In the first article of this three-part series analyzing private equity performance, we examined the performance cycles in the venture capital and private equity industries. We showed that this is a long-term investment asset class, and, while short-term performance is important, it should not be seen as an indicator of the long-term performance of the industry. We also showed the dangers of trying to “time” private equity investing.
In this installment, we look at the controversial issue of top-quartile funds and explain common misconceptions. We also demonstrate why manager selection is so important.
Private equity performance goes through cycles, and it is the prudent investors who refuse to try to time the market. Rather, they develop and try to adhere to a consistent investment strategy. However, not all asset classes are created equal, and it is a nave investor who treats venture capital or buyout investing like a stock index and simply throws money into the investment without examining which types of funds to invest in, and, more importantly, which managers to invest in.
It is important to understand the long-term characteristics of private equity and examine the first choice that an investor is often faced with when looking at private equity investing. Do you invest in venture capital or in buyout funds? Figure 1 illustrates an important point about investing in private equity: You should invest in both.
Figure 1 tracks performance for venture capital and buyout funds separately over five-year rolling horizon periods. This is done by calculating an internal rate of return (IRR) over a five-year period for all funds regardless of vintage and for all year ending periods. By way of example, we determined the return for the five-year period ending Dec. 31, 1992, by calculating an IRR from Dec. 31, 1987, to Dec. 31, 1992, and then from Dec. 31, 1988, to Dec. 31, 1993, and so forth. This gives us a trend line of intermediate-term performance. The result in Figure 1 shows one of the basic choices that limited partner investors make.
Assuming their asset allocation model tells them to put some capital in private equity (whether it’s venture, buyout or mezzanine funds), this chart illustrates clearly why LPs invest in both venture and buyout portfolios. The inverse correlation between the two asset classes, which is clearly evident in this figure, provides diversification and a hedge against long-term losses even though it also attenuates some of the clear winning years.
It’s easy to see that when venture capital returns were in the doldrums in the mid 1980s, it was the heyday of buyout groups, and the inverse trend can clearly be seen in the technology boom of the 1990s. It’s this diversification that gives investors expectations of long-term returns. Draw a line through the two to average the returns of both asset classes and you get total private equity returns, which is illustrated in Figure 2. This can be interpreted as the returns expected if a hypothetical LP investor invested in every venture and buyout fund formed from 1969 to the present. The inverse correlation gives you an averaged return between the two asset classes.
But this presents a bit of a problem: If an investor is not discerning and simply treats private equity as a market basket index, the resultant private equity returns are not fundamentally different than a frequent alternative-the S&P500 index which we’ve superimposed in Figure 3. Private equity returns in total were superior to the S&P500 in the early 1980s, but they have been more convergent in most recent times. Figure 3 illustrates that LP investors simply can’t throw money at any private equity investment and expect superior returns. They have to choose either an allocation or manager selection criteria that will give them superior returns.
Who’s Top Quartile?
That brings us to the “top quartile” question, one of the most quoted yet misunderstood statistics published. LP investors often quip that they’ve never seen anything but first-quartile funds because it seems that all funds make that claim. In fairness to general partners, however, I’ve never opened up a financial magazine and seen a mutual fund that advertised itself as No. 2 in its class. Everyone is No. 1. Kidding aside, it is obvious that not all funds can be top-quartile funds, and no investor can invest only in top-quartile funds as someone has to back those funds with lower returns.
A few definitions are in order. A “quartile” is the return that demarcates quarters in a sorted array of statistics. So three quartiles will separate a sorted list of data into four equal quarters. In mathematics, you start quartiles calculations from the bottom of a sorted list, so, strictly speaking, the first quartile marks the bottom of a series and the fourth quartile marks the top. However, in finance that designation won’t give us the classifications that are intuitive, so we typically turn them on their head and designate the first quartile as the best returns and the fourth quartile as the worst. The first is the point one quarter from the top of a sorted series, which separates the series into the top and second-from-the-top quarters. The median (or second quartile) separates the second and third quarters and the third quartile separates the third and fourth. Sometimes we use the term upper quartile and lower quartile to avoid criticism from our stricter statistician and mathematician friends.
