Why EV/Revenue Is Dangerous
Be careful when using EV/Revenue to compare software businesses
I first learned about EV/Revenue as a first-year in college, probably at one of the Wharton finance club meetings, as a way to value companies. As a third year now in Penn, who’s also doing a term-time internship in growth equity, I’m always looking at EV/Revenue multiples. However, after reading Jared Sleeper’s article, I’ve realized that I’ve taken this multiple for granted… I’ve never really sat down and thought about it until reading his article (I mean unless you’re a software finance nerd, why would you right? lol)
Due to the fast pace nature of school or work, we learn and apply. There’s not much time for us to sit down to understand the “why” or the “how” something works. Feeling inspired by Jared’s article, I thought to use this piece as a way to do some much needed reflection on this multiple and perhaps why it could be dangerous if not used properly.
We first must need to understand why software investors use the EV/Revenue multiple. In our classes, we learn about the Discounted Cash Flow (DCF) method and P/E ratios. Both of these methods falls short when analyzing subscription based businesses.
Firstly, as we know, the DCF is riddled with assumptions. Beyond the final years of the DCF that we model, we can approximate the “terminal value” or the value of all subsequent cash flows to perpetuity using the perpetuity growth method. As we know this terminal value is based on a constant growth rate and WACC, which feels a bit silly. As a result, we introduce the exit-multiples approach. We might apply an earnings multiple to the last year projected figure. But this method doesn’t make sense for software businesses.
Why?
Conventional thinking of earnings/cash flow doesn’t necessarily apply well to software. Contradictory to what we’re taught at school, investors don’t want SaaS companies to break profitability too soon. Wait, what? I thought cash was king.
I thought the same, and there are certain truths to it (such as the shift towards profitability and growth). However, as high-level blanket statement, there are a couple reasons why, as investors, we don’t want to see profitability too early:
Scale Issue: SaaS companies tend to acquire revenue through S&M effort near the beginning to generate long-term recurring revenue streams. Without consistent S&M spend, the company won’t be able to scale
TAM Issue: It could also be a sign that the market is saturated and there’s no further incremental revenue that can be achieved
Management Issue: It could be a sign that they’re unable to efficiently deploy capital to acquire customers (we saw this this past year when sales efficiency deteriorated and as a result SaaS companies got punished for it)
Now - with that out of the way, why did we end up settling on EV/Revenue? Well, we sorta already answered it.
Most companies aren’t profitable because they shouldn’t be - and it would be silly to use like an EBITDA or FCF multiple on companies that are not profitable
It’s the “least worst” option because even if you go down to EBIT, you can wind up unprofitable - so the safest bet is to stay high up in revenue land
This tells us something right? How can we compare operational efficiency of companies if we’re only looking at revenue? This leads me to the next section where I’ll talk about the inefficiencies of EV/Revenue and why we should be cautious.
Obviously, we all understand that we compare multiples only when they’re apples-to-apples. For example, we can’t compare software multiples with manufacturing multiples. Everybody knows this.
However, even comparing EV/Revenue within software companies can be quite problematic:
Gross vs Net Revenue: When revenue is recorded as gross revenue, especially in ad tech or marketplace companies, we’ll often get a highly inflated number - thus distorting the true multiple for the company
Services Businesses: Some “software” businesses might have a significant professional services operations (which tend to have lower margins), thus operating at lower gross margins compared to pure-play software
Payment Businesses: When software companies include payment processing capabilities, they incur direct costs for each transaction. Unlike traditional software that scales well, the COGS of payment businesses scale linearly with revenue and thus resulting in lower gross margins
With that said, one of the biggest issues with EV/Revenue is accuracy. We look at how “software” is doing and we lump companies together when they probably shouldn’t be lumped together. Why do we do this then?
It all boils down to simplicity. In fact, that’s why we use multiples. It’s easy, straightforward, and simple to communicate. Rather than needing to project cash flows for the next hundred years, let’s slap on a 10x multiple and call it a day.
So, if I don’t use EV/Revenue, what multiples should I use?
Honestly, you should probably still use EV/Revenue, but be mindful when considering comps. Looking at the real world (from investment banks to growth equity shops), all the professionals use the EV/Revenue multiple. However, we should start thinking about other multiples that might be better suited for these particular companies in our models.
For fintech, maybe it’s EV/GP or EV/EBIT? For pure-play software, maybe we keep it at EV/Revenue, etc.
Generally, the more capital intensive a business is, the further you go down in the P&L/CFS as you have to adjust for different efficiencies (Opex/CapEx levels). This explains why industrials business trade on e.g. EBITDA less CapEx and why software companies trade on revenue multiples. But as I mentioned before (the 3rd time now), software is a really broad space. Fintech software, for example, often trades on GP in order to adjust for the transaction cost.
We’ll keep it here for now. Next time, I’ll talk about Growth-Adjusted Rule of 40 in valuations :)