Dry Powder Dilemma –

Dry powder. Overhang. Un-invested capital. These are some of the terms used to describe one simple but complex concept that is usually posed as a question: “How much money is there to invest in deals?” Whatever you call it, it’s top of mind. There is a growing unease in the marketplace that in spite of the unprecedented investment binge of the late 1990s, there is an exorbitant amount still waiting to be put to work.

We estimate that venture funds had more than $100 billion in dry powder at the end of last year, while buyout funds had more than $120 billion ready to be put to work (see Figure 2). It is clear that the industry has been sitting on the sidelines with the capital it raised over the past two to three years. Even if we remove 20%-25% of the capital because of the downsizing of funds that has occurred this year, that still leaves more than $150 billion available to be invested by the private equity industry.

The Real World

The amount of dry powder is not an academic discussion. Limited partners typically pay management fees on committed capital, not on the actual capital invested. That means they are paying fees on hundreds of billions of un-invested dollars. For that very reason LPs have put pressure on VCs to reduce the size of their funds. After months of internal debate and discussions with LPs, several venture firms, including the esteemed Kleiner Perkins Caufield & Byers, have told LPs that they plan to draw down less capital than expected, ultimately resulting in lower fees.

That isn’t the only way that the dry powder issue affects LPs and GPs. LPs are eager to have their money put to work as they usually have allocation models that are based on actual deployed capital, not committed capital. Moreover, the amount of capital available means there are more competitive dollars available for deals, and that could create an overheated valuation environment if all GPs start looking at any one sector. Finally, GPs who raised their first fund in the waning years of the 1990s will find it difficult to raise additional capital, and they could suffer a capital crunch if they are not among those who have a large reserve built up.

The VC Burn Rate

Looking at the paid-in capital to committed capital (PICC) ratio which is the cumulative five-year take-down figure divided by the five-year investment figure is one way to determine how quickly venture and buyout funds are putting capital to work. It shows that even given the tremendous investment surge in the late 1990s, fund raising outpaced investment. The early 1990s, for example, show that buyout funds were running through capital quickly as they invested 60% of the capital they had raised in the previous five years.

Fun with Numbers

Another way of examining this is to look at the percent of un-invested capital, or the reciprocal of the PICC ratio. Figure 3, which tracks this from 1980 to the third quarter of 2001, shows that the private equity industry overall has invested about 50% of the capital it has raised in the prior five years. From that perspective the current environment is fairly typical of the nature industry. The buyouts industry went through a period where it, too, saw commitments outpace investments in the mid-1980s. But the VC industry typically has had about 60% invested at any one time in the last few years.

These cycles of over-investment and commitment appear to be part of the industry’s nature. What is different now is the sheer magnitude of capital raised and deployed in the past five years.

Beep, Beep!

The venture industry behaved a little like Wile E. Coyote does when he chases the Roadrunner off a cliff. As long as he doesn’t look down, he doesn’t fall. But, like Wile E., the industry has looked down. After seeing how far the bottom is, the industry is rightfully concerned that over-investment will hurt returns, and that will make fund raising much more difficult in the years ahead. Thus, more and more funds are downsizing their commitments.

But it’s a double-edge sword. A few short quarters ago, VCs were being chastised for investing too quickly and too much, especially in the overheated Internet sector. The effect of the pull back in investing over the last year has obviously increased the amount of capital left on the sidelines. But this creates an asset allocation problem for LPs. Once the LP is relieved of some of its committed capital obligation, it means that the allocation that the investor targeted has changed, in some cases quite significantly. For some LPs who have target allocations, this means they have to do some major rebalancing of their portfolios. This is not always an easy task when private equity is involved.

Any discussion of dry powder would be incomplete without a discussion of how dry powder figures are determined.

Simple Math?

On the surface, it seems to be a simple piece of math. Venture Economics (VE), the publisher of Venture Capital Journal, releases commitment numbers and investment (disbursement) numbers quarterly. It seems like you should be able to just subtract one from the other to determine how much is left, but it’s not that simple.

We often field inquiries where a casual observer says, “What is this that you are saying about the dry powder estimate being $50 billion or $75 billion? Last year you published that the industry invested $103 billion and raised $105 billion so there must be only about $2 billion left over to spend.”

There is more to this question than meets the eye. Unless you keep a running bank balance in your checkbook, it is impossible to tell how much money you have in the bank unless you can count all the deposits, checks and withdrawals made, and most importantly, a beginning bank balance to start with. To that end, the overhang estimate is nothing more than a bank balance.

If it were only as simple as balancing a checkbook. Before you can figure out what you’re dealing with, you have to understand the terminology. To avoid confusion VE has used precise terms for private equity transactions. Capital raised by funds from limited partner investors is termed “commitments” or “fundraising.” Capital invested by funds in portfolio companies is termed “disbursements.” Once capital is committed by limited partners it may be called down by funds for actual disbursement. These disbursements from LPs are termed “capital calls,” “takedowns” or “paid-in capital.”

Dig In

Now that we’ve established the terms, lets dig into how dry powder estimates are actually made. One of the first problems you’ll encounter with these estimates is that virtually all research firms include investments by all investment entities, whether they have committed capital or not. Researchers often mistakenly publish disbursement data in portfolio companies regardless of whether there is actual committed capital at play.

