Last spring I wrote a blog post that used some basic statistics and a dataset pulled from my models of 40 public software companies to find out which variables are best for explaining SaaS valuation multiples. The data showed unequivocally that both growth AND profitability mattered, and that they mattered in approximately equal measure. Somehow SaaS investors always seem like growth hounds and some overly simplistic valuation approaches rely on growth alone.
It is now fall, and I've just finally gotten around to updating all of the models I maintain for second quarter earnings- just a few weeks before third quarter earnings start (perils of having a day job!). Armed with new data, against a backdrop of record highs in the stock market, I took another look using the same methodology to see if investor preferences had changed.
Last spring, a regression using the metrics above revealed this as the best equation for estimating the multiple of a SaaS company:
* Note: I use EV/GP instead of EV/Rev- here's why.
So, for example, a company that grew gross profit 50% at a 10% FCF margin in 2016 would get a multiple of 2.1 + 18*.5 + 18*.1 = 12.9x 2017 gross profit.
Re-running the same regression today, here's the result:
Our hypothetical company that earned a 12.9x gross profit valuation in May would now get a multiple of 14.2x 2017 gross profit with the exact same metrics, good for a 10% return- not bad for a summer, especially with no change in fundamental characteristics.
A few takeaways:
1) Valuations overall as implied by the equation have gone up, all via the intercept. This suggests a rising tide that lifted all boats.
2) There's some slight evidence that investors now put more emphasis on growth, but growth and profitability are still pretty equivalent in their contribution to valuations.
3) Remarkably, the best explanatory power STILL lies in 2016 growth/margin numbers. 2017 numbers work, but not as well. I can't quite figure this out, as most public markets investors are forward looking.
4) This isn't perfect- a few companies dropped from the dataset due to acquisition, both having been on the lower end of valuations (JIVE and XTLY).
When I ran the regression using 2017 numbers, here was the result:
This is also an interesting outcome, with a higher intercept and a clear preference for growth over profitability. I plan on re-running the numbers after Q1 results next year to see if this holds and if the explanatory power shifts to 2017 numbers once they're firmly in the rearview.
Here's how the data looks using the most accurate (2016 numbers). I've included the trendline from the spring in blue- everything else is current.
Of course, this graph is hugely flawed, as it includes no information content past 2016. The data shows that including 2017 information helps, as do some other variables- but good old 2016 numbers remain the best two-variable set, and you're probably viewing this on a 2D screen. =)
Here's the chart using 2017 numbers. Simplistically (and not a substitute for actual investment research/opinions), companies above the trendline are expensive relative to their growth/profitability characteristics, while companies below are inexpensive.