In May 2012, three members of the Ewing Marion Kauffman Foundation's investment staff, Diane Mulcahy, Bill Weeks, and Harold Bradley, published "We Have Met the Enemy... and He Is Us," a detailed review of roughly twenty years of the Foundation's investing in venture capital (VC) funds. The Foundation began investing in VC funds in 1985 and had been a limited partner in more than 100 funds run by more than sixty general partners, many of which the authors describe as top-tier. As of February 29, 2012, its endowment stood at $1.83 billion, of which about $249 million was invested in or committed to venture capital and growth equity funds. The paper is now fourteen years old and its data run through 2011.
The authors' conclusion was aimed inward. They argued that the central problem was less the venture capitalists than the way institutional investors, including the Foundation itself, choose and pay them.
"Investment committees and trustees should shoulder blame for the broken LP investment model."
— Mulcahy, Weeks and Bradley, Kauffman Foundation, May 2012
This article summarizes what the paper found. It then adds what later academic research says about two of the paper's central themes, the cost structure of venture funds, much of which is fixed, and whether strong performance repeats, and notes where the evidence is less one-sided than the paper's headline.
A wide gap between the few and the many
The Foundation's data covered 99 venture funds with results measured through the end of 2011. Of those, 50 had a net multiple below 1x: their total value, realized and unrealized and after fees and carry, was below the capital the Foundation invested. Another 33 landed between 1x and 2x. Only 16 funds, roughly one in six, reached 2x or better, and six of those reached 3x or more. The average across all 99 funds was 1.31x. The sample includes funds raised as recently as 2011, which had not had time to mature, and the paper does not separate them, so results for those funds may change.
Net Multiple of Invested Capital, 99 Kauffman Venture Funds
Net of fees and carried interest, as of December 31, 2011
Mean net multiple across all 99 funds: 1.31x. 16 of 99 funds reached 2x or more.
Source: Mulcahy, Weeks, and Bradley, "We Have Met the Enemy… and He Is Us," Ewing Marion Kauffman Foundation, May 2012, p. 18.
Results shown are the actual, net-of-fees results of the Kauffman Foundation's venture capital fund portfolio as of December 31, 2011. Net multiple is total value (distributions plus remaining estimated value) divided by capital paid in, after fees and carried interest, as reported by the Foundation. They are not the results of any IFA account, strategy, or model, and are not hypothetical. Past performance is not a guarantee of future results. Individual investor accounts (performance, allocation, etc.) may vary.
View chart data as a table
| Net multiple (TVPI) | Number of funds |
|---|---|
| Below 1x (capital not returned) | 50 |
| 1x to under 2x | 33 |
| 2x to under 3x | 10 |
| 3x or more | 6 |
| Total | 99 |
The authors also found that fund size mattered. None of the Foundation's funds that raised more than $500 million returned more than twice the invested capital after fees. A Silicon Valley Bank study of 850 venture funds from 1981–2003 vintages, cited in the paper, found that 51% of funds larger than $250 million failed to return investor capital after fees, while smaller funds returned more than 2x about 34% of the time.
Measuring against the market, not against other funds
Venture investing ties up money for a decade or more. The authors regard it as riskier and costlier than a diversified stock portfolio, and they argue that an investor is only compensated for those burdens if a fund beats what the same dollars would have earned in public markets. For that reason, the Foundation came to prefer public market equivalent (PME) analysis, measured against the Russell 2000 Index, over the "top quartile" and "vintage year" rankings common in fund marketing. Those rankings compare a fund only with other venture funds, so a top-quartile fund can still trail the stock market, or even lose money.
The authors set their bar at an excess return of about 3 percentage points a year over the index, which for a ten-year fund corresponds to a PME of roughly 1.34. Of 96 funds from 1989–2010 vintages, 25 cleared it. The ten-year hurdle is shown for reference; funds from later vintages had not completed ten years. The results differed sharply by era.
Share of Kauffman Venture Funds That Beat the Russell 2000 by at Least 3 Percentage Points a Year
By vintage group; fund data through December 31, 2011
Outperformance measured by the public market equivalent (PME) method: each fund's actual net cash flows compared with the same dollars invested in the index on the same dates.
Source: Kauffman Foundation (May 2012), p. 45; percentages calculated from the fund counts shown. Benchmark: Russell 2000 Index, which tracks the performance of smaller-capitalization U.S. public companies.
