- Does US Dept of Energy (DOE) Distort CleanTech VC Co-investment Networks? - A Social Network Analysis
The study period is 2005 to 2011. The empirical data on 637 VC firm co-investment relationships comes from CrunchBase, which draws on public news on investment rounds and such. Here's what the co-investment network looks like:
Here's the degree distribution of nodes (VC firms). From this you can see that DOE is highly connected and therefore important, but clearly not the only one:
Here are the hypotheses that were tested:
- H1. The co-investment network structure is governed by a preferential attachment mechanism (‘rich get richer’) centered on DOE, and therefore exhibits a power law distribution in degrees.
- H2. Removing the DOE will significantly distort the network topology.
- H3. VC’s who co-invest DOE suffer a status drop in subsequent syndication transactions.
- H4. Ventures that receive DOE funding are less attractive to high status VCs in subsequent funding rounds.
All of the hypothesis tests failed to reject the null hypothesis, and therefore none offer support for the assertion that the DOE distorts the CleanTech venture capital market. All the details of the tests are in the presentation, above. (Note, these tests don't prove that DOE is helpful, smart, successful, etc. at CleanTech investing. Only that it doesn't "ruin the neighborhood" for VC or ventures.)
I also created a history-friendly simulation to explore the causal mechanisms for co-investment that are plausible. After much exploration (a.k.a. fiddling and trial and error), I was able to create a simulation that came close to replicating the empirical case. What's most interesting is that it was necessary to have a heterogeneous population of co-investment strategies. This, too, supports the conclusion that the CleanTech VC network is 'healthy' and does not suffer from the distorting effects of the DOE (or any other single VC).