Neat strategy. It would be interesting to actually browse through the study's data set and isolate Founder-CEO companies. If there's any conceivable link, then maybe that explains the 30%? However, I don't think there would be many founder CEOs within the dataset.