It’s hard to ignore the fight-or-flight impulse, but private equity investors should do everything they can to avoid getting scared out of the market by short-term losses. If history is a guide, those who flee will have a hard time getting back in, while those who hang on will be rewarded for their patience.
This is the first in a series of articles focusing on the performance and restructuring of the industry. Our aim is to help investors put short-term losses into perspective and hopefully avoid the same mistakes they made during a similar downturn 20 years ago.
Let’s start with the numbers. There is no disputing that short-term industry performance has been negative over the last two years (see Figure 1). In fact, the last nine quarters have produced negative returns for each of the preceding 12 months, according to the Thomson Venture Economics (VE) Performance Database. The years 2001 and 2002 posted returns of 27.7% and 23.3%, year over year, respectively. These are the steepest one-year declines in the VE Performance Database, which tracks quarterly and annual performance statistics for all periods since 1969.
What is lost and misleading in all the negative press about these short-term numbers is that the industry has been on a 30-year tear of positive returns for each year since 1970-with only 1974 and 2001/2002 as exceptions.
At the height of the technology boom in 1999 the industry posted a 166% return year on year. This is a long-term investment class, and short-term performance often masks this (see Figure 1 and Figure 2). The long-term performance of the industry is quite robust at 16.6% net returns to limited partners compounded annually over the last 20 years. This compares favorably to the S&P 500 return of 11.6% in the same period. In fact, on a vintage year basis, the venture industry has beaten the S&P 500 in 16 of the last 23 years (see Figure 3).
If you examine venture capital performance over the last 20-plus years using a rolling investment horizon of five years this cycle is readily apparent (see Figure 4). The performance of the industry in the early 1980s mirrored the technology boom of the era followed by the decline in the late 1980s to early 1990s. The latest positive performance cycle began in 1991 and continued until 2000. It is these technology booms where venture capital (especially early-stage VC) both creates and reaps the benefit of investment.
But unlike the public markets, where you can participate in a boom profitably almost at any time, investors in venture capital funds are constrained by the fact that they can’t just jump into and out of the market. For example, the best performing vintage year ever was 1996. It benefited from the boom of 1998-2000.
Technology Cycles are Normal
It was the unprecedented performance of 1997-1999 that led to an environment of over-investment and concentrated investment in relatively few years. Those who invested in funds prior to this period have done demonstrably better than those who invested during the boom. This should come as no surprise to any investor investing in either public or private markets. Virtually everyone inside and outside of the industry claimed to know it was going to be over soon, but nevertheless, investment, fund-raising, performance and the IPO market took off on a rocket ship ride that soon fell back to earth.
Venture-backed IPO data clearly shows how the highs and lows of the venture industry are tied to technology boom and bust cycles (see Figure 5). Those who invested early in the venture investing cycle made out better than those who invested late, as it should be.
That the venture industry is so intrinsically tied to technology investment and concomitant corporate investment cycles should not be surprising. I often quote Harry Rein of Canaan Partners, who quipped in the mid-1990s: “No one has repealed the business cycle.” Other than growing companies in a laboratory or a cave insulated from the world at large, companies financed by venture investors are subject to the same economic factors that affect all companies.
A Lesson Learned …
The tremendous returns of the late 1990s increased demand for investment in venture funds to the point where the industry raised more money in two years (1999 and 2000) than it had raised in the prior 20 years (see Figure 6). In the following two years the total amount of money raised by venture funds decreased to levels that mirror 1995-1996. Many venture investors haven’t shown any interest in going back to the LP community for more capital, choosing instead to spend more time with their companies and to try to grow the best and weed out the rest. Not only has the total capital decreased substantially, but the fund sizes have been decreased as well. The average fund raised in 2002 was $130 million compared with an average of more than $210 million in 2000 (see Figure 6).
In addition, the recent trend of funds “downsizing” has continued to decrease the amount of available capital, although the total amount of un-invested capital is still at record levels. The downsizing of the industry is not necessarily a sign that venture investors bet wrong on their fund-raising efforts in 1999-2000.
“Givebacks” to LPs are simply the result of GPs adjusting their funds to fit the new realities of the market. In the frenzied market of 1998-2000, investors would put large amounts of capital into companies because they might only have one chance to invest in a company before the price got too expensive for a follow-on deal. Additionally, companies funded during that period were expected to grow very quickly to reach critical mass for an exit, thus requiring more capital in a short period of time. That being the case, it was not unusual for one venture fund to do all the investing in a company with few if any syndicated partners.
Now that valuations are at reasonable levels, pricing of deals has substantially reduced the need for the massive amounts of capital required during the boom. And since the investment pace has slowed, capital can be apportioned over longer periods of time, creating more opportunities for syndicated deals or follow-on financings by other funds.
One of the benefits overlooked as a result of the givebacks is that it has allowed LPs to rebalance their private equity portfolios. They have been able to optimize their allocation models by shifting their commitments to a later period-a second chance, so to speak, to re-optimize the temporal allocation of their portfolios.
… And a Lesson Forgotten
However, Figure 6 illustrates a lesson forgotten by some limited partners. One basic principle of investment/trading/speculation is that as market prices increase (in our case prices can be taken as valuations and performance), prudent investors typically decrease the size of their incremental investments rather than increase them. This results in the cost basis of their investment being low enough so a small or medium downturn in a market won’t devastate their portfolios. What the LP community did in the boom was exactly the opposite, investing more and more as performance increased. The consequences should be obvious:
• By focusing so much capital in one period, the investor is guilty of trying to “time” the market, which is difficult enough (most would say impossible) in the public markets and simply doesn’t seem to be possible in private equity.
