Published in UT San Diego, July 14, 2014
Birds of a feather get fleeced together. At least that is the result of a study on venture capital done by our friends at the Harvard Business School. It seems that there is a tendency for venture capitalists “whose ethnic and educational backgrounds are similar to their own” to team up together. And in so doing, they lose some of their critical thinking, according to “The Cost of Friendship,” written by Paul Gompers, Yuhai Xuan and Vladimir Mukharlyamov.
In other words, when you go to the valley and the lead VC wants to “syndicate” the deal, he calls his buddies, his pals, his cohorts — who are very much like himself. And the entrepreneur is somewhat at the mercy of the VC cabal. For example, I have been told that there have been instances of collusion on pre-money valuation, but don’t quote me on that dirty little secret.
On the other hand, the lead VC wants some reassurance that he is not out of his mind, so finding another true believer gives comfort. However, the Harvard study says the following, “the closer affinity between two venture capitalists who co-invest in a new company, the less likely that the company will succeed.” It feels a bit like “Sophie’s Choice.” The lead VC will not invest without another VC and if they know each other, your company is doomed.
The theme is that in an early-stage company, you want people around the table who will challenge each other and the basic business assumptions. Even more chilling is that when two VCs co-invest, they have similar funds and attitudes and personalities. They are equals. And this cuts across all boundaries, including where they went to school and ethnicity — and across these sectors, the percentage decrease in success was approximately 20 percent. The VCs would do better to pick their partners asymmetrically rather than argue with the entrepreneur about a 20 percent increase in the pre-money.
Now comes the fishhook. The lack of success was not in picking bad deals; it was in the decisions that followed the investment. The basic job of the board — hiring and firing the CEO, devising strategy, choosing other board members, identifying strategic partners — that is where the negative affinity effect was most pronounced. The researchers call this “the cost of friendship,” and they attribute it to groupthink.
On a personal note, one of my companies is about to take a Series A round of investment. The only issue that was contentious on the term sheet was “who was going to be on the board.” There were a few people with whom we were less comfortable in terms of personality — but at the same time, they could add real value. Needless to say after sharing the above research we opted for more “diversity” even though it will mean biting our tongues on occasion.
Lastly, allow me to comment on Aereo, the platform for streaming network television shows, which the U.S. Supreme Court recently ruled was illegal. Five large venture firms, including one with Barry Diller as the lead, invested in Aereo. As of the end of June, they decided to close the company, and the loss was $97 million. This was an interesting bet, because there was “no plan B,” according to Amish Jani, founder of First Mark Capital, an early investor. It was a “bet the farm” type of deal, and Mr. Jani said, “We’d do it all over again.”
Chris Fralic, First Round Capital, knew there was a risk — he says, “that’s why they call it venture capital.” I love his honesty, but at the same time, there is a famous book, “Getting to Plan B” by Randy Komisar, and of course the main premise is to be willing to iterate and evolve the elements of the company. But sometimes “push all the chips into the center of the table” is the right strategy. After all, if Aereo’s technology had been ruled legal, the $100 billion network television industry would have been turned upside-down.
So betting $97 million against that market size. Yeah, I see the point. But I wonder if the VCs who were all friends when they funded — are they still friends?
Rule No. 359
With friends like these, who needs enemies?
###