Investors cannot invest directly in any index. Indexes are unmanaged and do not reflect fees or expenses. Results shown are the actual, net-of-fees results of the Kauffman Foundation's venture capital fund portfolio, not IFA accounts, and are not hypothetical. Group sizes are small (13 to 48 funds), and some of the more recent funds were still active when measured, so their results could change. Past performance is not a guarantee of future results. Individual investor accounts (performance, allocation, etc.) may vary.
View chart data as a table
| Vintage group | Funds clearing the bar | Funds in group | Share of funds |
|---|---|---|---|
| Before 1995 | 10 | 14 | 71.4% |
| 1996–2000 | 8 | 48 | 16.7% |
| 2001–2005 | 2 | 21 | 9.5% |
| 2006–2010 | 5 | 13 | 38.5% |
| All groups | 25 | 96 | 26.0% |
Ten of 14 funds raised before 1995 cleared the bar. For the 1996–2000 group, the figure was 8 of 48, and for 2001–2005 it was 2 of 21. In the authors' experience, it has been improbable since 1995 that a venture fund would deliver better net returns than the public market. That is a statement about one portfolio, not the whole industry, but it is consistent with a broader study: Harris, Jenkinson, and Kaplan (2014), examining nearly 1,400 U.S. buyout and venture funds, found that venture funds outperformed public equities in the 1990s but underperformed them in the 2000s.
Why the comparison, and what differs. Venture fund results are compared with the Russell 2000 Index to show whether investors were compensated for the illiquidity, cost, and risk of venture funds relative to a liquid public alternative. The two are materially different: venture funds are private, illiquid, and long-term, while an index is unmanaged, liquid, and cannot be invested in directly. The PME method applies a fund's cash flows to an index and does not adjust for differences in risk or liquidity. Small-company stocks carry their own risk, including the possibility of sharp declines.
What the fees pay for
Venture funds are typically paid through a management fee, usually about 2% a year of committed capital during the five-year investment period, and carried interest of about 20% of profits, the arrangement known as "2 and 20." The Kauffman paper cites a study of venture funds raised from 1993 to 2006 in which about 90% charged fees of 2% or more and 95% charged 20% carry. During the investment period, the fee is based on capital that investors have promised rather than capital invested or earned, so it is paid whether or not the fund performs, and it grows with fund size. On a $1 billion fund, 2% is $20 million a year. After the investment period, the fee usually steps down but continues through the fund's life.
Academic work on private equity compensation points the same way. Metrick and Yasuda (2010) built a model of expected manager revenue for 238 funds raised from 1993 to 2006 and estimated that about two-thirds came from components that do not depend on performance. For venture funds, the Kauffman paper reports the study's median estimates at $14.61 in fees versus $8.20 in carry for every $100 under management.
Estimated Median Manager Revenue per $100 Under Management, Venture Capital Funds
Model-based estimate for funds raised 1993–2006; estimates as published 2010, as reported by the Kauffman Foundation, May 2012
Median estimates of expected manager revenue over a fund's life, per $100 of fund capital.
Source: Kauffman Foundation (May 2012), p. 33, reporting estimates from Metrick and Yasuda, "The Economics of Private Equity Funds," Review of Financial Studies (2010).
These are model-based estimates of expected revenue to managers over a fund's life, not amounts actually paid, and are not investment performance.
View chart data as a table
| Revenue component | Dollars per $100 of fund capital |
|---|---|
| Management fees (fixed, not tied to performance) | $14.61 |
| Carried interest (performance-based) | $8.20 |
"...pays VCs more for raising bigger funds."
— Mulcahy, Weeks and Bradley, describing the "2 and 20" structure
The authors illustrated how these fees stack. In a hypothetical example using their own assumptions, a firm raises a $250 million fund, then a $350 million fund in year three, then a $500 million fund in year five. Annual management fee income reaches $19.5 million in year five and $22 million from year six, under the assumption that fees stay at 2% of committed capital for all ten years. This example is hypothetical, uses assumptions chosen by the Kauffman authors, and does not reflect any actual fund. Actual fees vary and often step down after the investment period. The authors presented it to show why investors should understand a firm's economics before committing.
Rate is not the only cost. Twenty-three of the Foundation's funds were more than ten years old, and eight were fifteen years or older. Some managers charged additional fees on the remaining portfolio of older funds, for example 1.5% of the cost basis. The authors also objected to the market-standard general partner commitment of about 1% of a fund, arguing it insulates managers from poor results; the Foundation said it expects a 5% to 10% commitment.
Does success repeat? The evidence is more mixed than either side suggests
Investors often judge a new fund by the manager's last one. Whether that works is among the more contested questions in the research, and the answer depends on the type of fund, the period, and what the investor can actually see at the time of the decision.
