Is Venture Capital Broken?

A recent report by the Ewing Marion Kauffman Foundation raises serious questions about the degree to which venture capital deserves emulation.

The report, provocatively titled "We Have Met the Enemy and He is Us" [PDF], summarizes its findings thus:

Limited Partners — foundations, endowments, and state pension funds — invest too much capital in underperforming venture capital funds on frequently misaligned terms. Our research suggests that investors like us succumb time and again to narrative fallacies, a well-studied behavioral finance bias.

One of those narrative fallacies relates to the most basic thing that investors care about: results. Most of the funds in which Kauffman invested failed to beat public market indices, despite the higher-risk nature of their work. Kauffman found that larger funds in particular provided disappointing returns.

Kauffman didn't say it was opting out of venture capital. Rather, it plans to seek smaller funds, shift some of its money towards public equities, and co-invest in later-round deals with seasoned investors. It also called on investors to work to better align investor and venture capitalist incentives.

Within Innosight, where I lead venture investing, the report led to an invigorating debate about the industry. We debated two possible stories behind some of the report's statistics.

The first story relates to plain old-fashioned competition. Over the past decades, foundations, pensions, and other investors have poured money into venture capital, particularly in the United States. The number of funds has exploded. If one assumes that the number of great startup ideas is relatively stable at any given time (a contentious but simplifying assumption), increased supply has to decrease returns. A result that is exacerbated by the simultaneous decrease in the cost of innovation and the rise of new approaches, like Y Combinator's systematic incubation factory, Kickstarter's crowd-funding platform, and super angels like Ron Conway. Barriers to entry are decreasing and disruptive entrants are surging, a recipe that both Michael Porter and Clayton Christensen could agree augurs poorly for industry returns.

The alternative and potentially more troubling story is that the venture capital approach is fundamentally flawed. The industry has unquestionably helped to support the formation of world-changing companies such as Microsoft, Cisco Systems, Google, Facebook, and countless others. Yet over the past decades, venture capitalists have shown signs, despite all their experience and skills, that they've not gotten better at what they do.

Maybe that's just the way it is. You can't know ex-ante which idea is the right one. If that's the case, though, it would be far better to simple spur more Y-Combinator-like incubators that follow the Steve Blank gospel of getting out of the building and iterating to discover product/market fit.

A related possibility is that a majority of venture capitalists use flawed theories of startup success. Clayton Christensen likes to describe how, in the early stage of theory development, people make predictions based on observed correlations. For example, people observed that things that flew had feathers. So people hoping to fly created large feathered wings. More advanced theories pinpoint causal mechanisms. For flight, that was Bernoulli's Principle, a theory that explained the concept of lift and why modern aircraft are not covered in feathers.

Much of the venture capital industry seems stuck in the feathers-and-wings stage of theory development. Many successful venture capitalists observe directional patterns. "Every time I have succeeded," they might think, "I've backed an 'A team,' that has targeted a hot market space." There's no doubt that's true, but that is a statement of correlation, not causality. Even worse, the narrative fallacy means that people are likely to construct stories about the past that might not have been precisely true, making future predictions even more dubious.

Google Ventures is an intriguing example of a venture organization putting more science behind its correlative analysis. As detailed in a recent Fast Company article, it has a team of statisticians crunching data to determine patterns of success in startups. For example, its team found the perceived wisdom that failure is beneficial isn't backed by data. Almost 30% of businesses founded by someone who had succeeded once succeed again, versus 15% of businesses founded by someone who failed once.

The Google Ventures approach has its limits. Without understanding why a pattern exists, making reliably accurate predictions is hard, at best. This is where good business theory comes in. Work by Christensen and other innovation thought leaders has built a body of theory about why certain ideas succeed and others fail. Other academics, in particular Noam Wasserman at the Harvard Business School, have provided useful theories to inform the critical choices facing entrepreneurs.

It's not hard to see the venture capital industry trifurcating. There will always be top-tier firms who give their seal of approval to a select group of startups. Expect to see continued growth among incubators that support fast-cycle iteration. And in between, look for a rise of companies that take the Google Ventures approach even further to approach the funding of startup companies scientifically. The rest? Well, Darwin has a way of working out such problems.

Press spacebar to pause and continue. Press esc to stop.

Bloomberg reserves the right to remove comments but is under no obligation to do so, or to explain individual moderation decisions.

Please enable JavaScript to view the comments powered by Disqus.