A Brief History of University Startup Exits

By: Manny Stockman

  • Cutting-edge research?
  • Experienced leadership running a team of sharp 20-somethings?
  • Core technology and science addressing established problems in a hard sector (with millions of dollars in non-dilutive grant funds)?

Most startups would be lucky to check the boxes on these questions. But in the world of university startups, doing so is the minimum exit velocity that any professor, grad student, or post doc must meet to spin out a company from their institution. Unfortunately, no checklist is sufficient to guarantee a university startup’s success; fundraising, sales, and scale-up are topics that aren’t covered when PhDs study for their qualifier exams.

But, fortunately, there have also been a large number of great university startup exits that illustrate the potential of the technologies spinning out of these labs, and the success of these exits inspire the academic entrepreneurial flywheel (link).

At OUP, we have found the university startup ecosystem to be a rich area for investing. I thought it would be worthwhile to revisit the interesting statistics around university exits, and use this post as a means to give credit to the academic origins of many well-known companies. We do a decent job of tracking this kind of data for the 80+ universities that have partnered with us, although it is far from perfect.

In our system we have recorded ~400 startup IPOs and acquisitions that are attributable to research that was performed in the depths of university laboratories. Over 300 of those liquidation events had exit valuations over $50M, a threshold above which venture investors often begin to see good outcomes. I expect there to be inaccuracies in some of the exit values, as well as some missing data points (please let me know). The data spans just over two decades of exits:

Scatter plot of startup exit valuations vs. venture capital raised. Our data is imperfect, so I expect some great startups have been left out. Not all company names could be shown on a single plot. Circles and crosses distinguish tech startups from life science startups, respectively. Public company IPO exit values are calculated in one of two ways: the majority are from OUP’s fund model, which uses a 6-month, post-IPO average market cap; and for newer exits we simply used the post-IPO market cap. (Sources: OUP Data, PitchBook Data Inc., public investor documents)

You’ll notice this is a log-log plot. Blame Stanford spinout, Google. Their 2004 IPO for ~$23B is such an outlier that I am removing it from the dataset for the remainder of the visualizations in this blog post. Ideally, a startup is up and to the left — a region primarily occupied by software startups. On top of making really good web browsers (Google, Netscape), some of the most successful university startups have set the course for virtualization as we know it today (VMWare) and have enabled bloggers to create sharp-looking plots (Tableau).

Sector Distribution

The university ecosystem stretches across all sectors, and so we have two sector-focused teams at OUP: a life science team and a tech team. Here’s why:

(Sources: OUP Data, PitchBook Data Inc., public investor documents)

It’s good to see that the value of two very important sectors in today’s investing landscape are mirrored in the university sector data. The uneven distribution in favor of Therapeutics and IT exits should encourage active VC firms to explore university deal flow.

Given this exit data, the real question is whether universities today are learning from exit trends and spinning out companies in the sectors that are most likely to lead to a liquidation event. Since this is a blog post on exits, not the active university funnel, I’ll just quickly interject that our database reveals that about 30% and 20% of the current university startups in our system are in the Therapeutics and IT/Software sectors, respectively.

If I were to dive deeper into the exit data, it would be interesting to see which sectors in the academic entrepreneurial space have been the least efficient at producing value for investors (e.g. sector exit dollars divided by sector investment dollars). My instinct is that it would be Cleantech/Energy given the fast pace of investing in that sector years ago and the lack of positive outcomes. This is not to say that universities are not leading the world in cutting-edge research in that field, but commercialization proved to be much more complicated than anyone anticipated for solar and battery technologies.

Time to Exit

One of the questions we often field at OUP is whether the raw technology emerging from the university ecosystem tends to create startups with longer times to exit than those borne by normal means (as an aside: what is “normal” for core technology startups?). When I look across the landscape, subjectively, I can find just as much diversity in the technology readiness of startups in many top VC firms’ portfolios as in the university ecosytem. Objectively, I consider the number of years that it has taken successful university startups to exit:

Scatter plot, by sector, of Years to Exit vs. Exit Year for over 300 university startups that had an exit valuation greater than $50M. The diameter of the circle is proportional to the exit valuation. Years to Exit is defined as the number of years between the company’s founding and its exit. (Sources: OUP Data, PitchBook Data Inc., public investor documents)

The flood of therapeutics IPOs in the past several years is clearly seen in the overwhelming number of yellow circles on the right side of the scatter plot. In the tech domain, Netscape, Oculus VR, and Akamai stand out as +$B exits after less than two years of operation. After filtering this data for deals that have had exits of at least $50M, only Therapeutics (n=155) and IT/Software (n=47) have enough samples to provide meaningful statistics.

Not quite a Gaussian, but helpful to see that the peak of the distribution of Years to Exit for IT/Software and Therapeutics is 6–7 years, well within the time frame of venture capital. (Sources: OUP Data, PitchBook Data Inc., public investor documents)

Therapeutics startups exited in an average of 7.5 years, while IT/Software companies exited in an average of 6.4 years. Looking at the histogram, it seems reasonable to assume that most successful university startups within these two sectors will exit in 4–8 years. So even if a university spinout has earlier stage technology compared to the average venture-backed startup, some other forces at play are able to accelerate exits for successful companies within attractive timelines for venture. I personally believe that the magic formula is a motivated academic paired with an experienced core-technology entrepreneur. Alums of the academic flywheel understand this, and so do the university tech transfer offices that are doing more work than ever to promote those marriages.

University Productivity

On that last note, some fantastic reports have recently been conducted that researched which university undergraduate programs have produced the most VC-backed entrepreneurs (link). Much career development happens after one leaves one’s undergrad institution and before one actually becomes an entrepreneur, so that report is less useful to VC firms looking to understand what happens inside a university. If you are interested in the sheer entrepreneurial productivity going on within the halls of a university, you may appreciate this listing of institutions whose startups have exited (in total) for greater than $2B:

Caltech
Columbia
Duke
Emory
Harvard
Hebrew University
Illinois Urbana-Champaign
Johns Hopkins
Michigan
MIT
Penn
Princeton
Scripps
Stanford
UC Berkeley
UC Los Angeles
UC San Francisco
Univ of Colorado-Boulder
Univ of Southern California
Yale

Given the uptick in and resources being devoted to university entrepreneurship on campuses across the country, I anticipate this exit dataset will only grow larger over the next several years. Our goal at OUP is to help instigate these successes by speaking with academics about their entrepreneurial ambitions, and to hopefully find a way to be a part of a few of their great exits.

The data presented in this blog post is derived from a variety of sources including OUP’s internal database, PitchBook Data Inc., press releases, and publicly filed investor documents.