Notes
1. In my comment paper (Cumming 2011), not only did I replicate the same summary statistics as Da Rin, Nicodano, and Sembenelli (2006) in Exhibit 1, I also extended the sample to include OECD data to show what happens when you extend their measures to other countries. Da Rin, Nicodano, and Sembenelli (2011) argue this extension is a mistake or misunderstanding on my part [which is bizarre because economists normally like out of sample tests!], so here in this reply to their reply I will focus on the data used in their paper to highlight why their measures and arguments are flawed.
2. Similar differences are observed when one compares high-tech venture capital with non high-tech VC (as done in Da Rin, Nicodano, and Sembenelli 2006). Here, for reasons of space, I focus on early stage/late stage venture capital, but the same problems apply to Da Rin, Nicodano, and Sembenelli's (2006) use of high-tech/non high-tech venture capital.
3. Venture capitalists predictably invest in early versus late stage investments over time for reasons presented in Cumming, Fleming, and Schwienbacher (2005) (see also Cumming, Fleming and Schwienbacher's 2001 working paper). Gompers et al. (2008) confirm this point using similar regressions as in Cumming, Fleming, and Schwienbacher (2005).
4. See Cumming, Fleming, and Schwienbacher (2009) on the presence and determinants of stage style drift among venture capital and private equity funds.