Venture capital limited partnership agreements: Understanding compensation arrangements

University of Chicago Law Review

Volumn 76, Issue 1, 2009, Pages 161-218

  Litvak, Kate ^a  

^a UNIVERSITY OF TEXAS   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 66249085920     PISSN: 00419494     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Conference Paper

References (49)

1. For surveys on this topic, see generally John E. Core, Wayne R. Guay, and David F. Larcker, Executive Equity Compensation and Incentives: A Survey, 9 Econ Policy Rev 27 (2003) (synthesizing prior research on equity-based compensation, including a discussion of how compensation is used to align incentives, how equity incentives are measured, when such compensation is deployed, and why researchers have argued it ought to be effective);
2. John M. Abowd and David S. Kaplan, Executive Compensation: Six Questions That Need Answering, 13 J Econ Perspectives 145 (1999) (explaining recent advances in economics literature on executive compensation);
3. Kevin J. Murphy, Executive Compensation, in Orley Ashenfelter and David Card, eds, 3B Handbook of Labor Economics 2485 (Elsevier 1999) (describing executive incentive contracts and surveying empirical and theoretical research on executive compensation).
4. For a summary of the recent literature taking this view, see Lucian Arye Bebchuk and Jesse M. Fried, Executive Compensation As an Agency Problem, 17 J Econ Perspectives 71, 72 (2003) (arguing that the current use of executive compensation to align the incentives of managers with those of shareholders may not be effective since the process by which executive compensation is set is burdened by the very agency problems equity-based compensation attempts to alleviate).
5. See also generally Lucian A. Bebchuk and Jesse M. Fried, Pay without Performance: Overview of the Issues, 30 J Corp L 647 (2005) (arguing that flawed compensation agreements are widespread and problematic and suggesting reforms for greater transparency).
6. Paul Gompers and Josh Lerner, An Analysis of Compensation in the U.S. Venture Capital Partnership, 51 J Fin Econ 3 (1999).
7. See id at 27-28.
8. Id at 14.
9. Id at 21-22.
10. See note 17.
11. See, for example, Steven N. Kaplan and Antoinette Schoar, Private Equity Performance: Returns, Persistence, and Capital Flows, 60 J Fin 1791, 1794 (2005) (pointing to Gompers and Lerner's finding that VC compensation is largely uniform, and finding it "puzzling that [persistently high] returns to superior skill [of some VCs] are not appropriated by the [general partners] through higher fees and larger funds").
12. Andrew Metrick and Ayako Yasuda, The Economics of Private Equity Funds *4 (Working Paper, Swedish Institute for Financial Research Conference on The Economics of the Private Equity Market, Sept 2008), online at http://ssrn.com/abstract=996334 (visited Jan 11, 2009).
13. See id at 15-16.
14. Id at 10.
15. Id at 12.
16. See generally Michael J. Halloran, Lee F. Benton, and Jesse Robert Lovejoy, 1 Venture Capital and Public Offering Negotiation §15 (Aspen Law and Business 3d ed 1996 & Supp 2008) (describing the typical structure of management fees and expenses of venture capital partnerships).
17. See generally Victor Fleischer, The Missing Preferred Return, 31 J Corp L 77 (2005) (pointing out that venture capital partnership agreements do not use hurdle rates, while buyout funds do, and discussing possible reasons for the difference).
18. See Steven Kaplan, Case Study, Accel Partners VII 2 (Chicago 1999), online at http://faculty.chicagogsb.edu/steven.kaplan/teaching/accel7.pdf (visited Jan 11, 2009);
19. Josh Lerner, Case Study, A Note on Private Equity Partnership Agreements (Harvard 2000) (explaining the structure of private equity partnerships in the form of a business school case study).
20. VentureXpert is a large commercial database containing comprehensive information on venture capital firms and funds, executives, and companies backed by private equity. It is widely used in academic research in finance, law, and accounting. See, for example, Gompers and Lerner, 51 J Fin Econ at 14 (cited in note 3).
21. One top venture capital attorney put it this way in email correspondence: "Unlike hedge funds you will never see the management fee based on [the asset value] of the [venture] fund. One of the many reasons for this is that the assets held by these funds are illiquid and difficult to value." Email from anonymous attorney to Kate Litvak (Jan 15, 2004). Gompers and Lerner treat the managed-capital base as equal to the fair market value of the fund's investments, rather than their cost basis. See Gompers and Lerner, An Analysis of Compensation in the U.S. Venture Capital Partnership, 51 J Fin Econ 3, 42 (1999) (cited in note 3). This is likely to be a miscoding, although I cannot be sure because I do not have their agreements.
22. See Part VI.B.
23. For more details, see Kate Litvak, Firm Governance As a Determinant of Capital Lock-in *8-9 (University of Texas Law and Economics Research Paper No 95, Mar 2007), online at http:// ssrn.com/abstract=915004 (visited Jan 11, 2009).
24. One can imagine an explanation in which the manipulation incentives provided by a managed-capital base offset the VCs other manipulation incentives, including those provided by the desire to show a high internal rate of return for the current fund when raising the next fund and those provided by distribution rules. But any offset would be rough at best, and none of my interviewees suggested this explanation.
25. See, for example, David Toll, Private Equity Partnership Terms and Conditions 38 (Dow Jones 3d ed 2003) (noting that, in the private equity context, "[t]he rationale [for switching from committed capital to managed capital as a basis] is that [the VC] will incur greater expenses during the investment period, when the team is putting the money to work. Subsequently ... [the VCs] expenses related to this fund can be tied to the specific companies remaining in the portfolio, and should therefore be reimbursed accordingly").
26. The model year has to be earlier than 1997 to ensure that all of the fund's activity is included.
27. Fleischer, 31 J Corp L 77, 78 (cited in note 14).
28. See Metrick and Yasuda, The Economics of Private Equity Funds at *10 (cited in note 9).
29. Gompers and Lerner, 51J Fin Econ at 14 (cited in note 3).
