One of the arguments often advanced in favour of hedge funds and private equity funds is that the managers of these funds have interests that are aligned with investors’ (financial interests, that is, as opposed to an interest in expensive yachts, fine art or fox hunting). Incentive fees are an important way to achieve such alignment of interests. But as regular readers are aware, the inherent asymmetry of incentive fees makes them a band-aid solution at best.
Instead, the gold standard of management alignment is usually a manager’s co-investment in the fund. Still, a co-investment of $300,000 from a start-up manager could represent a large portion of their liquid net worth, while a co-investment of $3 million might be a small portion of a more seasoned veteran’s total liquid net worth. For this reason, many investors will ask a manager to reveal the portion of their net worth in their fund.
While this line of inquiry is common during institutional due diligence on hedge fund and private equity funds, co-investment data is not generally divulged to publicly-available databases. There is a void in the academic research when it comes to manager investment data…until now.
David Robinson of Duke University and Berk Sensoy of Ohio State managed to put their schools’ basketball rivalry aside last year in order to cooperate on a groundbreaking study of private equity funds. The study was based on proprietary data from a “large (anonymous) institutional limited partner with extensive investments in venture capital, buyout, real estate, distressed debt and fund-of-fund private equity funds.” It contained 990 funds managing $677 billion from 1984 to Q2, 2010. So although there was clearly an inherent selection bias in the data, it was “ideally well suited to understanding private equity cash flows, performance and contractual arrangement” according to the authors.
The data set included cash flows in and out of each fund as well as the GP’s own investments in each. Quarterly cash flows allowed them to better compare performance of the PE funds to that of public equity markets. And GP investments allowed them to see who had the most “skin in the game.” The chart below was created with data from Table 3 in the duo’s paper and shows the mean management fees and GP commitments across various private equity strategies.
If you assume a PE manager’s own valuation is about 2 times recurring (management fee) revenue and that the funds in question are the sole fund for each GP, then you might say that GPs bet about half the value of their own management firm on most strategies. Most, but certainly not all. Real estate managers place bets that are most likely much more than the worth of their own firms. Notably, with the exception of funds of funds, the highest management fee is accompanied by the lowest GP commitments. While this may look like managers would rather charge fees to clients and not themselves (thus adhering to time-honoured “OPM” principles), many would point out that VC is more idiosyncratic and volatile than the other strategies and therefore requires more diversification on behalf of investors – including the GP itself.
(In case you were wondering, the carried interest percentage was about 20% across all strategies.)
They also found that larger funds (holding strategy constant) had lower management fees and higher carried interest. This will make intuitive sense to those who contend that larger managers are essentially more confident in themselves. Or, alternatively, that larger managers can “keep the lights on” with a lower management fee percentage and that they figured they might as well propose a higher incentive fee as a quid pro quo from investors.
You’d think that boom times, i.e. when managers are most confident, would also lead to higher carried interest. Yet this wasn’t the case. In fact, the researchers found that “increased GP bargaining power” lead to a shift toward management fees during booms times for each respective strategy.
Similarly counterintuitive was the duo’s finding that large funds are more likely to have a large GP allocation (on a percentage basis). While it seems to contrast with the “bargaining power” argument above, they suggest that this may be since,
“…LPs are concerned about GP incentives in large funds and accordingly are more likely to require above-normal GP capital commitments to help align incentives.”
Since Robinson and Sensoy had access to quarterly cash flows in and out of funds, they could more closely compare PE performance with that of public equities. To do this, they assumed that all distributions were invested in the S&P500 on the day of their issuance. Distributions and (their converse) capital calls often depend on the general state of the economy and markets. Clearly there will be more distributions in boom times due to easier exits (IPOs, private sales, etc.). Capital calls also rose during bull markets, but by a lesser amount than distributions.
There is a common assumption that private equity is a “liquidity sink” – that when market liquidity dries up, PE funds ramp up the capital calls. But curiously, Robinson and Sensoy actually found that both distributions and capital calls had a similar sensitivity to the TED spread most of the time. During the financial crisis, for example, distributions and capital calls dropped as PE managers crawled under the nearest rock. But the “component of calls not explained by [concurrent drops in] public equity valuations and the TED spread spiked,” suggesting some sort of yet-unidentified secular influence that was unique to the crisis.
In any event, once they accounted for the reinvestment of these distributions into the S&P 500, they found that PE funds produce 1.5% – 2.5% excess performance per annum. The “public market equivalent (PME)” for each strategy illustrated the multiple of the S&P 500 delivered by the strategy over the full time series. A PME of 1.1 means the strategy beat the S&P 500 by 10%. While there were significant ranges in the PMEs for funds within each strategy, most delivered more than just the S&P 500.
As always, it appears that VC may be driven more by the fact that hope springs eternal, rather than what the track record by academics attributes to it. Perhaps a good reason for managers to have just a little less skin in the game when it comes to VC.