|
Investment Dealers’ Digest, Oct. 23, 2000
Cover Story
New Economy, Bad Math
Street analysts take a deserved rap
for their e-commerce valuations
By Avital Louria Hahn
The rise and fall of Amazon.com Inc., the bellwether
e-commerce stock that has lost about 80% of its value over the past year,
has sent Wall Street analysts into a tizzy. When the company’s shares
plunged 50% this summer, analysts turned on Amazon CEO Jeff Bezos, the
former Wall Streeter they used to idolize. Now, they complain, Amazon is
providing ambiguous information about its future revenues, refusing to
control its costs and overstating the value of its investments.
Amazon’s smoke and mirrors never much bothered analysts
before. Indeed, without any solid facts, Wall Street analysts could spin
just about any kind of story they wanted on the company’s prospects. As long
as investors kept buying the tale, no problem. But the shift of sentiment
has been swift and brutal, and now Wall Street analysts and the e-commerce
companies they follow are having a “credibility problem,” according to Gary
Lutin, who heads the eponymous investment firm, Lutin & Co. He also has
organized an analyst group grappling with the problem of Internet valuation.
“The practices that have been observed in some cases are
sufficiently questionable,” says Lutin. Analysts shoulder a good portion of
the blame. Should an analyst have an inaccurate interpretation of a
company’s data, analysts are now wondering, is it the company’s
responsibility to correct him? But, asks Lutin, do analysts even call back
the company to ascertain what exactly the numbers mean? When Bezos told
analysts in a September conference that online sales would grow 50% per year
over the next 10 years, some took that to mean Amazon’s online sales would
grow by that much. Others said that Amazon did not say what percent of that
growth it would capture.
Amazon, and the analysts who’ve pumped up its stock, may be
taking the heat now. But they can’t be blamed for getting the ball rolling.
That ignominious award goes to Internet pioneer Netscape Communications
Inc., an unprofitable company that was awarded an unheard of $2.2 billion
market cap at the end of the first day its shares traded in the public
markets: August 9, 1995.
Netscape’s initial public offering, which was posting a $4.3
million half-year loss when it went public, forced Wall Street analysts to
change their entire world view. The stock had opened at $75 after being
priced at $28, which itself was more than double the price that underwriters
Morgan Stanley & Co. and Hambrecht & Quist Inc. suggested. When it finally
closed the day at $58.25, analysts were baffled. If they followed their
normal way of analyzing stock prices—looking at the price, dividing it by
the earnings and coming up with a price-to-earnings ratio—Netscape seemed
outrageously overvalued. If Internet stocks were too expensive, that meant
analysts have to they tell investors not to buy the hottest stock around.
Just as important, it meant that no more Internet companies
would find their way to the IPO market—a prospect that investment bankers
surely were determined to avoid. The alternative was for analysts to totally
recast their analytical tools, cross their fingers and hope for the best.
Out of that process came such new terms as forward multiples—a way of
justifying high stock prices based on expected future growth—and a revamping
of the discounted cash flow model, which projects future revenues.
“Starting a couple of years ago, you just couldn’t justify
stock prices using P/E valuation,” says Kris Tuttle, managing director of
research with Wit SoundView Inc. “Yet you wouldn’t want to be out of the
market during those years, which made people look for alternative ways to
justify valuations.”
But you also wouldn’t want to be in Internet stocks this
year. Hundreds of billions of dollars of wealth in Internet stocks have
disappeared since the Internet bubble finally burst this spring. The
greatest carnage was in the heavily hyped e-commerce sector, where growth,
market share and brand name became the more important indicators of a
company’s prospects than that old-fashioned notion of actually making a
profit. Many companies in that sector, if not bankrupt, are off 90% of their
highs. The Nasdaq, where most of the Internet stocks are traded, has lost
about 40% since its high on March 10.
Much of the e-commerce meltdown actually began in 1999,
before this year’s full-scale Internet crash. And even in the bullish first
quarter of 2000, e-commerce stocks like 1-800-Flowers.com Inc. and
barnesandnoble.com Inc. had been sliding for months. But when the sector’s
darlings, such as Amazon and Priceline.com Inc., finally succumbed to the
rule of gravity, the Wall Street analyst community took note.
Hypervaluations are nothing new, of course. Historically
they’ve existed for the first few years of any young, emerging industry,
explains Michael Parekh, managing director and Internet analyst for Goldman
Sachs, e-commerce’s most important underwriter. He points out that a similar
phenomenon occurred in the biotech boom of the 1980s. But he concedes that
the earlier boom was not of the same magnitude as what occurred in the
Internet mania, during which stocks soared several hundred percentage points
on the first day of an IPO, even without any earnings to show for it.
