# From Centaur to Unicorn in 30 Months

In an article posted in July 2020, I wrote about VC-backed company valuations and how they’ve evolved over the last 20 years. Specifically, I looked at the top 15 highest valued startups around the world — what I unimaginatively call “winners” — and how their relative quarterly valuations changed from 2000 to 2020. Some of the observations I made in the article were obvious, namely, winners are getting bigger and emerging globally. Other observations weren’t so obvious: the gap between big winners and *really *big winners is widening, valuations of winners reacted differently to the 2000, 2008, and 2020 macroeconomic shocks, and winners are winning faster. This last point is the topic of today’s article. To start, it’s worth unpacking exactly what I wrote back in July:

Winners are winning faster.The implied conditional probability of garnering a $1B+ valuation within 30 months of reaching a valuation of $100M+ has increased from 4% in Q1 2010 (N=25) to 9% in Q1 2017 (N=109).

In other words, the share of “centaurs” (companies valued at $100M+) that convert into “unicorns” (companies valued at $1B+) within 30 months of becoming a centaur is increasing. For example, in Q1 2010, 25 companies raised a financing round at a valuation north of $100M. Of those, only 1 (4%) went on to raise at a valuation north of $1B within 30 months of reaching centaur status. In Q1 2017, just 7 years later, 109 centaurs were minted, with 10 (9%) converting into unicorns within 30 months. Here’s what the trend looked like in the intervening quarters:

If we’ve established that winners are getting bigger and, by extension, centaurs are minted at increasingly high valuations, then it’s no surprise that the centaur-unicorn conversion rate is increasing. After all, it would make sense for a company that goes from a valuation of $90M to $500M to have a better shot of reaching unicorn status than one that goes from $90M to, say, $150M. Here’s the thing: the median first-time centaur valuation has stayed flat from 2010 to 2017. The same is true for first-time unicorn valuations. This is what the trend looks like from 2010–2017:

So the share of centaurs that convert into unicorns within 30 months of becoming a centaur is increasing, but these milestones are admittedly a bit arbitrary. Generalizing to a range of starting and ending valuation bands would surely be more informative, particularly if we also sample by sector or geography. This will be the topic of a future analysis. It is worth noting, however, that the 30-month window was derived empirically. When searching for the window following a company’s centaur round that yields the statistically most significant relationship between centaur-unicorn conversion (within said window) and time, the optimal choice is 30 months (see below Poisson-like distribution). I hesitate to call this a “make-or-break” point for a centaur’s shot at becoming a unicorn. Such a conclusion by itself would be reductive. What’s clear to me is that this 30-month window is *not* arbitrary.

The reason we need a window in the first place is to compare data from different time periods. Of the centaurs minted in Q1 2010 and Q1 2017, for example, the share that *eventually* become unicorns (3/25 vs. 16/109, respectively) will be artificially inflated for Q1 2010, discounted for Q1 2017. This is because there are centaurs minted in Q1 2017 that will one day become unicorns, but haven’t yet had the time to scale. Thus, we’ll only be able to get an accurate comparison between different time periods if we look at a window of time following the quarter in which a centaur is minted. The optimal window happens to be sufficiently small for us to draw meaningful conclusions today. I feel confident saying that the odds of becoming a unicorn (within 30 months) for a centaur minted today is about 10% ± 5%.

This brings me to my next point: the subtle distinction between talking about “probabilities” vs. “shares” or “conversion rates.” By way of analogy, if we have 100 fruit and want to know the share of apples that are green, we start by counting the number of apples among the fruit, then we count the number of green apples among the apples. Say there are 10 apples, 3 of which are green. Stating, “the share of apples that are green is 30%” is akin to stating, “given the fruit you have is an apple, the implied probability that it’s green is 30%.” The former language is more intuitive, the latter lends itself to pithier, punchier conclusions. Now that we have a useful analogy — fruit are VC-backed companies, apples are those valued at $100M+, and green apples are those that reach a $1B+ valuation in ≤ 30 months after passing the $100M mark — I hope the slight semantic difference between implied conditional probabilities, P(unicorn | centaur*), and conversion rates is made clear.

A keen eye will realize that we haven’t once talked about the *rate *of valuation growth, just the odds of jumping from one valuation to another. The thing is, the rate of valuation growth hasn’t changed much. If we look at the time it took to go from centaur status to unicorn status for the companies that did so within the 30 month window, the average rate is pretty much constant at 1.7 years, or 20 months. Here’s the supporting QoQ scatter plot:

The upshot is that winners aren’t actually winning faster. Rather, the frequency of centaurs converting into unicorns within a well-defined window is increasing. I have two hypotheses for why this is happening: (1) more capital in the private markets driving more consolidation around emerging winners, or (2) the median valuation of *outlier *centaurs is increasing QoQ, but the overall median valuation of first-time centaurs remains constant around $200M (see above). The latter theory would confirm my suspicion of higher-valued centaurs having a better shot of passing the $1B mark in ≤ 30 months. Overall, it’s important to recognize that the relationship we’ve examined is hardly a relationship at all. An R² of 16% means we see only a sliver of the the big picture. A VC-backed company valuation is dependent on dozens of variables, not least of which are fundamentals like revenue growth and steady state profitability. The goal of this study was not to derive some airtight relationship that we would expect to hold for years to come (e.g., Rule of 40). If 2020 has taught me anything, it’s that the world is largely unpredictable. The purpose of this article was to explore a pattern, however tenuous, and lay out my findings as clearly as possible. I hope this exploration is insightful for entrepreneurs, investors, and all those with a fascination threshold as low as mine.

Disclaimer: Views are my own and may not reflect those of my employer.

Source(s): Thank you to Bailey York at PitchBook, Inc., for helping with data acquisition and teaching me what “centaur” means.

*Due to a lack of data visibility, some companies are quoted as passing the $100M and $1B marks in the same quarter. I excluded these cases from the above analysis. For companies that passed the $100M+ mark, raised at a down round below $100M, then proceeded to raise a round at a $1B+ valuation, I looked at the time between the *first *centaur minting quarter and ultimate unicorn minting quarter. There were very few of these cases.

Support: low sample size affects the relationship explored above.