This weekend, I spent some time updating the financial models I keep for about forty public software companies, ranging from Instructure to ServiceNow, to reflect their Q1 earnings results. As a self-professed nerd for software metrics of all kinds, I then took some time to play with the newly freshened dataset that results from pulling all of the models together.
I set out to answer a perennial head-scratcher:
How are SaaS companies valued, and what does that tell us about what investors want to see?
First, we can attack the question with some intuition. I've written before about why we need to use high level SaaS valuation multiples in the first place, and why I vastly prefer EV/gross profit (EV/GP) over EV/revenue. Unless a company is nowhere near mature gross margins, that is the best approach. Once I decided what multiple to use, I tried to use data to determine what investors use to drive it. The obvious candidates are growth and profitability- but what kind of growth, and what kind of profitability?
I had some thoughts, but instead I played around systematically in Tableau with some of the usual suspects. The candidates for growth were revenue growth (either 2017/2016 using my projections or 2016/2015 using historical results) or gross profit growth (with the same time frames). The candidates for profitability were adj. free cash flow margin (which subtracts stock based compensation expense), free cash flow margin, net income margin, and non-GAAP net income margin. In each case, I tested the explanatory power by simply adding the two values together. For example, a company with a 35% gross profit growth rate and a 10% free cash flow margin would get a score of 45%, and then I regressed the observed EV/GP multiples against the scores for each company I build models for. Here's what came out:
What the tables tell us:
1) Investors clearly favor free cash flow over net income when valuing SaaS companies. This is a subject for another post but net income is a vanity metric for software companies and investors have caught on.
2) Gross profit growth is slightly better than revenue growth. This makes sense if you think of gross profit as a proxy for recurring revenue, which ties well into why I prefer EV/GP over EV/Rev.
3) Historical results are better at predicting current multiples than expected future results. I would have expected projected future growth/profitability to matter, but investors seem to care more about what has already happened.
4) Investors don't care about share based compensation (whether they should or not is a different matter!). The bulk of the gap between Adj. FCF and FCF and Net Income vs. Non-GAAP income is share based compensation expense, and excluding the expense made for better predictions than penalizing companies for it.
Here is what the best fit metric (2016 FCF margin + 2016/2015 gross profit growth) looks like plotted against EV/2017 gross profit:
Implications for the growth vs. profitability question:
That's a satisfyingly pretty plot, but there's more work to do here and one more surprising conclusion (at least to me). Here is the equation for the above regression, if we break gross profit growth and free cash flow margin into two separate variables rather than blindly adding them together:
What is interesting here is that the coefficients on growth and profitability are almost exactly the same (the actual difference is .0135). In other words, the data shows that investors are ambivalent between an extra point of free cash flow margin or gross profit growth: a company growing 50% with a 0% free cash flow margin and a company growing 25% with a 25% free cash flow margin garner about the same EV/GP multiple. When it comes to growth vs. profitability investor view them as perfectly interchangeable.
On measuring SaaS valuations over time:
When trying to compare SaaS multiples over time, it is far better to use the coefficients in the equation above than a simple plot of multiples over time for a cohort of companies (a staple of VC blogs). The valuation given a certain level of growth and profitability is the right way to frame the question. The underlying metrics of SaaS companies ebb and flow as growth for the category overall slows, and for a true apples to apples comparison of valuations at different times we need to adjust for these changes. Going forward, I'll post quarterly updates for subscribers (more if market action warrants) on how the growth and profitability coefficients that my simple regression spits out change over time.