Then there are the cases that make the history books: the
Holland tulipmania of the 1600s or the South Sea Bubble. Most analysts don’t
like to make such a comparison, but given the greater size of today’s
capital markets, the total dollars lost in those crashes were probably puny
compared with today’s Internet rout.
Now the Internet’s early, outrageous days of growth are
over. Companies are maturing, and Darwinian selection is weeding out dozens
of e-commerce, mostly business-to-consumer, companies. Bankruptcies of dot-coms
are now as commonplace as IPOs once were. As a result, more and more
analysts, including Tuttle, are returning to the basics of analyzing
companies’ fundamentals. Instead of looking at what might happen, they look
at what has happened, with some sober adjustments for emerging companies.
As the market looks back at the frenzied hype of recent
times, it is clear that some analysts were more than just optimistic. “The
meteoric valuations of business-to-consumer stocks in the Internet’s heyday
were what drove everyone off the beaten track of traditional P/E valuations”
is how Wit’s Tuttle captures the detour that just about everyone—bankers,
analysts, investors—had taken from the somewhat sane company valuations
employed four or five years ago. What emerged was a system that allowed one
to justify almost anything.
“There is serious problem with the way analysts are working
today,” says Eric van der Porten, a hedge-fund manager at Leeward
Investments LLC. To point out how easy it is to poke holes in much of what
still passes for analysis, he picks apart a June research report on Amazon
by Donaldson, Lufkin & Jenrette analyst Jamie Kiggen. Just before Amazon’s
stock took its June dive, Kiggen made Amazon one of his top picks, with a
price target of $140, a value he derived by assuming that each Amazon
customer would spend $2,400 at the retailer. Even if that were true, says
van der Porten, “it would take 30 years before [Amazon begins] to see a
return on its investment.” Kiggen could not be reached for comment.
Noticeably, DLJ was one of the managers of Amazon’s
controversial $600 million European convertible bond issued in February—its
last visit to the capital markets. Van der Porten blames the faulty analysis
on Street analysts’ focus on corporate finance revenues instead of their
historic role of serving investors. “Far too many are beholden to investment
banking interests, and too much compensation is driven by investment
banking,” he says.
That’s a widely acknowledged problem, which the Internet
only worsened. Wall Street in recent years scrambled to win the mandates for
Internet stocks, with top-tier investment banks Goldman Sachs and Morgan
Stanley Dean Witter leading the way. As more speculative companies could tap
the capital markets for their funding, the rules quickly shifted. “We ended
up with the criteria of venture capitalists as opposed to the criteria of
mature industries,” explains Goldman’s Parekh. “What we are trying to assess
is the opportunities around these new markets, the value propositions of
these new products and services, the business models of these companies, and
most important, the management teams, because that’s really most of what you
have at the early stages of those industries.” (Morgan Stanley analysts
declined to be interviewed for this article.)
The Internet certainly brought with it a new set of
problems, including “the fear of instant global competition,” explains
Parekh. While the previous generation of technology companies, such as
Oracle Corp. or Cisco Systems Inc., typically had three or four years to
develop a product, build the U.S. market and build profitability, then use
the revenues for expansion, the Internet compressed the time frame, he
explains. “The minute your site is up there and accessible to the world, you
have to scale your business in every level to a much larger market from day
one,” he adds. Knowing how brutal the environment would be, analysts might
have been more cautious, instead of so giddy, about the future.
Method in the madness
That isn’t how it worked. To understand just how wild things
got, it’s perhaps best to start back with the Old Economy, old-fashioned
method of determining whether a stock’s price is too high, or too low: the
P/E ratio. Typically forward-looking, price-to-earning ratios usually are
based on next year’s earnings forecast. Microsoft Corp., for example, is
expected to earn $1.90 per share in 2001. Given the stock price of $61 last
Thursday, its P/E ratio of 32 remains high by historical standards. And
that’s after the stock has dropped by about 50% from its 52-week high.
The historical average P/E for the Standard & Poor’s 500
over the past 50 years is 15. The S&P 500 is trading at a P/E of about 30
today. Even with its 40% downturn, the P/E ratio for Nasdaq is still about
100.
Indeed, profitable tech companies like Microsoft also became
overvalued due to the new math designed to sell IPOs of unprofitable
Internet companies. With Internet stocks profitless, growth, not earnings,
became the only way stock price of any tech company was judged. “Wall Street
switched to revenue multiples because everyone was losing money and there
weren’t any earnings to value off of,” says Laurence Goldberg, a senior
e-commerce investment banker with Credit Suisse First Boston. “People used
revenue multiples as a sort of a proxy for very high growth, emerging
companies that don’t have earnings yet.”
