tracking

Venture capital is the highest performing asset class available to investors today, but it is far from a level playing field.

In truth, a large portion of the return in the asset class comes from a few top tier funds and the balance from smaller emerging manager funds. The bulk of other funds underperform.

This has led to a perception that returns in venture are a function of the manager – bright insightful investors with a perception of the future somehow can ordain positive outcomes. In reality, manager skill is not the sole determinant of outcome. In fact, outcomes for early stage investment, from Seed through Series A, is more subject to randomness but favors a certain selection bias.

Think about it like what’s happened in the automotive industry over the past 40 years. Up until the 1970s, advances in automotive design were personality driven, with larger than life engineers stylizing vehicles, marketing a dream. Automotive design was a “craft” driven by the chief engineer.

But that all changed when Toyota entered the market, transforming the automotive industry with a novel (at the time) application of statistical methods to manufacturing. This upstart company from Japan took quality data from its manufacturing process and applied it to design, developing integrated manufacturing methods, which improved quality and reduced cost.

Almost overnight, the entire shifted from a craft industry to what we now call “lean” manufacturing.

This transition from opinion to statistics has since transformed many industries in the same way, and venture is next.

The proof comes in the form of recent data from the Kauffman Foundation. Through statistical analysis, Kauffman reviewed 100s of startups, both successes and failures, to tease out the predictors of success. The researchers found that an investor saw higher returns if they did more diligence on the startups they invested in, if the diligence was conducted by someone with domain experience, if the investors invested in more than 15 deals, and if they kept close track of the company post investment.

In a sense, the Kauffman study found that an investor will do better by avoiding bad deals through deeper diligence. The diligence process remove obviously bad deals, which is a controllable risk, leaving behind only those investments that are subject to external risk factors that are outside of one’s control.

Kauffman’s data suggests that a portfolio needs to include 15 or more investments to achieve venture returns. Further, it suggests that although it can be difficult to determine which of those 15 deals will deliver dramatic returns, the statistics prove that as long as an investor remains invested in all 15 of them a few of the investments will deliver a return that provides a portfolio return annualized at 27%.

Can anyone reliably pick which of the 15 will succeed? The short answer is “no.” But, frankly, at a 27% portfolio return it does not matter.

For decades, venture has been a craft industry, overly reliant on reputation and experience. But these personality-driven venture funds are, frankly, limited in their scope. It is hard for them to successfully invest more than $50M per year because they are burdened with the craft nature of the business.

Skill does still matter, but in today’s market the application of lean, statistical methods to venture provides the opportunity to scale this asset class. Think about it. Scaling up to support investment in Series A deals on the order of $300M-$400M per year seems a bit crazy. But, while uncommon in the venture industry, companies like Boeing and Merck place these types of investments in R&D each year. Why don’t big companies have venture-like returns? Because they limit their investment world to their own internal opportunities, rather than seeking true diversification. Venture investors don’t have that limitation.

And that’s where it gets exciting.