My last post was about the process I used to uncover a simple equation that has significant explanatory power for SaaS EV/Gross Profit multiples. Using just two variables, last year's gross profit growth and last year's free cash flow margin, I was able to get to an R^2 of .69, which isn't too shabby (my statistics background consists of a few low-level college courses, so I know just enough to not make overly broad claims about R^2).
No one who knows me will be the least bit surprised that my next step was to throw a ton of other variables into the regression to see if anything had explanatory power beyond growth and margins. Here are the variables I tested:
1) Does the company have a full SaaS model? A few companies in the dataset don't.
2) Does the company sell to a specific vertical? Software companies in this bucket sell to a specific industry (think Veeva in life sciences or Guidewire in P&C insurance).
3) Is the company based in Silicon Valley?
4) How much stock comp does the company pay employees? I think SBC is a real business expense, and it is not captured by the current framework since it is a non-cash expense.
5) How is the company's "growth + profitability" number (which I creatively call the magic number) projected to change in 2017 vs. 2016? This was an attempt to include some more current data about how the company is expected to perform based on my financial model.
Here are the results of this regression (the R^2 inched up to .75):
The takeaway is surprising: none of the variables mattered, expect for whether the company targets a vertical. Vertical software companies enjoy a statistically significant multiple bump vs. horizontal software companies, even after adjusting for growth and profitability. The bump is not small: a vertical software company, all else equal, has an EV/GP multiple that is 1.8 higher than the rest of the group. Considering the group average is ~10x gross profit, that is an almost 20% bump in valuation, after adjusting for growth/profitability. Here's what the valuation chart from my last post looks like, colorized with vertical focused companies in red:
Why might investors prefer vertical software companies to their horizontal brethren? In general, public software companies have one of two debates surrounding them:
1) Concerns around competitive threats
2) Concerns about the size of the addressable market
Generally speaking, horizontal software companies are enmeshed in debate #1. The bull-bear tug of war around Workday, for example, revolves around the impact of Oracle/SAPs efforts to harden up their competing products in enterprise HCM/ERP in time to fend off competition from Workday. On the other hand, no one doubts that Workday is competing for a huge addressable market.
Vertical software companies typically have much less direct competition, but more limited addressable markets. Investors seem to prefer this because, while it means growth may slow at some point, most of the truly rough single-stock blow ups in software have come as a result of aggressive competitors entering the market and disrupting deals (see: DATA, JIVE, TWLO). Investors seem to prefer the long term safety of a dominant market position in a "niche" vertical to the open ended potential of an unlimited but competitive market.
Do you have other ideas for variables you'd like to see tried out in the regression? Leave a comment and I'll see what I can do!