One widely used methodology in recent years is known as the
relative price-to-sales calculation. That method takes the company’s price
per share, multiplies it by the number of shares outstanding, and divides it
by projected revenues, not earnings. But instead of looking one year ahead,
as was typically the case in a P/E analysis, analysts looked five to 10
years out, given the fact that the Internet—particularly e-commerce—was just
getting off the ground. According to that method, Netscape traded at about
68 times price to sales, or a P/E of 343 after a hypothetical “E” had been
plugged in, when it went public. Many that followed in its wake priced even
higher. Soon after it went public in 1998, eBay Inc. had a P/E of 1,000, and
today, at $53 per share, it is trading at 620 times earnings and has a 46
price-to-sales ratio. (eBay is one of the few profitable Internet
companies.)
Now analysts say that methodology was flawed. That’s because
the price-to-sales formula was too easily manipulated. First of all, the
projections of revenue growth were, by and large, wild guesses. Such
intangibles as brand names or the much-ballyhooed first-mover advantage were
also ascribed a value, with little evidence of their worth. And there was a
feeding frenzy: Once growth numbers for a leading e-tailer company, such as
Amazon, were accepted, other companies’ market values were derived from
those numbers.
If, for example, a company traded at 10 times its
price-to-sales ratio, and a similar Internet company traded at 15 times, an
analyst could with no qualms raise his price target for the lower-valued
stock, saying it should rise to the level of its peers. Of course, that’s
precisely why the whole house of cards tumbled. Once one company’s stock
collapsed, so did those of its peers.
Another commonly used Internet valuation methodology was an
adaptation of the traditional discounted cash flow (DCF) method, which has
also been used by Wall Street for traditional, profitable companies. DCF was
recast to ascertain the value of certain Internet companies, including
Amazon. This methodology projects the future cash flow over several years
and discounts it back at about 5% or 10% to reflect factors like interest
costs.
Some of the more cautious analysts, like Jonathan Cohen,
formerly an Internet analyst at Merrill Lynch & Co. and now head of European
M&A at Wit, have used this analysis. But as with the other methodologies,
there was still a lot of guesswork involved. For example, in 1998 Cohen had
a $50 call on Amazon, based on this analysis, while Henry Blodget, then an
analyst at CIBC Oppenheimer, used the same approach to come up with a $400
price target for Amazon, which it met within months. (Cohen left Merrill,
and was replaced by Blodget.)
How could the two men have had such vastly different
projections? According to Blodget, it all boiled down to whether you
thought, as he did, that Amazon was really an Internet company, or whether
it should be compared with a traditional retailer, as did Cohen. While Cohen
looked to Wal-Mart Stores Inc. to figure out a valuation for Amazon, Blodget
found parallels with Yahoo! and eBay, companies that grew faster than
retailers because of the Internet. While Blodget, who is still bullish on
Amazon, correctly predicted the $400 high, it is Cohen who looks plenty
smart today: Amazon is trading at about half the $50 call he made.
But analysts got caught up in Internet mania early in
Amazon’s life. “We were trying to figure out what percentage of all retail
sales could go online, what percentage of that market can Amazon take,”
explains Goldman’s Parekh, who also used the DCF methodology. “And if they
have that percent of the market, what would be the margins from that over a
five- to ten-year period? If those are the margins, what would be the
implied operating income and net income, and if that’s the net income over a
five- or ten-year period, we discount it back at different rates, what does
this imply for enterprise value?” he asks.
There’s an inherent limitation to this method, he explains:
“You can put in assumptions to make either a bullish or a bearish case
because you are projecting financial sensitivities out five or ten years.”
Even if two analysts were using the exact same model, a small difference in
their assumption can make a huge difference in the projected stock price,
adds Blodget.
A closer look
After this spring’s Nasdaq correction, analysts’ calls were
examined more closely. “It was the catalyst that forced people to
re-evaluate the practices that had been used for justifying ratings of stock
prices,” says Wit’s Tuttle.
Amazon survived the correction, but finally, one analyst
broke the spell. On June 22 Lehman Brothers convertible analyst Ravi Suria
rattled the Street with his outspoken assessment of Amazon, which he said
had “the financial characteristics that have driven innumerable retailers to
disaster throughout history.” Suria, who judged Amazon as a retailer and not
as an Internet company, did not see good prospects for the company’s
survival and noted that the company’s debt-to-asset ratio had deteriorated.