Confusion about quartiles lies in what the quartile designates. In our above example, the first quartile is not the return of the funds in the top quarter. It is the return that is at the bottom of the top quarter-the line you have to cross in order to be a top-quartile fund. Confusion occurs when you ask for the performance of top-quartile funds. Are you asking for the quartile itself, which defines the beginning of the top quartile? Or, do you want to know the performance of funds that are above that line, that is, the funds in the top quarter? The difference is significant.
Let’s look at an example. The top quartile (as published in the Thomson Venture Economics Investment Benchmarks Report) for venture funds formed in 1998 is 62.2%, as of as of Dec. 31, 2002. That means that a fund must post a return of at least 62.2% to be above the top quartile of that sample.
Now let’s take all the funds above the top quartile to see how they performed as a group. The group of funds with returns above 62.2% (or top quartile funds) had a pooled IRR of 197.3% as of Dec. 31, 2002. This can be compared to a 34% average return for the entire 1998 vintage year sample. That means that the overall return of the “best” funds in a vintage year is significantly higher than the average for that sample.
Quartiles Are Relative
The results are always relative to the vintage year. Venture funds formed in 1988 would need a return of just 18.5% to be considered a top-quartile fund. All the top-quartile funds taken as a group would have returned 30.8% to their investors.
We can take each vintage year as a class and choose all the top-quarter funds in each vintage year to create a sample of top-quartile funds. Figure 4 summarizes the results returned by this rarified group of funds and indicates why investing in them is so coveted. This chart plots the cumulative return of this group of funds, and it charts their cash-on-cash returns by a cumulative-to-distribution ratio.
If we take the funds above the top quartile for funds formed in 1996 (which we designated in our last article as probably the best performing group in history), the entire group had an IRR of 188% and had distributed 12.97 times paid-in capital as of Dec. 31, 2002. That’s a feat that will probably never be repeated but also clearly indicates the reason why so many investors were attracted to this asset class in the last half of the 1990s. To be considered a top-quartile fund in vintage year 1996 you would need to post a return equal to or higher than 95.4%.
Getting back to our investment horizon analysis, let’s calculate five-year horizon returns for our top-quartile fund group (see Figure 5). The results clearly outperform the S&P500 of the prior horizon chart and clearly indicate why the mythical (some would say “mystical”) top quartile is so important to investors. As stated before, there is no way for an LP investor to get these returns unless they invested only in top-quartile funds, which they obviously can’t know before the fact. Since you can’t know which quartile a fund will ultimately wind up in, you have to use as much due diligence as possible to increase the odds that you will have at least some funds in the top quartile.
Experience is one of the most important characteristics cited by investors looking at funds. Does a firm have a record of performance? And how do you choose a new group with its first fund, since some of these new funds have the chance of being stellar performers?
There is a rule of thumb-often quoted, but never attributed-that fund No. 3 is an important one. The thinking is that anyone can raise two funds, because you can raise the second fund before the first fund shows results. The third fund, on the other hand, separates the winners from the losers, because the only way to raise a third fund is to show satisfactory results from your first two funds.
Performance statistics validate this observation. To examine this question, we stratified all the firms in our database into firms with only one fund under management, two funds, three funds, all the way up to firms with five or more funds. We then calculated IRRs since inception for these funds/firms as well as the first-quartile breaks and distribution to paid-in ratios.
Private equity is no different than other investing in that there is a Darwinian process at work, so it is natural to expect that firms with funds with more roman numerals after their names would have better performance. Figure 6 seems to validate this: It shows that firms with one fund had an 11.8% return, while funds with five or more funds had a return of 20.3%. Expectations seem to be validated. Firms with more funds have a better track record-it’s the self-selection process at work-except for funds belonging to firms with only two funds under management, which had a 4.6% IRR.
If our anecdotal observations were valid, we would expect this group to be littered with a lot of funds that did not raise the crucial third fund, ostensibly due to poor performance, thus dragging down the performance of all firms with only two funds.
On average funds in this group have returned just 43% of paid-in capital, compared with 83% returned by one-fund families and a return of 1.70 times paid-in capital by firms with five or more funds under management.
Does this mean an investor shouldn’t invest in firms with only one or two funds? No. It just means that you have to do more due diligence and look for desired characteristics when choosing managers.
Which brings us to manager selection. We propose that this is the most important part of the asset allocation process, that you can’t simply invest in private equity without knowing something about the risk/return characteristics of sub asset classes and their managers.
