Mike, our suspicion grows - Fleck is actually Burke or vice versa. -g-
I am sure you would love this piece:
stocksite.com
------------
June 17, 1998
Q&A with Bill Fleckenstein
Bill Fleckenstein is a market contrarian and a frequent guest on CNBC. He answers your
investment questions each Wednesday in our new Q&A forum. Send them to
fleckenstein@go2net.com.
I continue to receive email from investors who insist that
technology has been beaten up over the last few months and that
we must be nearing the next great buying opportunity.
Unfortunately, these people are missing the message that I have
tried to convey over the last year. Buying on the dip worked for a
long time, that I will concede, but this strategy will only lose you
money going forward. While a 10 or even 20 percent correction
in technology might seem significant to the "new era" investor, it
is really only the beginning for many of
these troubled technology companies.
For those familiar with a piece I authored
a few months back called "In Love with
Technology" the reasons for why I believe
so strongly that technology shares are far
from reaching any sort of a bottom are
clear. For those unfamiliar with the piece,
I'd like to run it in place of this week's
Q&A. Your questions and comments, as
always, are welcome.
It's easy to see why people love
technology. Life without television, fax
machines, cellular phones, or personal
computers would be far less enjoyable,
and life without modern medicine would
certainly be shorter. In addition,
technology continues to decline in price. The combination is hard
to hate.
Conversely, the business of technology is hard to like. Why?
First, the business of technology is obsolescence. In order to
stay ahead of competitors, you cause your own products to
become obsolete. It is a run-until-you-drop type of business. As
a result, you must constantly put all of your cash, and often more,
back into the business, and yet you still cannot erect barriers to
entry. So, ultimately, all products become commodities with the
attendant small profit margins. Bill Gates illuminated this point
best in his recent Senate testimony: "The cost of computing has
decreased 10 million fold since 1971. That's the equivalent of
getting a Boeing 747 for the price of a pizza." Investors tend to
be blinded by the benefits they see as consumers of technology
and ignorant of the pitfalls they face as fractional owners of the
businesses.
Commoditization and price collapses are what allow for
mass-market penetration. Unfortunately, once we have a big
up-cycle for technology issues, such as we have just witnessed,
venture capitalists have mountains of capital and can raise even
more. In 1997 venture capitalists raised a record $11.4 billion, a
16% increase over 1996 which was a 42% gain over 1995.
Then, they steal managers away from the previous winners to
create even more competition. What this creates is an
environment like we have now of "profitless prosperity", where
product prices drop, margins collapse, consumers of technology
get a great deal and the businesses suffer cut-throat
competition. Ultimately, there are very few long-term winners.
For every Microsoft or Hewlett Packard, there are hundreds that
languish or wither to nothing. It doesn't take that long to go from
darling to dog. Look at Netscape and Digital Equipment, or for
those with longer memories, Wang Computer, Computervision,
and Prime Computer.
It is precisely because everyone can see the wondrous beauty of
technology that the world has created so much capacity, setting
the stage for the destruction of profitability.
The combination of the first post cold war economic boom, the
stock market boom and the personal computer boom created a
period for technology investors from 1992 through 1995 that is
the best we have ever seen and may be the best we will ever
see. Until Asia fell apart last summer, the economic backdrop
was close to nirvana with every corner of the world making a
dash for growth, while worshipping technology as the religion of
the new era.
Given that the world economic outlook was so powerful until
recently, it is shocking to note that so many large technology
companies had earnings disappointments in 1996 and 1997. In
fact, in the personal computer, semiconductor and related
businesses only Microsoft and Dell came through this period
without an earnings disappointment. Why? Because the peak
for these businesses was really late 1995/early 1996.
The anticipation of Windows 95, and the enormous demand for
all PC-related products that it was supposed to induce, caused
massive stockpiling of semiconductor and other components. In
early 1995, we had an unprecedented period of strong pricing
and chip shortages, which resulted in double and triple ordering;
a fact which was denied by Wall Street even as it was occurring.
This, of course, created record backlogs and caused U.S. and
Japanese corporations, as well as the governments of Taiwan,
Korea, and Singapore to spend well in excess of $120 billion on
semiconductor fabrication facilities. This overcapacity problem
is not the inventory correction Wall Street has been wishing it to
be for two years. It is the early stages of the bust that always
follows the boom. Data from the Semiconductor Industry
Association backs this up: In 1995 worldwide semiconductor
sales were $144 billion; in 1997 they were $137.2 billion. In my
opinion, 1998 will be considerably lower than last year.
Microsoft is a unique high-technology
company. It is one of a kind, a monopoly.