Within a fund, early numbers can mislead. The Kauffman authors found that their funds' reported net IRRs, in aggregate, tended to peak around month 16, and that the median time to the first capital call of a manager's next fund was about 27 months. In the Foundation's portfolio, then, managers typically went back to investors while the earliest, and usually highest, marks were on the books. Only 25 of 86 funds (29%) followed the textbook J-curve. The authors suggested that early valuations, which general partners set themselves, may help explain the pattern.
Did Early Results Hold? 88 Kauffman Venture Funds vs. the Russell 2000
Paper's "VC PME" label at 27 months and at the end; fund data through December 31, 2011
"VC PME" is the paper's label for a return of 3 to 5 percentage points a year above the Russell 2000 Index, measured by the public market equivalent (PME) method.
Source: Kauffman Foundation (May 2012), p. 47. Benchmark: Russell 2000 Index, which tracks the performance of smaller-capitalization U.S. public companies.
Investors cannot invest directly in any index. Indexes are unmanaged and do not reflect fees or expenses. Results shown are the actual, net-of-fees results of the Kauffman Foundation's venture capital fund portfolio, not IFA accounts, and are not hypothetical. Past performance is not a guarantee of future results. Individual investor accounts (performance, allocation, etc.) may vary.
View chart data as a table
| Outcome | Number of funds |
|---|---|
| Had VC PME at 27 months and still at end | 6 |
| Had VC PME at 27 months, not at end | 16 |
| No VC PME at 27 months, had it at end | 13 |
| Never had VC PME | 53 |
| Total | 88 |
Early results were an imperfect guide. The paper uses the label "VC PME" for a return of 3 to 5 percentage points a year above the Russell 2000 Index, measured by PME. Only 6 of the 22 funds that had VC PME at 27 months still did at the end, while 13 funds that did not have it at 27 months eventually did. Early weakness was more informative: 53 of the 66 funds that lacked VC PME at 27 months (80%) never had it.
"...the J-curve effect is mostly notable by its absence."
— Mulcahy, Weeks and Bradley, Kauffman Foundation, May 2012
Across funds, the academic record is not one-sided. Kaplan and Schoar (2005) documented strong persistence in venture and buyout fund performance. Later research by Harris, Jenkinson, Kaplan, and Stucke, using cash-flow records from institutional investors rather than the database behind the earlier study, which they say has flaws, found that venture persistence continues but weakened after 2000, while buyout persistence faded considerably. When they limited the analysis to what an investor could see when a new fund was being raised, the prior fund's interim results, they found little evidence of persistence for buyouts and continuing but reduced persistence for venture. They point to interim performance being only moderately correlated with final performance, and to managers avoiding fundraising when interim results are poor.
These findings can be read alongside the Kauffman paper. The Kauffman authors argued that great venture returns are concentrated in a small number of funds that new investors struggle to reach, and the paper cites estimates of the number of top-performing venture funds ranging from about 10 to about 50, out of several hundred. In our opinion, the practical problem for many investors is picking funds from noisy early signals, at a time when each additional fund a manager raises adds a new stream of fees.
Other academic research at a glance
| Study | What it examined | Finding relevant to fees or repeatability |
|---|---|---|
| Kaplan and Schoar (2005), Journal of Finance | Returns, persistence, and capital flows in private equity funds, using the Venture Economics database | Found strong persistence across successive funds, for venture and buyout. Later research found problems with the database used. |
| Metrick and Yasuda (2010), Review of Financial Studies | Expected manager revenue for 238 funds (94 venture, 144 buyout) raised 1993–2006, from one large investor's records | About two-thirds of estimated revenue came from fixed components not sensitive to performance. Venture managers earned more revenue per dollar managed than buyout managers. |
| Harris, Jenkinson, and Kaplan (2014), Journal of Finance | Nearly 1,400 U.S. buyout and venture funds, using investor cash-flow data from Burgiss | Buyout funds beat public markets consistently. Venture funds beat public equities in the 1990s and trailed them in the 2000s. |
| Harris, Jenkinson, Kaplan, and Stucke (2020), NBER Working Paper 28109 | Whether a manager's performance carries from one fund to the next, using Burgiss data | Venture persistence continues but declines after 2000. Buyout persistence fades, and using only information available at fundraising, little persistence for buyouts. |
Sources as listed at the end of this article. Descriptions are summaries of the authors' stated findings. Past performance is not a guarantee of future results.
Where the picture is less one-sided
A fair reading of the Kauffman paper requires some context.
