Report: Public Tech Company Revenue and Earnings Multiples 2015-2019

How do investors in public companies value performance? We analyzed 60 public tech companies' revenue multiples and earning multiples from 2015 to 2019 to find out.

As 2019 rolls on and the wave of tech company IPOs continues, it’s striking how much of the value in the public markets has increasingly accrued to software and tech companies.

For the four years ending June 30, 2019, the NASDAQ composite (which is heavily weighted towards information technology companies) has increased from ~5,000 to ~8,000, or about 12.5% per year. (Compare this to S&P 500 index’s ~8.6% annual growth for the same period: roughly 50% higher return.)

Because private company transactions (and valuations) are driven in part by the activity in the public markets, we keep a close eye on public market trading. But valuation for private companies is nuanced and not always driven by current revenue and earnings, so to us, value-to-revenue and value-to-profit ratios are important signals worth tracking.

$1 invested in each company in our basket of tech companies ("$1/each") or a $1 invested based on the relative size of each company in our basket of tech companies (defined later) would have materially outperformed both the NASDAQ and the S&P 500 over the past 4 years.

What exactly do we mean by “value?” The value of anything is the risk-adjusted, time-adjusted, value of the cash flows it provides (and the process of determining value this way is called the Discounted Cash Flow method).

Since risk and future cash flows are uncertain, an investor may want a discount on future payments in exchange for tolerating that uncertainty, because who knows how the market will feel about products and services in the future?

This is where multiples come in. Multiples compare current performance to the current value of the enterprise—for example, a company with risk-adjusted, time-adjusted cash flows that sum up to $2B (i.e. a company that has a $2B value) and produces $1B in annual sales would have a 2x EV/sales multiple—and they are, effectively, a shorthand way of comparing the estimated present value of future cash flows (i.e. value) to current performance.

But fundamentally, they’re still shorthand for “what positive cash flow(s) will I receive, when, and with what certainty.” Given that there is so much uncertainty about the future, we don’t explicitly think in terms of “risk-adjusted, time-adjusted value of the cash flows,” but that perspective is there all along in the logic of analyzing multiples.

We wanted to take a look at the current and recent investor thinking with regard to publicly traded tech companies. The full report will be available soon (and you can sign up to receive it here) but we’ll be sharing some high-level results over the next few weeks. Let’s get started.

Model your investments‘ exit and fundraising scenarios in minutes. Sign up for early Waterfall Modeling access:

Analysis of Public Tech Company Valuation Multiples, 2015-2019

To do a deep dive, we constructed a basket of ~60 U.S.-headquartered tech companies that went public prior to June 2015 (sorry, Uber). Let’s explore this “Tech Basket” and ask the following two questions:

  1. How has the public market been valuing each dollar generated and earned by the Tech Basket, as measured by Enterprise Value to Sales (“EV/Sales”) and Enterprise Value to EBITDA (“EV/EBITDA”)?
  2. When comparing clusters within the Tech Basket (Ecommerce, Digital Media / Social, etc.), are there are any notable trends?

A few caveats about the population:

  1. As companies operate in the intersection of many industries and disciplines, the single-variable industry classification loses its explanation and model-fit ability, so these clusters won’t be fully representative of every company in their industries. You can see the individual companies that were selected in the image here, with additional analysis below. In the extended version of this report, we analyze a wider set of comparables within each industry (sign up to receive a copy here). For the purposes of this report, we will refer to the grouping of these companies as “clusters.”
  2. Given the size of these clusters, some of the analysis may yield limited deductions.
  3. Comparing companies across clusters (given the low number of observations) should be done with some caution.

The Value of the Tech Basket

Now let’s take a snapshot of each cluster of companies within the Tech Basket, as of June 30, 2019. We ranked our clusters from largest to smallest according to::

  • Median Enterprise Value (“EV”) in billions
  • Median Enterprise Value to Sales (“EV/Sales”)
  • Median Enterprise Value to EBITDA (“EV/EBITDA”)

As analysts, we already know that certain clusters in the Tech Basket tend to have “winner-takes-most” characteristics, with a few companies significantly outperforming the others in their clusters. These companies’ presence in the Digital Media / Social cluster (Netflix, Google, Facebook) and Software Solutions cluster (Salesforce, Oracle, Microsoft, Adobe) gives them much higher median EVs than the others. For the same reason, they also enjoy some of the highest EV / Sales ratios. Fintech as a cluster typically presents higher-than-usual natural barriers to entry (technical complexity, regulation, and subject-matter-expertise). Due to this fact, the companies in the Fintech cluster enjoyed higher EV / Sales ratio as of June 2019.

Given the highly competitive nature of these clusters (and the Tech Basket in general), companies spend and invest heavily in growth, which means higher expenses and lower EBITDA—so analyzing EV / EBITDA alone (without more context) may produce the incorrect conclusion (because low EBITDA margins would result in otherwise higher EV / EBITDA ratio).

In the median EV / EBITDA table above, we can see the median EV / EBITDA of Marketplaces, Gaming, and Ecommerce rank second through fourth. However, this is mainly because their EBITDA margins are on the lower end of the spectrum in the overall Tech Basket. When we use the Tech Basket average EBITDA margin (instead of cluster-specific EBITDA margin), Digital Media / Social, Software Solutions, and Fintech take the top three positions of EV / EBITDA instead.

From the resulting spread, we can see that investors are comfortable with much higher valuations relative to EBITDA in clusters that expect to invest in growth instead of delivering earnings in the short term. The premium they pay for such companies takes into account the cost of delivering higher than expected earnings in the future.

