Within my (hand selected) peer group of investors round sizes increased by 30–40% for A/B/Late Stage companies from 2013 to 2014, and almost doubled for C Stage companies in Q2/Q3 of 2014 compared to 2013. Is this a bubble? Are we paying overzealous valuations? Are VCs careless and imprudent? Is it just unsophisticated later-stage investors who get sucked into too-high valuations? Are we (as an industry) investing too much money too quickly? Or are the rest of us investing too slowly, given the recent unicorn exits?
I hate to be the harbinger. But I am German, it’s probably part of my 2.9% Neanderthal HV0/R1b1b2a1a1d haplogroup, so here we go: We’re going to be OK. Very Un-German. But sure, I can have some concerns — after all Germans are great at having concerns.
I am a principal at an institutional venture capital company with institutional Limited Partners (LPs). Next World Capital is focusing on expansion-stage enterprise software and IT infrastructure startups. Looking at the venture capital industry through lenses of averages — such as average return or average invested capital or average round sizes — is a dangerous idea. Like in any other industry there are great VCs, successful VCs, and a long tail with all kinds of VCs with all kinds of different goals and objectives. For example, some corporate investors use venture capital as a financial instrument to achieve results somewhere else on the P&L and balance sheets. Similarly, medians only tell you the story that it’s risky and pretty hard to be great as a VC, just as it is pretty hard to create a successful startup. That’s why it’s called venture capital and not municipality bond.
Why I Use Peer Groups
I have a list of investors — individuals as well as firms — that I trust, that I get great dealflow from, that I admire for their remarkable industry insights, and that I co-invest with. That list has currently 57 entries, and I internally refer to it as my “VC Watchlist”.
Whenever pundits, analysts, and doomsday sayers release their latest view on “the demise of the VC industry” or how “Silicon Valley startups are getting funding” I am rarely concerned or interested unless it directly impacts either my peer-group or my fund’s limited partners.
How To Measure The Mindset Of Your Investor Peer Group
I use a variety of tools to track portfolio companies of these investors, Mattermark, Datafox, Quid, and Google Alerts. I filter their portfolio companies by our own investment focus which excludes B2C, biotech, green/energy, etc. If you think of this as my portfolio of preferred dealflow then I have at least 542 A/B/C stage companies to follow closely plus a whole bunch that are categorized as “unknown stage”:
About 1,050 startups from the “VC Watchlist” portfolio had a funding round in 2013 or 2014. About 116 have exited since then. 392 startups (37%) who raised a round in 2013 did not raise a round in 2014, among them 78 startups (7%) who raised an A round in 2013.
Some companies raised their last funding round more than 24 months ago and I am not filtering which of these companies are still operational. The “stage” classification is not always accurate, but for this exercise Mattermark’s data is pretty good.
I can visualize the data by quarters and stages in a tukey boxplot (aka whisker diagram) that nicely shows medians, quartiles, and variances (mind the logarithmic Y-axis!).
Here is the same data with medians and means by stage in a table:
10–30 companies per quarter per round is hard on the lower limit for establishing any sort of stable pattern. But you can see that within my “VC Watchlist” portfolio the median funding round size for A, B, and Late Stage companies increased by 30–40% from 2013 to 2014 and stayed there, while median funding round sizes for C-Stage companies almost doubled in Q2/Q3 of 2014 and then came a bit down again.
Did investors in these 2014 rounds (which were not necessarily led or participated by any of the investors on my “VC Watchlist”) brainlessly pump money into these companies? Beyond anecdotal evidence or gut feeling, were VCs acting careless or imprudent?
Well, we don’t have round valuations, and some of these deals might have come with a lot of structure (participating liquidation preferences with several multiples, full ratchet anti-dilution protection, liquidation preferences for IPOs, etc.). What we do know is that most of these investors will have some return expectation on their capital.
What Are VC Return Expectations?
(This section will have some wild generalizations, but bear with me)
VCs generally would like to make some money and raise another fund. With a heavy dose of generalization, institutional VCs would like to generate a 3x return rate minimum to be able to raise another fund (and also make some money for themselves). Together all investors in a company hold a certain percentage of equity; in early stage companies maybe about 30-40%, in late stage companies together maybe up to about 80%, leaving 20% for employees and founders — for my purpose I want to err on the safe side for VCs, so I am putting out rather generous numbers here.
When investors pooled $2,530m in 223 Series A stage startups and are on average holding 40% per startup, then we have a proxy for a rough total aggregated valuation of all 223 Series A companies in this set, which is $6,325m. But we also know that a certain percentage of these companies are going sideways, just so returning investors’ money. And some startups are going to die. Let’s assume a third of the 223 Series A startups are likely to have a minimal or no return — a rather generous assumption — and about 50% will go sideways . But that means that a small percentage — 17%, or about 37 startups — will have to make up for the expected returns, with more funding pooling in and the percentage of VC ownership in these companies getting higher. In fact, we can do the math now: each of the 37 successful startups currently in Series A state will later on have to exit for an average of $477m if we ignore for a moment that ownerships will change in future rounds (quite a big IF). We can do this for all the amounts above as well as for the total, and it looks something like this:
Exit expectations for companies that raised a round in 2014 seem to be much higher than the ones that did not raise a round in 2014 but in 2013. However, when I look through the list of companies from both years I also do not have an “OMG, startups that got funding in 2014 are so much cooler” moment. There are definitely differences in how segments are approached. Mesosphere, DigitalOcean, CloudPhysics, and Docker are good examples for a shift in investor confidence for 2014 for a new sector. But the “last funding round in 2013" list also contains exits from Aerohive, Jive, LinkedIn, Marin, MobileIron, Mojave, Rally, TubeMogul; as well as active companies such as Lithium, MapR, MongoDB, or Zuora (one of our companies) that all made it onto the list of Venture Beat’s 27 tech IPOs to expect in 2015.