The company’s stock lost close to 30% in one day.
By then most e-commerce stocks had been bludgeoned in the
market. But it wasn’t due to a slew of sell reports from Wall Street
analysts. Suria’s report was an anomaly in that regard. Its timing—so
crucial on Wall Street—was impeccable because it hit the market when
investors were ready for such sober news. Indeed, a similar report from
rating agency Moody’s Investors Service was virtually ignored a year
earlier. Moody’s had placed a Caa1 rating on the company. “Caa1 means that
if you had to liquidate the company today, it is not anticipated based on
the current outlook that all the debt obligations would be covered,”
explains Russ Solomon, a senior analyst with Moody’s.
“[Amazon] has a lot of intangible assets like customer base
and information about customers,” noted another Moody’s senior analyst,
Marie Menendez. “Those are good things to have. But in terms of risk,
looking at the fundamentals and the fact that it’s a very early-stage
company, and [looking] at the fact they have chosen to finance themselves
through debt rather than through equity—all of that makes the company, from
a credit perspective, highly risky.”
Amazon’s debt load has only increased since that report was
written. And after Suria’s now-famous call, the rest of Wall Street’s
analyst community finally followed suit. A string of analyst downgrades,
including Lehman’s own Internet equity analyst, Holly Becker, followed.
Although the analyst community is starting to shift its
perspective, there is still plenty of speculative analysis on Internet
companies. On Aug. 22, for example, Salomon Smith Barney analyst Tim
Albright kept a buy recommendation on Priceline.com Inc. with a $130 price
target. The stock closed that day just under $25. Albright lowered his
rating a month later.
Prospects for Priceline, which were already iffy, have since
gotten worse. The company, which gained notoriety (and Wall Street support)
by offering the unique Internet service of name-your-own price discounts on
groceries, gas and airline tickets, was unable to get suppliers to provide
the requested discounts. Instead, it was subsidizing the discounts for
groceries and gasoline, sold separately under a Priceline unit, WebHouseClub
Inc. from a $373 million war chest it raised for that unit. Priceline was
forced to close WebHouseClub down on Oct. 5, by which time it had used much
of the cash. Priceline’s stock sank to $5, and a class action suit followed.
The suit, filed on behalf of investors who bought the stock between July 24
and Sept. 26, charged that the company disseminated misleading information
as to when the company would become profitable.
It may take such actions for change to occur. What makes the
Priceline and Amazon situations so troublesome is that these two were long
considered to the have winning models on the Internet, where brand and
market share are so highly regarded.
As the Internet correction continues, Goldman’s Parekh sees
it as an end of the first stage of a business cycle, signaling a return to
more sane valuations. Indeed, more analysts are now abandoning the
doctored-up business models that allowed one to justify almost anything, to
more traditional models based on the classic price-to-earnings ratio
.
“People finally realize that they want order, that the
markets don’t work without it,” notes Lutin.
Many Wall Street analysts nonetheless complain that the
sell-off is overdone. Amazon-bull Blodget is one of them. “It’s important to
step a good way back and look at the fact that five years ago there was one
Internet company and that was AOL, and it was worth a billion dollars, and
now five years later there are about 400 companies with a collective value
of $750 billion, which is an extraordinary value-creation period, and that
is after the brutal shakeout we have had,” he says. Given the fact that the
Nasdaq is down 40% from its highs, and many Internet stocks have sunk to
tiny fractions of their former worth, that means hundreds of billions of
dollars in market cap has been lost.
And with it, a lot of good will. “Everyone went from looking
at the glass from being totally full to being totally empty,” says CSFB’s
Goldberg. “Investors only want to talk about stories where you can show a
model where the company is going to be profitable within a year,” he says.
“And they need to get walked through exactly how this is going to happen,
because they don’t believe it anymore.”
Meanwhile, Moody’s Menendez, who has proven to be something
of a seer in the e-commerce field, recently published a report on prospects
for online retail sales in the coming months—that important pre-Christmas
rush. Menendez believes that this year, “consumers are likely to shop a name
they know,” and that will most likely mean brick-and-mortar stores or even
catalogs.
On top of that, she throws cold water on the conventional
wisdom regarding the supremacy of e-commerce. “E-tailers were initially
founded on the theory that a virtual storefront would be much less expensive
than a chain of brick-and-mortar stores,” notes Menendez. “The prodigious
amounts of marketing expenses by e-tailers have proven otherwise. Relative
to gross margins and sales, the 1999 advertising expenses of many pure-play
e-tailers was actually greater than the combined advertising, marketing,
catalog and rent expenses of traditional retailers last year.”
|