VC life can get pretty busy, and for much of 2018 I fell behind at keeping my SaaS company models up to date. As market conditions have changed, I've gotten a bunch of requests for an update- where do multiples sit? Are SaaS companies overvalued or undervalued? Tough questions to answer (especially the latter one), but I decided to take a look to what insights I could find in the data.
To answer them, I've spent some spare time refreshing models for 33 public software companies (one day I'll get to adding some of the newly IPOed ones). The preliminary results are pretty counter-intuitive and there are some neat dynamics at play.
First, a quick refresher- I value recurring revenue companies using EV/Gross Profit (read why here). In the past, I've found through trial and error that investors tend to value companies based on a combination of the growth rate and the free cash flow margin the company earns (or, in the case of startups, the burn rate required to fuel that growth). It makes sense- all else equal, higher growth is better and so are higher margins. These two variables have historically correlated to the multiple on gross profit investors are willing to pay for public software companies (the only ones for which we have such detailed data). Here's what the chart looked like just over a year ago:
So a company with a 40% growth rate and a 10% free cash flow margin (40%+10%=50% on the x axis) traded at about 12x 2017 gross profit in October 2017. I used a regression to estimate attribution between growth and free cash flow margin, and found them to be approximately equal- investors were treating a company growing 40% with a 0% margin and a company growing 0% with a 40% margin about the same.
This year, as I redid the work, the correlation was much less impressive visually. I was intrigued, so I reran the regression, and the correlation between free cash flow margin and multiple has disappeared. In today's market (even after the recent decline) investors are valuing software companies almost exclusively based on their growth rate:
To make sense of this, I took a look at another chart I use frequently, a simple scatter plot of 2018 growth rate (y axis) against free cash flow margin (x axis). As always, the free cash flow margin is adjusted to penalize companies that issue a ton of stock based compensation, which otherwise isn't accounted for as an expense:
Cross referencing the two charts, a few points stick out:
1) At the high end, there are four companies in a league of their own when it comes to growth (at least, among the group I have models built for): MongoDB, Okta, Shopify and ZScaler. These are the same four companies that really drive the correlation between multiples and growth- all four trade at 24-28x 2018 gross profit, some of the highest multiples I've ever plotted. Though their margin profiles vary, there is no relation between their margins and the multiples they trade for within the group.
2) Investors still do pay-up for cash flow- the issue in the data is few public software companies have much of it. Ansys (ANSS), Aspen Tech (AZPN) and Veeva (VEEV), the three most cash-generative companies, all trade at a premium multiple vs. what their growth would imply alone. For the rest, there's no discernible pattern- a muddled middle.
How'd we get to a place where the equation I was so proud of lost its relevance? There's only one way- high growth companies have outperformed so far this year, meaningfully. This chart makes the relationship clear:
If this seems like an obvious point, remember that it is not- software growth rates are relatively predictable year to year, and so the correlation doesn't look too different if I use the 2017 growth rate. Investors have clearly changed their appetites in 2018 in favor of growth, without too much regard for the costs.
At this point, I had one last curiosity- was this a result of the ebullient markets earlier in the year, with investors now retreating to slower-growth names in the last few months? To answer, I picked a date for the "peak" (10/3, the date markets really starting rolling over) and took another look at returns vs. growth:
The answer is a resounding no- the higher growth names have been more resilient since the NASDAQ peaked. It is not a perfect correlation and there are stock specific factors (earnings misses at companies like ELLI and APTI), but still, there is no evidence of a retreat to cash generative companies.
At least for now, software investors are back to caring almost exclusively about growth.