30. For example, if their investor chooses not to invest in funds with a carry percentage greater than 20 percent, that could explain why only one of the ninety-four VC funds in their sample has a carry percentage above 20 percent. See Metrick and Yasuda, The Economics of Private Equity Funds at *10 (cited in note 9).
31. See generally Franco Modigliani and Merton H. Miller, The Cost of Capital, Corporation Finance, and the Theory of Investment, 48 Am Econ Rev 261 (1958) (presenting the theory that in an efficient market, and in the absence of information asymmetries and costs associated with taxes and bankruptcy, the firm's value is not related to the means by which it is financed).
32. A rare exception is direct traditional borrowing by the fund from an investor or a VC. Interest on the borrowed funds, however, is a different issue from interest paid by VCs on, say, early overpayment of profits.
33. See Metrick and Yasuda, The Economics of Private Equity Funds at *4 (cited in note 9).
34. See discussion in Part II.C.2.
35. See generally, for example, Jonathan Axelrad and Eric Wright, Distribution Provisions in Venture Capital Fund Agreements, Venture Capital Rev (Nov 1997), reprinted in Memorandum from Fund Services Group, Wilson Sonsini Goodrich & Rosati, to Private Equity Fund Clients (Aug 5, 2001), online at http://www.wsgr.com/PDFSearch/1363214.pdf (visited Jan 11, 2009).
36. For a few funds in my sample, VCs were required to repay excess carry net of the tax that they already paid on it. One agreement went further and stated that VCs do not have to return the amount of taxes paid but have to return the amount of future tax benefits from taking the loss. I ignore this complication in the discussion below and in the regressions. If included in the analysis, these net-of-tax provisions would further reinforce my main claim-that distribution rules significantly affect and usually increase the NPV of the carry that VCs actually get-because in most cases, VCs get to keep both the tax they paid and the future tax benefits of any loss.
37. This assumes that the risk of subsequent securities litigation against the fund or the VC (who will be indemnified by the fund) or other unusual events is low enough to be ignored.
38. Partnership agreements typically allow VCs to opt out of distributions to themselves if they so wish. If VCs think that distributed securities are overpriced, they can simply refuse to take securities at that distribution. If VCs refuse a distribution, they receive a credit in their capital account that will eventually be paid out of the proceeds from the sale of other companies (albeit without interest).
39. For a full description, contact the author for a technical appendix.
40. Because full information on still-active funds is not yet available, I use average investment and distribution schedules from a pre-bubble vintage year (1992), provided by Sand Hill Econometrics, to predict expected distributions for funds raised after 1996. Thus, for a fund raised in 1998,1 have data through 2006 and need to estimate year 10 and year 11.1 use an average 1992 fund to establish the trend (percent change in distributions between year 9, year 10, and year 11). I then use that trend to extrapolate year 10 and year 11 for my 1998 funds on the basis of real data for year 9. In regressions that use distribution rules as variables, I limit the sample to funds raised in 1997 or before, to limit the potential impact of relying on extrapolated data.
41. For example, for the Return First rule, the algorithm is as follows: For each month of the fund's existence, examine the fund's distributions to investors and the amount that investors contributed to the fund to date; if the former is lower than the latter, VCs get zero in that month; if the former is higher than the latter, VCs gets the carry percent of the difference between the former and the latter. Sum the undiscounted payments to VCs in each month of the fund's existence through the end of year eleven. Calculate "carry under the agreement"-the difference between the total undiscounted fund return and the total undiscounted investor contributions, multiplied by the applicable carry percentage. If the total undiscounted payment to the VC at the end of year 11 is higher/lower than the carry under the agreement, the overpaid party transfers the amount of overpayment to the other party without interest at the end of year eleven. This amount is called "clawback." Using a 10 percent discount rate, calculate the net present value of all payments to investors and VCs, including payments in midstream and the clawback. The sum of all discounted payments to (by) VCs is the measure of the NPV of carry under the Return First distribution rule.
42. See Gompers and Lerner, 51 J Fin Econ at 27 (cited in note 3).
43. For anecdotal evidence that some high-performing VCs raise their carry percentage, see Kaplan, Accel Partners VII at 1 (cited in note 15). Accel already charged a 2.5 percent management fee and 25 percent carry, and wanted to raise its carry to 30 percent. Kaplan writes: "At a 30 percent carry, Accel would join a select group of private equity firms that included Bain Capital; Kleiner Perkins Caulfield & Byers; and, under some circumstances, Benchmark Capital." Id.
44. None of the VC firms in my sample changed law firms across funds. Thus, the intra-law firm differences cannot be explained by lawyers moving from one firm to another and taking both their clients and agreements with them.
45. In regressions that use my agreements-based sample of sixty-eight, the sample size does not allow me to use venture-firm fixed effects and clusters plus vintage-year fixed effects. Therefore, I have to use weaker specifications.
46. See Murphy, Executive Compensation at 2524 table 7,2526 figure 6B (cited in note 1).
47. For funds using managed capital as a base, I calculate the cumulative fee over eleven years based on Sand Hill Econometrics data on capital calls and distributions, and express that value as the percent of committed capital.
48. No fund in my sample used a managed-capital base in early years and switched to committed capital in later years.
49. I use venture-firm fixed effects rather than random effects, as in some earlier tables, because of Stata limitations on the use of random effects with ordered probit models.