Monopolies in capitalism are quite rare
and almost never seen in technology. It
also may be one of the best-managed
companies of modern times. The
question for investors is: what is it worth?
Presently the market capitalization is
$220 billion, or 3.1% of GDP. In the first
seven weeks of 1998, Microsoft added
roughly $55 billion of market
capitalization, which is about the amount
the U.S. economy will grow in the first
quarter of 1998 (if growth is 3%) and four
times trailing 12-month revenues of $13.1
billion. During its most recent conference
call the guidance was for 2-3% sequential
growth and 10-12% year over year growth
for the next few quarters. Investors are
paying 50 times earnings and 17 times
revenues for a big pickup in the growth
rates sometime in 1999 or beyond.
Potential problems with the Justice
Department apparently are of no concern.
Intel, valued at $138 billion, is probably
the most influential stock in the computer
technology universe. It better reflects the
PC marketplace than Microsoft, which enjoys a better position
and can raise prices and grab share in new markets. In
November of 1997, Worth magazine devoted most of the issue
to Intel, and Time magazine voted Andy Grove "Man of the Year".
This may give some insight into just how popular it is with today's
"new era" investors.
Intel fared better than other semiconductor manufacturers and its
peak didn't occur until the fourth quarter of 1996. A combination
of new entrants into the PC market (the Japanese) and higher
average selling prices on its MMX chips created the illusion that
the business was much better than it really was. At the time,
Intel's revenues and units were growing much faster than those of
the PC industry - which is what it sells into, so it was clear then
that something was not quite right. Wall Street, instead of
recognizing the period for the aberration it was, in typical fashion
extrapolated it for the foreseeable future.
In early 1997, Intel was believed to be a monopoly that could do
no wrong and was expected it to earn $6.00 for the year. Instead,
it made $4.00. This year $3.00 is a good, if not a generous
guess, and next year's earnings could go below $2.00 by my
reckoning. This is a perfect example of how fast things change in
technology. From total (apparent) command to fighting for
scraps in fifteen months, during a boom. Intel is in big trouble.
For all Intel's knowledge, it missed the transition to low priced
PCs, betting that people would keep paying up for horse power
rather than opting to get the computing job done as cheaply as
possible. The P2 was a major disappointment, not even coming
close to internal expectations. More importantly, there is no
longer a penalty for having a non-Intel processor in your PC, and
there are many suppliers to choose from. Virtually every vendor
has a non-Intel product as they try to cut the cost of PCs in the
unfolding price war. The processor has now become a
commodity, a statement that would have labeled you a nut just six
months ago.
Intel has pre-announced three times in the last nine months. The
last time they had quarterly earnings in the range they now
expect, the stock traded in the 40s. At that time, earnings were
growing by 40%, not falling by 40%. This will be the first down
year in nearly a decade, something that didn't occur even in the
last recession. Yet Intel still trades at about five times revenues.
Without a doubt, PCs are the biggest commodity in technology.
Everyone agrees that DRAMS are a commodity, but to make
them you need a very sophisticated billion dollar fabrication
facility (fab). PCs, by comparison, don't even require a clean
room, they are just plastic boxes full of chips.
It has essentially become a delivery business. (PC companies
spend almost nothing on research and development, the only
area of technology where this is true.) The new business model,
build-to-order, is essentially this: Call them up, tell them what you
want, and they'll ship it. If Federal Express or UPS did this, would
investors think it was such a great business? For some reason,
investors are captivated by PC manufacturers. Could it be
because they all have them on their desks, so they know it must
be a leading edge growth business?
Consumers have come to realize that $1,000 boxes will do the
job and are now shopping based on price, not processor
bragging rights. After all, many features of the $1,500 and up
models result in no discernible improvements for most users.
So, the PC industry is facing essentially
no revenue growth as 10-15% growth
worldwide is swamped by even greater
average selling price declines. We are
witnessing industry revenue growth at the
(end of?) a boom that is lower than what it
was in the last recession. This places
tremendous pressure on the margins of
the PC makers and their suppliers as they
try to reduce costs. Price concessions
are not difficult to come by as virtually
everything is in excess supply; however,
the advantages flow to all the participants
thereby negating any real advantage to
any particular PC maker. A classic
example of profitless prosperity.
PC revenue growth topped out in 1995 at
25% worldwide. At that time, revenue
growth and unit growth were fairly similar
as average selling prices were not under
too much pressure. That changed fairly
drastically in 1997, when a saturated
market was invaded by additional
capacity (the previously mentioned
reason for Intel's strong results in late 1996) as manufacturers
attempted to cash in on growth rates that had been extrapolated
but were erroneous.