Enthusiasm and optimism is especially high for Stage A startups, even more so for startups that raised their Series A in 2014. Series B startups expectations are lower. Is this maybe a sign of realizing that being successful is harder than most investors and startups thought? Or is this a sign that earlier stage investors are overly optimistic, or trigger happy, or loaded with cash?
Are Series B expectations influenced by less available Series B capital and fewer expansion-stage VCs to choose from which in turn drives down prices?
For SaaS startups their Annual Recurring Revenue (ARR) is often used as a proxy and rule-of-thumb for valuations. At great companies (team, product, traction, growth, low churn) we’ve seen ARR multiples of 8x-12x.
Did startups that raised their last round in 2014 show better and faster growing ARR than their same-stage siblings in 2013, justifying higher exit expectation at time of investment?
Did startups that raised their last round in 2014 have already higher ARR at the time of investment than their same-stage siblings in 2013? Were the 2014 startups bootstrapped longer or had more earlier-stage capital to get further along with their ARR when they raised their A-round?
But just to put this into perspective: If it takes 8 years on average for a currently Series A-Stage enterprise SaaS startup to successfully exit then startups that raised their A-round in 2014 need to either grow at 6% higher CAGR over these 8 years than their 2013 siblings; or start out with 30% higher ARR at time of investment, which at an initial 100–200% annual growth rate really only means about a quarter later funding or $250k or $500k more runway from their earlier round(s).
About 129 late-stage companies that raised their Late Stage round in 2013 or 2014 are expected to have an aggregated market cap of about $81bn based on my assumptions. Is that realistic? How many $627m+ exits (IPOs or M&As) have we seen in the past two or three years in this segment? Could a few outliers really skew the average that much?
How I Make Investment Decisions
As venture investors we are embracing risks and are making informed bets on outperformers (I wouldn’t bet on unicorns, but that’s another story). We’re about “highly asymmetric risk-reward through unfair information and operational advantage”, as my Partner Ben Fu puts it. The median expected outcome should be lower than the average, with a few 10x, 20x, or even 100x outliers massively raising the average. Of course we always assume that we have better information (which of course we do) and that our own startups are winning (which of course they are) while competitive startups (and their VCs) are not. But the average minimum exit expectation is still the same, regardless of different VCs leaning into different assumptions and views. Plus within my “VC Watchlist” portfolio a lot of co-investing is happening and we can’t all be right (and I neither think we’re all wrong).
The numbers above are a great reminder that venture investing is a deal-by-deal business. Recent expectations even within the portfolio of our trusted investor list are pretty aggressive. But as an expansion-stage investor B stage startups are my sweet spot. And here I seem to find a very fertile ground with more realistic expectations around the table.
Running the numbers also reminds me that a rigorous strategic and thematic approach to investing within specific sectors to build deep domain expertise is paramount. Not all of the 757 A/B/C/Late Stage startups can be winners (let alone unicorns).
Another thing should be clear: From that list of 116 exited B2B companies that raised a round in the past 24 months you can not conclude that the past 24 months have not worked out well for all VCs in that portfolio. Venture capital is not a short-term investment strategy with a two-year window. There is a reason why most funds run 7–10 years. It’s not a good sign if you have to argue with your LPs why your last year’s return is not like last year’s return in the stock market. VCs build portfolios for a reason. And there are still 757 A/B/C/Late Stage startups left in my “VC Watchlist” portfolio to realize that potential.
Within these 116 B2B companies we also did see some great success stories such as FireEye or LinkedIn. Some of these exits were great exits with great returns for VCs and their LPs at the time of the exit. Some of them have traded down from their peak and their market cap went below the calculated expected average of $442m, more in line with an expected median. Examples are Aerohive, Rally Software, Five9, Prolexic Technologies, Jive, Brightcove, Cryvera, Klout or Silver Spring Networks. And then there were lots of early exits and exits with “undisclosed prices”.
Gut check: Can top investors meet or exceed calculated averages or estimated medians on this list? Yes. I am not too worried about the industry as a whole, as long as the top quartile can show results. The set of later stage companies alone within my “VC Watchlist” portfolio has many examples with great exit potential such as MongoDB, New Relic, Cloudera, Nutanix, Stripe, Square, Box, Evernote, Blue Jeans Network, AppDynamics, Anaplan, Affirmed Networks, Okta, Taulia, Tintri, Pure Storage, Lithium, Jasper, Lookout, Jawbone, Ciphercloud, … ; and of course our own (by now later stage) investments with great potential such as Zuora, DataStax, Virtual Instruments, Host Analytics, or GoodData.
 with permission from The Metropolitan Museum of Art, http://www.metmuseum.org/collection/the-collection-online/search/468327
 Various answers and links at http://www.quora.com/What-percentage-of-startups-fail point to a 50%-80% failure rate overall.
 all data and graphs can be found on my Tableau Public page at https://public.tableausoftware.com/views/VCWatchlistPortfolio20132014/StageandLastFundingDate?:embed=y&:display_count=no
 I have to thank my outstanding colleagues @benfucious, @tomrikert, @craigalanhanson, and @tarun_kalra for many inspiring discussions we had on this topic. They helped substantially to shape this article.
Thorsten is principal at expansion-stage venture capital firm Next World Capital, an international expansion-stage venture capital firm investing in leading enterprise technology companies in the software, internet and mobile sectors. NWC is the only U.S. VC operating a full strategy, sales and business development platform to help its companies expand in Europe. NWC is headquartered in San Francisco, with additional offices in London, Paris, and Brussels. Our team prides itself on rolling up our sleeves and helping our portfolio companies address the challenges and opportunities inherent in building a world-class company.