Donaldson, Lufkin & Jenrette
Jamie Kiggen (jkiggen@dlj.com) 212.892.8985
Tim Albright (talbright@dlj.com) 212.892.6801
Hilary Frisch (hfrisch@dlj.com) 212.892.4374
Sender: jkiggen@dlj.com
The Internet Observer, 10.19.98
DLJ Internet Research
River Of Light
The art of valuation isn't hard to master, even though, when it comes to
Internet stocks, many view the effort as a desperate practice, the
calculations as so many leaky boats on an ocean of wishes. How many brows
furrow these days at the idea of more upside to these stocks? Lots, let us
tell you. But many of those same investors can't just walk (or swim) away,
transfixed as they are by a vague sense that there is something big going on
out there. Plus, the true believers are bubbling up in such great numbers
that they create an ever-strengthening undertow of intellectual opposition.
So in an effort to save as many as possible from the oozy deep, here's
another plank on the valuation raft we've been building for a couple of
years. And we promise to stop with the water imagery while we're still one
metaphor away from sounding like Louis Rukeyser.
As always, the debate is most intense around the consumer Internet stocks,
and the intensity often obscures some basic principals. For instance, it's
important to remember that the underlying value driver for any consumer
marketing company is its ability to attract, retain and profitably service a
set of customers. This notion is every bit as true for Internet companies
like AOL, Yahoo!, and Amazon as it is for mainstream consumer marketing
companies like American Express, Disney and Wal-Mart. We've been discussing
this specific concept for well over a year now (see our Internet Observer,
"What Do You Value?", 9/25/97), but, since the behavior of online consumers
has evolved pretty rapidly over the last twelve months, it may be worthwhile
to revisit our analysis of the lifetime value of a customer. And for that we
turn our focus to the most customer-centric company we know: Amazon.com.
Inevitably, the value of any given customer is a function of the profit
generated by said customer. In one former life we spent time as a direct
marketer, buried in the minutiae of response rates, customer life span,
revenue-per-customer, and, the Scylla and Charybdis of direct marketing,
fixed solicitation costs and conversion costs. Needless to say, we prefer
our present happy occupation, but we're grateful for the understanding of
customer valuation that life as a direct marketer provided. Our most simple
lesson is that there are three primary inputs into the calculation of
customer value: the customer acquisition costs, the revenue generated over
the life span of the customer, and the steady-state margin per customer.
This straightforward triumvirate naturally aligns with a classic discounted
cash flow approach to help assess what each customer is ultimately worth.
But the devil, of course, is in the details. So let's wrestle with the
devil...
We won't frighten (or numb) you here with a long narrative about our
underlying assumptions, since we have a spreadsheet for that (which you can
receive by bravely requesting it). And since one of our rules-to-live-by is
"avoid all religious wars", believe us when we say that our aim is not to be
missionaries about this or any valuation methodology. But a certain science
has grown up around the process of thinking about value, and we're big
believers in the basic logic behind that science. So a skeletal description
of the methodology should be helpful here.
First, we create a two-year simplified income statement for a single
customer, with year one bearing all of the acquisition costs and year two
showing a steady-state profitability level. After establishing a
revenue-per-customer figure (1998 revenue of $500 million divided by just
over 3 million customers, or $168), we then move down each expense line
item. Plug in Amazon's 1998 gross margin (22.6%), which yields a
contribution per customer of $38. Then wander into the weeds of sales and
marketing expenses, which we divide into customer acquisition costs (62% of
the $122 million line item, and allocated only to new accounts) and what
we're calling "non-discretionary" marketing expenses (which get spread
across all accounts). Finally, down in the depths where no light shines, we
divide a portion of R&D and G&A expenses (67% of such, to be precise) among
new accounts. And what does all this (and a lot more pencil-sharpening
behind the scenes) leave us with? A headache, as well as a year one
loss-per-customer of ($7) and a year two earnings-per-customer of $16.50
(since year two is absent the customer acquisition costs).
Now the fun really begins. Take our steady-state (i.e. year two) customer
earnings profile and apply a dividend discount model (stay with us here),
defined as D / (R - G), where D is the operating profit from the customer, R
is our cost of capital or discount rate, and G is our terminal growth rate.
Having emerged from the keeper-hole of the preceding paragraph with $16.50
of operating profit from each customer, we calculate a weighted average cost
of capital of 13.9% (see aforementioned spreadsheet for applicable capital
structure) and a terminal growth rate of 12.5%. Now, we can hear some
accusatory cries coming from the cognoscenti regarding that last number, but
remember that we're talking about the fastest growing company in our
universe (if not the real universe), and we need to allow for the assumption
of a steady-state customer model just one year down the road.
So, you should now hit the present value button on your calculator, and a
number appears that represents what we believe each customer is currently
worth: $1,125. But since our ultimate goal is to establish a target price
for the stock a year from now, we need to take one last step: estimate the
number of customers that could be acquired over the next twelve months, and
multiply the customer value by that total account number to arrive at an
enterprise value. Even though Amazon will probably end 1998 having achieved
a year-over-year account growth rate of around 2000%, we won't extrapolate
from there. Instead, we'll ratchet it down to 60%, arriving at a year-end
1999 customer base of 7.25 million. The math then gets us to a target price
of $170 per share, comfortably near the $175 target we've been carrying
based on a discounted cash flow equation for the whole company.