• Capital is not temporally diversified. That is, capital is so focused on one period of time that it’s the equivalent of laying all the money you have on one turn of the roulette wheel.
• It doesn’t take much of a downturn to create a loss, since the cost basis of the investment is very high.
• The upside of the current environment is that the givebacks allow LPs to reallocate commitments to more favorable vintage years.
Without detailing the various complexities of Modern Portfolio Theory (MPT), suffice it to say that limited partners commit anywhere from 2% to 6% of their capital into alternative investments, of which private equity is a subset. MPT attempts to optimize the risk and return in a multi-asset class, with risk typically being measured as price volatility over time. These allocation models are usually set for about five to six years. They are typically measured on a market value basis, so in the heady days of the late 1990s, as public equities increased in value tremendously, many of these allocation models became out of balance. The way to remedy this is to either sell the public equities or invest more in other asset classes. Most LPs chose to do the latter.
Another factor contributing to the level of commitments was that by 1997 the industry was reaping the benefits of almost 20 years of investment history, and distributions were outpacing new commitments. Thus, the industry became self funded and limited partners didn’t necessarily need to find new allocations of capital to balance their portfolio, since they could re-commit distributed proceeds to private equity. When the industry started its decline in late 2000, distributions declined and LPs had to find new sources for additional committed capital. They couldn’t depend on distributed proceeds.
A third factor was that even with the frenzied pace of investing in the late 1990s, LPs found that they needed to “over-commit,” since the actual amount taken down in capital calls was often much lower than their committed capital, as the industry struggled to keep pace with financing demands.
Since public and private equities were both driven down in value, the asset allocation decisions made were probably still in balance. But by late 2001 and 2002, when it was evident that public equity and private equity returns were not returning to the short-term highs of the 1990s, some of the asset allocation models were recalculated and rebalanced to reflect expectations of lower returns.
The Flaw in MPT
There is a glitch in applying MPT, as illustrated above. MPT assumes a few things that can be problematic when applying MPT to private equity in a mechanical exercise. Its most basic assumptions are that there are identical investment horizons, that each investor has the same perception and information relative to risk (measured by standard deviation), and that return and correlation between asset classes are the same. Another problematic assumption is that you can have perfect short selling opportunities.
Mathematically, the above assumptions can create an optimal portfolio. However for private equity there are some problems. One has to do with assumptions of risk and the other has to do with the ability to have rebalanced portfolios. Risk traditionally is measured in MPT using standard deviation of volatility-that is, how much the value of an asset class moves up and down over time. Secondly, MPT assumes you can rebalance a portfolio at any time.
Historical volatility is fairly easy to estimate in public equities. There are prices available on virtually any time period one chooses. However, when looking at private equity, volatility is a difficult thing to measure, since valuations are not priced frequently enough to accurately measure volatility. Given standard practices for valuation in the industry, volatility tends to be attenuated, since valuations typically don’t “occur” except during an actual transaction. For more detail on these caveats, see the extensive coverage of cross asset class comparisons with private equity in VE’s “Investment Benchmark Reports: Venture Capital.” The upshot is that practitioners have used sophisticated estimation techniques to try to get at private equity “true volatility,” but none has been satisfactory.
The bigger problem lies in the “perfect short selling” assumption, which assumes you can buy/sell the asset when you choose to invest in a two-asset portfolio-assume for now public stocks and venture capital funds. In the short run, the only way to bring a portfolio to optimal balance is probably controlled only by the amount of stock you can buy and sell. There are liquidity constraints and investment constraints in the venture capital fund that make it nearly impossible to balance a portfolio by selling or investing in venture capital funds. Since venture funds are closed-end vehicles, in the short run there are relatively few opportunities to invest in the “best funds” at any point in time. There is usually a long line of investors wanting to get into the best funds, so positions in these funds are typically rationed to previous investors first.
Another factor that needs to be considered is that once you have exited the venture industry, it is very difficult to ramp up your allocation in the asset class without taking on more risk. That’s because you typically have to invest in newer funds without the track record of older funds. This is the temporal allocation timing problem: Timing your entry into this market is next to impossible. As many institutions found out by trying to increase their allocations in the heady markets of 1999-2000, they were forced to invest in newer funds, which by itself is not a bad thing, but over-allocation into new funds takes on more risk of sub-par performance. There is anecdotal evidence that during the last downturn of the venture market in the 1980s, many institutions that exited the venture market after over-investing in the 1980-1983 technology boom found themselves in an awkward position when they tried to get back into the market in the early 1990s. They spent the better part of the next decade trying to rebalance their private equity portfolio to catch up.
The ultimate lesson here is that an investor has to already be in a venture fund prior to a technology boom that presages superior investment returns. You can’t time the private equity market. Investors that are out of the market or under-allocated in this market when these long investment cycles turn will always be chasing performance.
Jesse Reyes is vice president of research at Thomson Venture Economics (publisher of VCJ).
PART 2 of 3: All Funds Aren’t Created Equal
Click here for All Funds Arent Created Equal, which appeared in the July 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.