You might suppose that early-stage investments have had the highest returns, because their charter is to invest earliest in the life of a prospective winner and thus have the lowest cost basis on the winners. You might also suppose that funds that have shorter investment horizons would have lower returns because they have invested at higher prices. So, returns to later-stage funds should have lower returns but pose less risk, since they invest in extant companies with business models that are more mature. While early-stage investors may have higher returns, they are investing in companies with a longer investment horizon and more hurdles on the way to profitability.
We have summarized return results between sub asset classes in Figure 7. The chart shows returns since inception, regardless of vintage year, stratified by stage and fund size and for buyouts. As expected, the highest returns to venture capital funds were posted by early-stage funds, which had returns of 500 basis points higher than the next-best venture group. Buyout funds in the small fund group (under $250 million) had the best returns among buyout funds.
However, all managers are not created equal. While the early-stage group had the highest return, it also showed the greatest differences in returns between managers, as measured by standard deviation of returns in the third statistical column. Note the 68% standard deviation to early-stage funds. That means these funds have the highest level of “manager risk.” That is, the risk of choosing the wrong manager. Balanced and later-stage funds have a much lower risk between managers, but they also post much lower returns.
How do you decide among these alternatives? One barometer we propose is the statistical measure called “coefficient of variation,” which divides standard deviation by average return, normalizing comparisons between groups that take both risk and return into account. Some of these normalizing ratios are common in the public markets, although they measure different things. For example, there is the Sharpe ratio, which normalizes risk and return relative to an overall market index.
Our coefficient of variation-the VE Manager Risk Coefficient (MRC)-attempts to normalize manager risk selection with return. That is the last column on Figure 7.
Strictly interpreted, the MRC is the amount of manager risk you are willing to undertake as an investor for a 1% increase in return. For example, early-stage funds have an MRC of 3.2 vs. 1.8 for later-stage funds. This indicates that on a risk/return basis, other things being equal, later-stage funds have a better risk/return profile than early-stage funds.
(Rather than look at the MRC as an absolute measure, the statistic makes more sense when it is contrasted with the MRC of other asset classes.)
The Case for Managers
Does this mean that you should invest in just later-stage funds and not invest in early-stage funds? No. That is the wrong interpretation. It simply means that if you invest in groups with higher returns and higher MRCs, then you need to have strong manager selection skills to get the higher returns. As a limited partner investor, if you are going to invest in funds with high risk and return, you have to have the ability to get into the best fund families and have the time, energy and resources to make discerning choices between managers.
If you can’t get into the best funds, or you aren’t able to do the proper homework because of lack of resources, then you need to either invest in groups with lower MRCs or get a third party, such as an advisor or fund of funds, to invest for you. They have the means and resources to get into better funds and/or do better due diligence.
If you have the means or the ability to discern the best funds in the early-stage category and you are able to get into them, then by all means swing for the fences-at least judiciously. This is not meant to tell you how to allocate your resources, but how to take risk and return into account when investing in private equity.
As another example, look at buyout funds. Funds under $250 million have the highest return among buyout funds, but they also have the greatest difference between managers as measured by standard deviation. However it’s the mega-buyout fund category that has the highest MRC. That is, it has a high-risk measure with a lower overall return. Or, interpreted another way, you should be very careful when investing in mega-buyout funds, since the difference between managers is so much greater than the marginal return you get for that risk. Again, this doesn’t mean you shouldn’t put money into mega-buyout funds, but you do have to know whom you are investing in.
This is probably the mantra that should be adopted by all, that private equity investing-while now a mainstream asset class-is not homogenous in either risk or return. So, you have to strive to increase the odds of getting into the better funds with either better due diligence or handing off the due diligence and investment selection to a third party. This is not an asset class for the casual investor, but it can produce phenomenal returns to the prudent and discerning investor.
Jesse Reyes is vice president of research at Thomson Venture Economics (publisher of VCJ).
PART 1 of 3: Put Performance Into Perspective
Click here for Put Performance Into Perspective, which appeared in the June 2003 issue of Venture Capital Journal.
PART 3 of 3: Dare To Compare
Click here for Dare To Compare, which appeared in the August 2003 issue of Venture Capital Journal.