388  Using Multiples
Spotify offers music streaming on demand. Sirius XM combines on-demand 
and radio-format music streaming (via its 2019 Pandora acquisition) with its 
original business of satellite radio broadcasting. All three generate revenues 
from user subscriptions, but Spotify and Sirius XM also generate advertising 
income. For Netflix, the cost of streamed content is largely independent of 
the number of users; it produces an increasing portion of the content itself 
and does not pay per view to its content suppliers. Spotify, in contrast, does 
not produce content and pays an amount per song-play to the content own-
ers (record companies and artists). Sirius XM develops part of its content (for 
example, its radio shows) and buys music from third parties—also paying 
per song-play, but at lower rates for radio broadcasting than for on-demand 
streaming. In addition, the companies’ user growth rates for 2019 varied 
widely: about 30 percent for Spotify, 20 percent for Netflix, and 3 percent for 
Sirius XM (with Sirius active in the United States only and Netflix and Spo-
tify in many countries). Because of these differences in user economics and 
growth, an average of the value per user of Netflix, Spotify, and Sirius XM 
is not very meaningful for valuing another online music or video-streaming 
business. You would need to analyze the underlying business model in detail 
to understand which of the three is the most comparable business.
In the end, what matters is the underlying value creation, not the number 
of users, website hits, or unique visitors. In the late-1990s Internet examples 
from Fortune, Yahoo traded at a higher multiple than Amazon.com because in-
vestors expected that Yahoo’s profit per user would be higher than Amazon’s. 
Academic studies have demonstrated for these valuations that the number of 
unique visitors to a website or the number of pages on a site viewed per visit 
was directly correlated to a company’s stock price, even after controlling for 
the company’s current financial performance.13 The power of a given nonfi-
nancial metric, however, depended on the company. For portal and content 
companies such as Yahoo, page views and unique visitors were both corre-
lated to a company’s market value. For online retailers such as Amazon.com, 
only the page views per visit were correlated with value. Evidently, the mar-
ket believed that merely stopping by would not translate to future cash flow 
for online retailers.
Research has also shown that as an industry matures, financial metrics such 
as gross profit and R&D spending become increasingly predictive, whereas 
nonfinancial data tend to lose power.14 This indicates a return to traditional 
valuation metrics for new industries as they mature and once financial metrics 
became meaningful.
13 B. Trueman, M. H. F. Wong, and X. J. Zhang, “The Eyeballs Have It: Searching for the Value in Internet 
Stocks,” Journal of Accounting Research 38 (2000): 137–162.
14 P. Jorion and E. Talmor, “Value Relevance of Financial and Non Financial Information in Emerging 
Industries: The Changing Role of Web Traffic Data” (working paper no. 021, London Business School 
Accounting Subject Area, 2001).

Summary  389
A problem with all multiples is that they are relative valuation tools. They 
measure one company’s valuation relative to another’s, normalized by some 
measure of size, be it size of earnings, revenues, or number of customers. They 
do not measure absolute valuation levels. For multiples based on operating 
metrics, there is an additional challenge of interpretation, because you can 
only compare them across a very limited number of companies that have a 
very similar operating model. Financial multiples are easier to interpret and 
compare. Take the example of an EV-to-EBITA multiple of 20 times for a ma-
ture industrial company. Basic understanding of underlying value drivers can 
readily lead you to a first conclusion: this multiple reflects high expectations 
for ROIC and growth (at a reasonable cost of capital). But it is much harder 
to come to such a conclusion when you observe an EV multiple of $1,200 per 
customer.
Summary
Of the available valuation tools, discounted cash flow continues to deliver the 
best results. However, a thoughtful comparison of selected multiples for the 
company you are valuing with multiples from a carefully selected group of 
peers merits a place in your tool kit as well. When that comparative analysis 
is careful and well reasoned, it not only serves as a useful check of your DCF 
forecasts, but also provides critical insights into what drives value in a given 
industry. The distinction between operating and nonoperating results, capital, 
and cash flows should follow the exact same logic as applied in DCF valua-
tion. The most insightful multiples are those that compare operating value to 
operating results. Operating metrics such as mineral reserve size or number 
of subscribers can be used when these are clearly related to value creation. 
In all cases, be sure that you analyze the underlying reasons that multiples 
differ from company to company, and never view multiples as a shortcut to 
valuation. Instead, approach your multiples analysis with as much care as you 
bring to your DCF analysis.