How does Venture Money Change with the Market?

Photo Credit

No surprise, most flows of capital in the venture ecosystem are procyclical. When the markets are hot, venture capitalists, corporate venture capitalists, and angels all invest more capital into startups, and limited partners invest more into venture funds. When there is a downturn, the capital flows slow down across the board. There are organizational and inter-organizational dynamics that lead to these effects, but the underlying cause is human group psychology. When we actually quantify the changes, we find they are startlingly large.

The Data

The data in the graph and table below comes from the National Venture Capital Association and the University of New Hampshire Center for Venture Research (public data, plus private communication with Director, Jeff Sohl). Unfortunately, the available angel investment data does not reach back through the dot com period, and is only available on an annual basis for 2001 through 2003; semi-annual since then [1].

Some Ratios

Let’s compare the peak-to-trough ratio from the height of the dot com boom to the subsequent trough, and from the trough of the financial crisis to our recent warm market “Peak” [2]. The data in the table below is drawn from the same data as the graph above.

Average Quarterly Capital Flows in Billions of dollars, for specific four-quarter periods of time, with Peak/Trough Ratios.


The table and graph above show some really interesting things:

  1. The magnitude of the changes in capital flows are quite large. As is obvious from the graph above, and quantified in the table, the changes in capital flows are many-fold, not just tweaks around the edges. It is hard to overstate the impact of this effect on most aspects of the venture ecosystem. These changes are so large as to completely overshadow most other effects we see in venture.
  2. With the benefit of hindsight, these extreme swings in capital flows seem wrong somehow. But, at the time they are happening, it is very easy to get caught up in the prevailing market sentiment. If the head of your organization is screaming about lack of liquidity in your organization’s investment portfolio, the bias against committing any incremental dollars to the inherently illiquid venture asset class is very strong. Also, while these procyclical capital flows have an element of suboptimal “buy high, sell low” to them, the best returns in venture come from investing in the years leading up to a peak in the exit environment. It may be the case that “at the time of a new venture commitment, the best predictor of the market conditions at the time of exit is simply the current state of the market.”
  3. There were a number of large late-stage private company financings that were done in 2014 where much of the money came from non-traditional investors in venture ecosystem. This money is included in the VC Investing numbers above, but would not have corresponding VC Fundraising amounts. This means (i) the VC Investing amounts in 2014 are overstated for many types of longitudinal comparisons, and (ii) the gap between the 2014 VC Investing and VC Fundraising is driven significantly by this.
  4. Of the four types of capital flows in the venture ecosystem, Angel investments into Seed/Startup companies are the least affected by general market conditions. To be honest, this is the exact opposite of what I expected to see. Perhaps there is something about not being part of a formalized investment program that allows the Angels to avoid organizational groupthink that may be driving the huge procyclical behavior we see in the other three types of capital flows.
  5. As something of an interesting side note… The VC Fundraising data does not include the “negative capital flows” associated with the fund size reductions done in the post dot-com period. These were mostly done at the insistence of the investors. In Q2 ‘02, fund size reductions actually exceeded new commitments to VC funds, so technically VC Fundraising was negative for that quarter! Including these fund size reductions brings the VC Fundraising well below Angel investing for the four quarters in the Post Dot-Com Trough period. This is striking enough to warrant including as a separate line in the table.
Same as the table above, with an additional row: VC Fundraising number reduced by in-quarter fund size reductions, in the Post Dot-Com period.

The main takeaways:

  1. Capital flows in the venture ecosystem are hugely procyclical.
  2. Of all the capital flows, the Angel investments at the Seed/Startup stage are the least procyclical.

[1] For the annual, and semi-annual Angel investment data, I simply spread them evenly across the quarters within each of the reporting periods. It is important to recognize that only a minority of of the Angel investments in the CRV data are into Seed/Startup Stage companies. The CRV data does not break out deal size as a function of stage — in the absence of any data to the contrary, I assumed the round sizes were constant.

[2] Across all of the data sets, I used these four quarter periods:

  • Dot-Com Peak: Q4 ‘99 through Q3 ‘00
  • Post Dot-Com Trough: Q3 ‘02 through Q2 ‘03
  • Financial Crisis: Q4 ‘08 through Q3 ‘09
  • 2014 “Peak”: Q1 ‘14 through Q4 ‘14 (Angel data isn’t available for Q1 ‘15, yet)