But don't run off yet. There is a postscript to this discussion of valuing
Amazon's customer base, one that may well be more important than the
valuation process. Its called return on invested capital. Amazon's ROI is
assumed by many to be inherently negative. After all (the common wisdom
says), a company expecting to generate more than $26 million in operating
losses in this quarter alone must be struggling with an unfavorable business
model, one that will prevent them from ever getting ahead of the customer
acquisition cost curve. Not so fast. Our analysis shows that the customer
Amazon acquires today will generate positive cash flows over the life of
that customer with a net present value of $1125. So the $24 cost to acquire
a customer today looks downright cheap, considering Amazon achieves
breakeven in a little over two years and achieves a 4700% return on
investment over the long-term.
Of course, analyzing Amazon's real lifetime ROI per customer is more complex
than our little exercise suggests, given the leaps of faith one needs to
make to arrive at the average effective lifetime of the customer and other
nuances like that. But in our view it's beyond debate that Amazon operates
an increasing returns business: the more customers Amazon gets, the more it
can spread its fixed costs over the customer base, thereby freeing up funds
to acquire new customers. And a widening lead in total number of customers,
as Amazon has been achieving, provides the company with an inherent
advantage as it moves into new markets like music. Effectively adding these
new markets on top of Amazon's asset base accelerates asset turnover, and
asset turnover is one of the two components of the return-on-equity equation
(the other being return-on-assets). Ultimately, anticipated return-on-equity
is the primary reason to own any stock. While we'll have more to say on this
in the near future, for now we'll leave you with this cryptic truth:
Amazon's ROE begins with the value it derives from its customers. Stay
tuned.
In the meantime, we hope this discussion of customer value has been
illuminating. As always, the human aspects of any business are more
important than calculations or models, since execution is the ultimate
differentiator. And we expect Amazon's management team to continue executing
in Churchillian fashion, bringing us from the tossing sea of cause and
theory to the firm ground of result and fact.
=================================================
The DLJ Internet Observer, a biweekly research product of the DLJ Internet
Research team, is distributed through email, First Call and fax. To be
included on the distribution list simply send an email message to
eqinfomail@dlj.com with the phrase "subscribe observer" in the body of the
text, or contact your DLJ salesperson. To remove yourself from the
subscriber list, send email to eqinfomail@dlj.com with the phrase
"unsubscribe observer" in the body of the message.
Donaldson, Lufkin & Jenrette Securities Corporation, 1998. Additional
information is available upon request.
THIS REPORT HAS BEEN PREPARED FROM ORIGINAL SOURCES AND DATA WE BELIEVE TO
BE RELIABLE BUT WE MAKE NO REPRESENTATION AS TO ITS ACCURACY OR
COMPLETENESS. THIS REPORT IS PUBLISHED SOLELY FOR INFORMATION PURPOSES AND
IS NOT TO BE CONSTRUED EITHER AS AN OFFER TO SELL OR THE SOLICITATION OF AN
OFFER TO BUY ANY SECURITY OR THE PROVISION OF OR AN OFFER TO PROVIDE
INVESTMENT SERVICES IN ANY STATE WHERE SUCH AN OFFER, SOLICITATION OR
PROVISION WOULD BE ILLEGAL.DONALDSON, LUFKIN & JENRETTE SECURITIES
CORPORATION (DLJSC), ITS AFFILIATES AND SUBSIDIARIES AND/OR THEIR OFFICERS
AND EMPLOYEES MAY FROM TIME TO TIME ACQUIRE, HOLD OR SELL A POSITION IN THE
SECURITIES MENTIONED HEREIN. UPON REQUEST, DLJSC WILL BE PLEASED TO DISCLOSE
SPECIFIC INFORMATION ON SUCH POSITIONS OR TRANSACTIONS. DLJSC OR AN
AFFILIATE MAY ACT AS A PRINCIPAL FOR ITS OWN ACCOUNT OR AS AN AGENT FOR BOTH
THE BUYER AND THE SELLER IN CONNECTION WITH THE PURCHASE OR SALE OF ANY
SECURITY DISCUSSED IN THIS REPORT. OPINIONS EXPRESSED HEREIN MAY DIFFER FROM
THE OPINIONS EXPRESSED BY OTHER DIVISIONS OF DLJSC. DISCLAIMER FOR
INSTITUTIONAL CLIENTS OF THE EUROPEAN ECONOMIC AREA (EEA): THIS DOCUMENT
(AND ANY ATTACHMENTS OR EXHIBITS HERETO) IS INTENDED ONLY FOR EEA
INSTITUTIONAL CLIENTS. THIS DOCUMENT MAY NOT BE ISSUED OR PASSED ON TO ANY
PERSON IN THE EEA EXCEPT(A) A PERSON TO WHOM IT MAY LAWFULLY BE ISSUED, OR
(B) IN THE U.K., A PERSON WHO IS OF A KIND DESCRIBED IN ARTICLE 11 (3) OF
THE FINANCIAL SERVICES ACT 1986 (INVESTMENT ADVERTISEMENTS) (EXEMPTIONS)
ORDER 1995, AS AMENDED. |
