Barron's article part 2
Cover Story, Part 2
Cover Story, Part 1
From there, Cisco acquired more advanced Internet technology and optical
switching technology, maintaining its dominance in Internet gear. Last year's
$6.9 billion acquisition of Cerent Corp. underscored its determination to have
the latest and best data transmission equipment, even in the face of
competition from newer rivals such as Juniper Networks and Foundry
Networks.
Last year and this year, Cisco has extended its acquisition emphasis again. Its
latest targets are telephony over the Internet, data over cable TV lines,
wireless data networks and, for the first time, the specialized computer chips
that make all the routers, switches, networks and phone systems fast and
powerful. For a hint of how important acquisitions are to Cisco, consider
these numbers: In the past three fiscal years, Cisco spent $3.3 billion on
research and development internally and recorded an additional $1.5 billion in
purchased research and development.
Unfortunately for Cisco, the success bred by its acquisitions carries with it the
seeds of self-destruction. As the company bids higher and higher for its
targets, it drives up the market for all telecommunications-equipment
companies-including Cisco itself. Acquisitions come harder and higher.
Once upon a time, takeover artists looked for companies with shares trading
far below the value of total corporate assets. Today's takeover artists at
Cisco and other such companies can't do that. They are not buying assets,
they are not buying products. They are not buying profits and they are not
even buying revenues. They are buying people and half-formed technology
that the people may someday turn into products generating revenues, profits
and assets.
Cisco's takeover specialists run a risk: They must buy the right companies,
with the right people, developing the right products for a market that may not
exist for years after the deal is done. How well does Cisco do with its
acquisitions? Very well, it says. One target company that had $10 million in
revenues at the time of acquisition provided Cisco with technology that now
generates more than $1 billion of revenues. But not all acquisitions are so
successful, and for most of them, it's hard for an outsider to gauge success.
But Cisco does make it clear how high its hurdles are. The company has said
it expects eight out of 10 investments to be successful.
Until recently, Cisco accounted for most of its acquisitions through simple
purchase accounting. More recently, however, it has made several
acquisitions for prices that are astronomical even in this era of infinite
price-earnings multiples, and accounted for them by the pooling-of-interest
method. If these companies had been acquired for those prices with purchase
accounting and entries for purchased R&D, Cisco's reported earnings
probably would have vanished in 1999.
Pooling distorts earnings by failing to reduce them
with amortization of goodwill and similar
adjustments. And pooling can dilute the interest of shareholders. If Cisco
issues 100 million shares of stock (before a recent 2-for-1 split) worth about
$6.9 billion to take over Cerent, a private company with $10 million of sales,
stockholders ought to ask if the acquired outfit will be worth giving its owners
roughly 3% of Cisco. It's hard for anyone to imagine anything Cerent could
do for the company to justify the dilution inherent in Cisco's
pooling-of-interest acquisition. Even with Cisco's market cap of 39 times
revenues, Cerent ought to have $176 million of revenues to join the Cisco
family on a basis that's equitable to Cisco's existing shareholders. And if
Cisco's price should ever fall from here, the disparity would be worse.
Friday's acquisition of ArrowPoint should leave investors asking the same
question. Cisco is issuing about 90 million post-split shares, with a market
value of $5.7 billion, to acquire a company that had negative book value, has
never earned a profit and had sales running at an annual rate of $40 million.
Applying Cisco's revenue ratio, ArrowPoint ought to have sales approaching
$146 billion.
The Financial Accounting Standards Board has proposed doing away with
pooling-of-interest accounting. FASB says investors can be confused if two
equally sanctioned accounting methods produce vastly different valuations.
Last week Congress held a hearing on the proposed ban, and Cisco officials
joined executives of other high-tech companies in opposing it. They said it
would make it more difficult to do mergers that apply the capital of
established firms to the technology of new companies.
Cisco fans seem anything but confused by Cisco's reported earnings. They
have awarded Cisco a share price that's about 190 times reported fiscal 1999
earnings. Yet there are serious questions about the quality of earnings at Cisco
and other communications-equipment vendors.
In order to close their deals, they are giving generous financing packages to
their customers -- sometimes to customers whose ability to pay ought to be
more closely explored. Cisco failed to return calls on the question before
press time, but it acknowledged in a footnote to its most recent quarterly
report that it is experiencing increased demand for vendor financing, and has
taken on increased risk as a result. The company does not fully disclose the
extent of its vendor financing. A close reading of footnotes in the annual report
shows that "net investment in leases," a form of vendor financing, rose from
$190 million in fiscal 1998 to $500 million in fiscal 1999, but fell again to
$212 million at the end of the second quarter of fiscal 2000. Deferred
revenues, which include accounting for deliveries where collection is
questionable as well as other less problematic items, rose to $724 million in
fiscal 1999 from $339 million in fiscal 1998, and were not broken out in the
quarterly statement.
The company's P/E alone should give investors pause, even though analysts
strain to justify it. No established company has ever traded at such a high
multiple and failed to come a cropper in the end -- especialy not an
established company with such predictable earnings.
If Cisco sold at the multiples of its competitors, investors would be shocked.
If the market valued $1 of Cisco's earnings the way it values $1 of Nortel's
earnings, at a multiple of 100, Cisco stock would be selling for $35 a share. If
it could command Lucent's multiple of 46, Cisco's share price would be
around $16.
How much does a company have to earn over the next 10 years to warrant
a multiple of 190? By the old-fashioned one-to-one rule of thumb, matching
the growth rate with the P/E ratio, earnings would have to grow 190% a year.
Even bullish analysts do not believe that Cisco's profits will even double from
fiscal 1999's $2.5 billion, much less that they will do so every year for more
than a decade. If they did, the analysts would have to believe that Cisco's
existing businesses will produce profits of $2.5 trillion in 2010. If they believe
that, then they should consider selling every other company in their portfolios,
because a company that earns $2.5 trillion in 2010 will have taken over half
the world. Of course, a company that successful will probably sport a P/E
ratio way above a mere 190 times earnings, and it will be able to take over
half the world for stock.
Another question worth asking produces a different answer: If a hypothetical
long-term investor buying Cisco at 67 3/4 at the end of last week were
seeking what some bullish analysts expect, about 35% a year in price
appreciation, what would the stock be selling at in 2010? Answer: $1,300 a
share. At a multiple of 190 times earnings, Cisco would be making around
$6.80 a share, or about $47 billion. That's hefty, but a far cry from $2.5
trillion, the kind of earnings the P/E ratio implies.
The absurdity of such speculation points up the ultimate impossibility of
Cisco's acquisition binge: It can't go on forever. It also points up Cisco's utter
dependence on continuing an ever-increasing string of successful acquisitions. |
