% → 2.5% management fee	-1.3%
8% hurdle rate provision	+0.3%
Post-investment fee reduction	+0.6%

Performance persistence:

Despite poor average performance, there is strong evidence of persistence:

• Top quartile funds do outperform the S&P 500
• Prior fund performance is the single best predictor of current fund performance
• When past performance is included in regressions, all other characteristics (size, experience, etc.) lose significance

The puzzle:

Given these findings, why do investors continue to allocate to private equity?

1. Learning. The industry is relatively young; investors may be learning which GPs have skill.

2. Misleading metrics. IRRs and multiples—the standard performance measures in prospectuses—systematically overstate returns. IRRs are biased upward because poorly performing funds have longer durations.

3. Side benefits. Some LPs (especially banks and corporate pension funds) may accept lower returns in exchange for business relationships with GPs—consulting work, underwriting, lending opportunities.

Key implications:

1. The 2/20 fee structure may not be justified by returns. The paper finds gross-of-fees alpha of ~3%, but fees of 6%. GPs capture more than their value-add.

2. Fee structure matters. Management fees dominate carried interest in total GP compensation. This creates misaligned incentives—GPs profit even when funds underperform.

3. Information opacity enables rent extraction. The difficulty of benchmarking PE performance (no public prices, complex cash flows, voluntary reporting) allows the industry to maintain a narrative of outperformance that the data don’t support.

4. Skill exists but is rare. Top quartile funds do generate alpha, and there’s persistence. But the average GP doesn’t justify their fees.

After correcting for biases, private equity funds on average destroy value for investors while extracting substantial fees. Some GPs have genuine skill—performance persistence confirms it—but the industry’s opacity allows underperforming managers to keep raising capital. The economics of asset management look very different once the measurement is done right.