Median as Representation

Thus far, we have referenced the median data point for our analysis. However, it’s important to review the data’s distribution to get a better sense of the range and variability of the sample data.

The standard deviation (a measure used to quantify the amount of variation or dispersion of a set of data values) of each cluster’s EV / Sales helps illustrate, for instance, that the five Gaming companies are all close to 4.5x EV / Sales, with a low 0.8 standard deviation (or “st dev” for short), whereas the eight Cybersecurity companies have wildly different EV / Sales multiples (as shown with the 6.2x EV / Sales st dev). In other words, investors in Gaming companies are treating their future earnings potential a lot more regularly (or predictably) than investors in Cybersecurity.

Since markets and the risk tolerance of investors change over time for any given cluster, we wanted to explore how the distribution of valuation changes over the analysis period. To do so, we selected the EV / Sales 1st quartile through the 3rd quartile (“EV / Sales Band”) along with the median of each cluster between June 2015 and June 2019.

EV / Sales is simple and quick, but may be misleading:

We often look at a basket of comparable companies to assess how their EV / Sales ratio may be applicable to a subject/target company’s sales (and thus its estimated EV). It’s important to note that, although median and ranges provide for great macro views on the industry and as a frame for the analysis, a blindly-selected multiple shouldn’t be used as the basis of analysis.

Ecommerce, Electronic Equipment, and Cybersecurity have shown the highest stability over the past 4 years, where the EV / Sales bands have remained fairly flat.

Other industries like Travel are down, while Fintech and Software Solutions are up.

Digital Media / Social and Gaming are ultimately up during the analysis period (despite the volatile path over the past few years).

Let's do a deep dive into the EV / Sales band changes within the Marketplaces and Software Solutions cluster over the past 4 years to assess how the public markets have been valuing each dollar generated by these two clusters. Are the bands getting tighter or wider? Are the companies moving together, or away from each other?

Marketplaces: (Match Angi, Etsy and Grubhub) each moved up from a 3x-6x EV / Sales range to a 6x-10x EV / Sales range around Dec 2017-Jun 2018, whereas other companies in the space (Ebay, Shutterstock, Truecar, Care, and Yelp) all remained around 2x-4x EV / Sales. Selecting the right multiple would be critical as a 2x vs 10x EV / Sales multiples would result in an astronomical difference.

Software Solutions: A handful of high EV / Sales companies (Adobe, Salesforce, NewRelic) are at ~10x, while the remainder of the companies in the basket (Oracle, ADP, Citrix) remain lowly valued around 3-5x. Box and Microsoft are valued between the high and low, hovering between 5x-8x over the last 2 years.

Take a private company operating in the Marketplace or Software Solutions cluster with $1M in trailing sales. Selecting a 2x EV / Sales multiple or an 8x EV / Sales multiple both seem reasonable (given the aforementioned bands). But without considering the size, product portfolio, revenue and earning repeatability, unit economic study, growth channels, etc, a buyer could overpay by $6M, or a seller could undersell by $6M. That's why it’s critical to go beyond multiples (or at least understand where multiples are coming from) when we prepare our analysis.

How optimistic is the market about earning potential?

Almost across the board (except for Electronic Equipment and Travel), we are seeing higher EBITDA margins for the Tech Basket in 2019 than we did in 2015. Given that EBITDA is the denominator of our EV / EBITDA, all else equal, we would expect a decrease in EV / EBITDA for the Tech Basket.

But it seems like “all else is NOT equal”—except for Electronic Equipment and Travel, it seems that investors are rewarding these companies by providing them higher EV/EBITDA multiples as their EBITDA margins are improving.

For example, the Software Solutions Median EV / EBITDA has increased from 15x to 20x. In so many words, investors are saying that, without any risk or consideration for time, either:

A) Your same level of earnings will continue for 20 years (whereas it used to be 15 years), or

B) Although we had expected your earnings to increase 15x next year, we now believe that it will increase by 20x next year, and you will cease to exist afterward, or

(the most logical answer):

C) Some combination of A, B, and others. Higher EBITDA margins actually seem to lead to higher investor confidence.

Some winners, some “losers?"

Lastly, we wanted to answer the question: on an absolute basis, for each cluster, how did the EV / Sales and EV / EBITDA change for our nine clusters—are there any noticeable trends? For our starting point (and to take out “noise”), we looked at the average of June 2015 and December 2015 values (“Late 2015”), and the average of December 2018 and June 2019 as our ending point (“Early 2019”).

The Losers:

  • Travel, Cybersecurity and Electronic Equipment saw the largest decreases in their EV / Sales.

The Winners:

  • Biggest industry jump in median EV / Sales is Fintech (4.1x → 6.9x = +2.8x)
  • Followed by Gaming (2.1x → 4.6x = +2.5x) and Software Solutions (4.8x → 6.1x = +1.3x)

The caveats noted at the beginning of this analysis still hold, and what has happened in the past is not always a predictor of the future. As companies consider M&A transactions—or other activity where valuations have high sensitivity—they'll likely check out public companies' performance versus value, but they should remember that there is more nuance analysis to be considered than simple multiples.

Check back soon for more insights from our analysis, in which we’ll discuss companies’ multiples in the context of broader market performance. If you’d like to read the full report, email us at [email protected]

Model your investments‘ exit and fundraising scenarios in minutes. Sign up for early Waterfall Modeling access:

Report: Public Tech Company Revenue and Earnings Multiples 2015-2019