Venture Investor’s Playbook: Part 2
The Power-Law of Venture Returns
Chip Hazard; General Partner, Flybridge
This post is Part 2 of a series on Chip’s Playbook of 4 Success Factors for Early-Stage Venture Investors. You can see Part 1 here.
Success as a venture investor is defined by being in the top 10–25% of your peers and the only way to get there is by investing in outliers that drive massive returns. Apart from the financial gains, there is also a marketing and awareness benefit of being associated with huge wins.¹
As a result, it is crucial for venture investors to fully internalize the power-law of venture returns. The power-law states that a small number of wildly successful investments drive overall returns for the industry, individual funds, and specific venture investors within those funds.
We can see the power-law at work at both a fund level and a company level. First, let’s look at mature fund-level returns. If you go back ten years, for example, the pooled return for venture funds raised in 2009 and tracked by Cambridge Associates is 1.96x invested capital. These funds barely outperformed the S&P 500, which returned 1.8x over that time frame. The top 25% and the top 5% of the firms, however, returned more than 2.2x and 3.5x, respectively. As you move from the top 25% cut off and towards the top 10%, the more respectable returns compensate investors for the illiquidity inherent in venture investing. This is illustrated in the chart below.
For the years 2002–2012, which excludes both the late 90s, early 2000s bubble/crash years and more recent years in which portfolios are still immature, the pattern looks the same: pooled returns approximated the S&P 500, the top 25% delivered respectable results, and the top 5% drove extraordinary returns as shown below:
In conclusion: as a fund manager, you only prove your worth if you’re in the top 25%, and you’re only doing really well if you’re in the top 5–10%. How you get there will be the topic of subsequent articles.
The second cut of the data that’s illuminating to look at is the individual portfolio company investments that make up fund-level returns. Sticking with 2009, the first year in which the funds highlighted above would have made initial investments, there were 672 early-stage investments made in the Cambridge Associates dataset that generated a pooled return of 2.7x.² The company-level return distribution for these investments as of the end of 2018 was as follows:
This high-level data, which is how the industry reports returns, obscures two facts. The first is that the vast majority of industry returns come from the top 10%. The second is that if the top-returning company caps out at 5.6x, portfolio returns will be mediocre (more like 1.6x versus 2.7x).
We asked Cambridge for what the top 10% looks like, and that’s where the data gets interesting. The 69 US early-stage investments that comprise Cambridge’s top 10% from 2009 (measured by total value as compared to invested capital, otherwise known as TVPI in the industry’s jargon) had a:
- Minimum TVPI: 5.6x (the cut off for the top 10% in the table above)
- Maximum TVPI: 254.9x
- Mean TVPI: 18.6x
- Median TVPI: 8.5x
- Weighted Average (by invested capital): 15.4x
So, in fact, it is not that the vast majority of industry returns come from the top 10%, as I said above, but really from the top 5% and from within that top 5%, there are likely less than 10 companies that returned more than 30x. Using this data at a company level, the power-law of venture returns can be visualized in the two charts below:
In other words, the top 10% of the companies generated 57% of all returns for the investments made in 2009, and the top 25% of investments generated 85% of total returns.
While Cambridge Associates does not disclose specific companies, looking at Crunchbase reveals 2009 initial investments in companies such as AirBnB, Twilio, Square, SendGrid, and MongoDB (a Flybridge portfolio company I wrote about here), all of which certainly generated returns greater than 50–100x. I can guarantee the top 10% of funds from 2009 were invested in one of these or one of the other five or so companies that returned more than 30x.³ Invest in one of the 1–2% of companies funded in a given year that deliver exceptional returns and you are a hero, if not you are average.
Another observation from the 2009 Cambridge data is that more than half the investments lost money. Losing half the time sounds terrible, and a lot of investors and fund managers spend an inordinate amount of time understanding their losses. However, as long as your losses are not outsized in dollar terms relative to the winners in your overall portfolio, they don’t matter as much as you’d think. As the old venture saying goes, you can only lose 1x your money. To illustrate this, in a portfolio of equally sized investments from 2009, if all the losers turned into get-your-capital-back investments, it would improve performance by only 15%, whereas missing out on the top 5% would decrease performance by 66% and make it hard to generate attractive returns. Instead of focusing on investments you made that lost money, focus on what will drive more outliers into your portfolio. To learn from mistakes, focus on what outlier investments you missed and why. Here is a link to one painful mistake I made and my analysis of what I missed.
Finally, I would argue that the power-law has become more critical in venture capital over time as winners generate outsized returns more quickly and at a greater scale than before. Interconnected global markets and the power of network effects, whereby the strong get stronger with more success, results in winner-take-most markets on a global scale.
From what I remember of the top-performing Greylock funds from when I was there, there was a fatter distribution curve with fewer losses, more strong returns, and relatively few 100x returners. I suspect the lower impact of the power-law in the 1980s and 1990s was because of these factors.⁴
It is important, however, not to translate this understanding of the power-law into how you work with founders. Pushing every company, often prematurely, into being an outlier can be a disservice to founders. Similarly, just because losses matter less in early-stage venture capital, working hard on behalf of companies with issues is important not only in terms of fulfilling promises made to founders, but also to turn potential losses into gains.
In summary: success in venture investing is defined as being in at least the top quartile, if not the top decile, of investors. For early-stage investors, that requires investing in outliers — companies that can generate 50x return on your invested capital. This mindset is critically important. Avoiding fine-but-not-great companies is more important than taking a swing at a potentially great opportunity that may be a miss. Because when you invest behind an outlier, it makes all the difference in your fund’s performance.⁵
Part 3 covers how identifying macro-market trends early, and riding the waves of growth they create, is the next key to VC success.
 Name a top-tier venture firm, and their big wins easily come to mind: Sequoia (Google, WhatsApp), Accel (Facebook), Greylock (LinkedIn), Union Square (Twitter), etc.
 Pitchbook reports over 1,800 early-stage investments in 2009. The Cambridge Associates data of 672 is a subset, and likely biased upwards given the firms CA monitors.
 Another example of the power-law at work in a specific fund is Boris Wertz’s 2012 results at VersionOne posted his 2012 results. The data shows that with a 40x and a 10x returning investment in the portfolio, they’re doing great; absent those two companies, their performance is more average.
 Hoffman and Yeh refer to this in Blitzscaling as well.
 There are three excellent articles on this topic: Jerry Neumann’s post on “Power-Law in Venture” (also covered in a recent “Invest Like the Best” podcast), Clint Korver of Ulu Ventures’ post on the same topic, which I’ll return to later in this series when discussing portfolio management, and Correlation Ventures’ article “No, We’re Not Normal”.