Why is this important? Well, disbursements include investments made by corporate venture groups and captive and semi-captive groups of financial institutions. Investments by those groups don’t come from the institutional investors, and that capital isn’t counted in committed capital statistics for the industry.

So, the fount of capital available for investment is actually larger than the committed capital figures published by any research firm. There are more sources for investment than there are sources for commitments, which means that researchers either undercount the amount of capital available or overestimate how much is taken down from the commitment pool.

Yet another problem with estimates from research firms is that they are based on a calendar year, but capital under management is a figure that is most accurately counted based on the vintage year of venture funds.

Let’s look at this through a specific example. Say ABC Venture Fund closed on a total fund size of $500 million on March 13, 2000. The casual observer would say: “Oh, commitments for 2000 total $500 million.” But assume that the real story is that the fund closed on $450 million in 1999 and another $50 million in 2000. If you were not aware of the two separate closings, you could make a mistake and count the $450 million in both years. It takes tedious research to be sure that commitments are phased correctly and aren’t double counted. The best way to do this is to count fundraising once in the vintage year of the fund.

To get an accurate tally of committed capital, you can take two approaches. In the first, you count only those investments made by true funds that have true committed capital. This assumes that you can get a comprehensive accounting of all investments made, and that’s not really possible because the universe of private equity investors is so massive. Many times investments are allocated to a VC firm and not to a specific fund. That makes a true estimate of an individual fund’s capital balance tricky if you depend on reported investments by a firm.

The second approach, and the one that VE believes to be the most accurate, is to go up the food chain and measure actual capital calls against committed capital. One potentially reliable and accurate source is to use the data that VE compiles using its performance database that tracks the flow of cash to and from LPs by venture funds.

VE currently tracks more than $400 billion in institutional capital invested in more than 2000 private equity funds through our limited partner analytics services. Those data come from quarterly and annual audited statements from general partners in addition to surveys of general partners. Since these performance reports are used and vetted by institutional investors and VE serves as an analytics intermediary, this provides the most accurate accounting of how much capital is left over from capital committed.

Four Reasons in One Day

It should be noted that even this technique of tracking dry powder isn’t perfect for four reasons:

    * It doesn’t address evergreen captive investment vehicles such as corporate venturing programs. For the purpose of estimating dry powder, this has to be ignored as these investments are typically provided on an as-needed basis. We are only trying to get to the amount of capital left over from true committed capital.

    * It’s virtually impossible to cover the entire universe of funds, so statistical extrapolation has to be performed to determine an industry aggregate estimate.

    * Some firms are reducing the size of their funds drawing down less capital from LPs than planned and some are just closing down without drawing down any capital. These factors are incorporated into research as they become known, but estimates will always lag.

    * Firms are reserving more capital for follow-on investment to current portfolio companies. This is difficult to discern at any point in time, so any estimate of dry powder doesn’t mean that that amount is actually available for “new” investment in portfolio companies.

Given all the variables, the following methodology is the approach that we have refined over the past few months. First, using the VE performance sample, we compute the fund capital committed on a vintage year basis for each fund. Second, we compute the capital calls (also known as “paid in capital,” including management fees) for each fund in the performance sample. Third, based on the first and second figures, we calculate an outstanding balance figure for each fund. Fourth, we aggregate this to the entire sample to get a performance sample outstanding balance. Fifth, we extrapolate this estimate to the universe of funds that are not in the VE sample using a variety of proprietary statistical techniques. (For the result, see Figure 1.)

Even that isn’t enough. We refine this method because there is statistical “leakage” in the performance database over long periods of time. We use a rolling period estimate that encompasses the reasonable period of time that it takes a fund to draw down all it’s capital. We have experimented with several scenarios, but we have found that a five- or six-year period is more than ample time to encompass 99% of all capital calls and smoothes out statistical perturbations.

To refine our data, we take the following steps:

    * Estimate total capital (fund size) for the performance sample on a rolling five vintage year basis. For example for the year ended 2001, we computed total fund size for funds formed from 1997 to 2001.

    * Compute total takedowns for these funds for the same period of time.

    * Estimate un-invested capital for five-year period.

    * Extrapolate the overall figure to the universe of all funds with committed capital for the same five-year period. (Note that we haven’t completed 2001 data collection, so year-end statistics may vary.) The VE VentureXpert database has a module in the performance analytics section that calculates this estimate for the sample funds in the performance database. This sample number is the number that VE has traditionally published as its dry powder estimate of $30 billion to $50 billion. Starting late last year however, it became obvious that VE’s performance database was not as representative of the tremendous amount of capital raised in the last two to three years. Thus our sample estimate had to be extrapolated to the larger universe.

Testing Boundaries

As with all statistical analysis there are boundaries that you have to draw around estimates. However, the figures above seem to be validated by estimates privately made by some of Venture Economics fund-of-funds and advisory clients.

Some of the long-term issues for the industry are fairly easy to ascertain. Large amounts of capital in proportion to demand will probably lead to lower returns for the industry. Fund raising will be more difficult because LPs are already flush with investments. And, there will be increased competition to invest in deals.

What is not certain is whether the effect will be felt across all firms. Perhaps firms with long track records will be spared, while those who have neither the capital nor the confidence of their LPs will go away.

Jesse Reyes oversees research for Venture Economics. He can be reached at