- It is one investor's experience, and it is dated. The data run through 2011. The paper does not cover venture funds raised or measured in the years since.
- The Foundation did not abandon venture capital. The authors wrote that they planned to remain venture investors, with a smaller portfolio of perhaps five to ten partnerships chosen for track record and better-aligned terms.
- Top funds can deliver. Sixteen of the Foundation's 99 funds reached 2x or better, and most of its funds raised before 1995 beat the public market by more than 3 percentage points a year. The authors' point was that few investors can reliably identify and access such funds.
- The academic evidence on persistence is not uniform. As described above, venture persistence is better documented than buyout persistence, even after weakening post-2000.
- Public alternatives carry risk too. The authors chose the Russell 2000 because they believe it better reflects the higher price volatility and sensitivity to economic cycles of small companies. Those same characteristics mean it can fall sharply. Index funds do not remove market risk, and lower costs do not guarantee better results.
Questions the Kauffman authors raised for investment committees
The Kauffman authors directed most of their recommendations at investment committees rather than managers. Paraphrasing their themes, the questions they pose include:
- How does the fund's net performance compare with a public benchmark chosen in advance, using PME rather than rankings against other private funds?
- What are the manager's total fees across all of its funds, how are partners paid, and how much of their own money is committed to the fund?
- Is the allocation based on the committee's analysis of the fund, or on a target percentage that needs to be filled? (The authors call the latter "bucket filling.")
- How much of the manager's reported performance is realized, and how much rests on valuations the manager sets?
- What happens to costs and liquidity if the fund runs well beyond its stated ten-year life?
"...but LPs never walk away."
— A venture capital general partner quoted in the Kauffman paper, on why investors have little leverage
The authors' view was that investment committees hold more leverage than they use, because they can decline to invest. Whether the answer is a smaller allocation, better terms, or none at all is a decision each board makes under its own investment policy. These questions summarize the authors' themes; they are not IFA recommendations or advice for any investor.
--------------------------
Sources
- Mulcahy, Diane, Bill Weeks, and Harold S. Bradley. "We Have Met the Enemy... and He Is Us": Lessons from Twenty Years of the Kauffman Foundation's Investments in Venture Capital Funds and The Triumph of Hope over Experience. Ewing Marion Kauffman Foundation, May 2012 (SSRN abstract 2053258).
- Kaplan, Steven N., and Antoinette Schoar. "Private Equity Performance: Returns, Persistence, and Capital Flows." Journal of Finance, 2005.
- Metrick, Andrew, and Ayako Yasuda. "The Economics of Private Equity Funds." Review of Financial Studies, 2010.
- Harris, Robert S., Tim Jenkinson, and Steven N. Kaplan. "Private Equity Performance: What Do We Know?" Journal of Finance, 2014 (NBER Working Paper 17874).
- Harris, Robert S., Tim Jenkinson, Steven N. Kaplan, and Ruediger Stucke. "Has Persistence Persisted in Private Equity? Evidence from Buyout and Venture Capital Funds." NBER Working Paper 28109, November 2020 (SSRN 3736098).
- Silicon Valley Bank. "Dialing Down: Venture Capital Returns to Smaller Fund Sizes." 2009, as cited in the Kauffman paper.
Important disclosures
All investing involves risk, including the possible loss of principal. Past performance is not a guarantee of future results. Index funds do not eliminate market risk; when markets decline, an index portfolio declines with them. Lower costs do not guarantee better results.
Individual investor accounts (performance, allocation, etc.) may vary. Investors cannot invest directly in any index. The Russell 2000 Index, referenced in this article, tracks the performance of smaller-capitalization U.S. public companies; it is unmanaged and does not reflect fees or expenses.
All performance figures in this article are the actual, net-of-fees results of the Ewing Marion Kauffman Foundation's venture capital fund portfolio, or findings of the academic studies cited, as published by their authors. None is the performance of any IFA account, strategy, or model, and none is hypothetical. Unless otherwise stated, fund data run through December 31, 2011. The fee illustration in "What the fees pay for" is a hypothetical example constructed by the Kauffman authors and does not represent any actual fund or any investment performance. Sources for all statistics are stated with each chart and in the Sources list.
Index Fund Advisors, Inc. (IFA) is an SEC-registered investment adviser; however, such registration does not imply a certain level of skill or training and no inference to the contrary should be made. This article is for educational purposes only and is not a recommendation to buy or sell any investment. It summarizes third-party research; IFA is not affiliated with the Ewing Marion Kauffman Foundation.