The price collapse of PCs has been breathtaking. In the
consumer market in 1997, average selling prices plunged from
$1,642 to $1,169, a drop of 29% according to Computer
Intelligence. We are seeing the same size drops in the corporate
market thus far in 1998 as the price wars spread.
At the height of the Windows 95 euphoria, pundits believed that
the home market was going to be the industry's source of growth
because the business market was a replacement market. Now
the home market appears to be a replacement market as well.
Lower prices do not appear to be expanding the market much
and are instead cannibalizing higher priced sales. According to
the research firm Odyssey LLP, only 35% of the people buying
sub-$1,000 PCs are first time buyers, the other 65% are repeat
buyers. Nick Donatiello, president of Odyssey, states " [PC
makers] thought that it [price cuts] would put a kick in the growth
curve, that we were going to see much greater penetration as a
result of the reduction in price. It didn't happen." Odyssey also
found that PCs priced above $1,500 dropped to 24.8% of the
market from 59.3%, alarming news for high-end PC makers.
In the last year, every PC maker except Dell has either
pre-announced or missed earnings estimates, including IBM,
Hewlett Packard, Micron Electronics, Compaq, Gateway, and
Toshiba. Distributors are swimming in inventory thanks to
"channel stuffing" by IBM and Compaq. Compaq recently came
clean (it had been the most blatant channel stuffer) by announcing
that it would have an earnings shortfall of over $500 million, just
two days after Intel admitted its earnings would be substantially
lower than expected.
However, investors refuse to connect the dots concerning these
problems, with the stocks still selling at rich valuations relative to
any reasonable estimate of what they will earn in the future.
Wall Street appears to be the most myopic regarding Gateway
and Dell, vendors who have professed no interest in the
sub-$1,000 market. They plan on continuing to sell high priced
PCs with better margins. Good luck. Both stocks are near all
time highs even though Gateway stumbled in three out of the last
four quarters. Gateway sells at 57 times earnings, Dell sells at
46 times earnings. Dell's market capitalization is $45 billion, a
staggering 30% of total worldwide PC sales of $150 billion.
Since the semiconductor revenue peak, prices for SRAMS,
DRAMS, flash memory, and processors have collapsed as all
niche segments have seen pricing come under assault.
Backlogs in the industry have evaporated as most business is
done in a hand-to-mouth (just-in-time) fashion.
The problem is that once the $1-2 billion dollars are spent and
the fabs are constructed, the fabs will operate. If margins aren't
there in SRAMS, they will switch to flash, or DRAMS or
EPROMS, etc. Manufacturers will swarm over anything with profit
margins since capacity is fungible.
Programmable logic device suppliers like Altera, Xilinx, and
Lattice Semiconductor still sell at 5 times revenues even though
growth over the last two years has been minimal. They've missed
earnings estimates, backlogs are gone, pricing is weak, and they
face increasing encroachment from application-specific
integrated circuit manufacturers.
Makers of analog components, Linear Tech and Maxim, the most
economically sensitive and the most exposed to the glutted
notebook market, sell at 11times revenues. Investors continually
wait to get beat over the head with bad news rather than
anticipate where things are going and plan ahead.
No discussion of semiconductors would be complete without
discussing the stock market's favorite speculative technology toy,
Micron Technology. Amazingly, this money-losing producer of
commodity DRAMS was the fourth most active stock on the New
York Stock Exchange last year, just behind AT&T, ahead of
General Electric, Coca-Cola, Walmart, and Boeing. (This shows
how speculative the stock market is once one examines the
fundamentals at Micron Technology.)
In the most recent quarter they lost about $208 million from
operations. Semiconductor revenues fell to $283 from $440
million last quarter, down 35%. Even more stunning,
semiconductor inventories were almost $400 million, 140% of
revenues. (Looked at differently, this DRAM inventory would
weigh 78 tons.) They are currently losing over a dollar on every
part they make - a staggering fact when you realize that the parts
now sell for about $2. Bulls believe that as Micron transitions
from 16-megabit chips to 64-megabit chips, life will get better.
Nothing could be further from the truth. Asia-Pacific is ahead of
Micron in this transition but, more importantly, PC sales will be
insufficient to absorb the massive increase in supply
64-megabits will create. Sixty-four-megabits will be the same
disaster as 16-megabits. During the entire three-year life-cycle
of 16-megabit chips, cash flow has been negative at Micron! It is
doubtful if Micron will make any money in 1998 or 1999. Yet,
when you adjust for their publicly held PC business, the
semiconductor business effectively has a market cap of $6 billion
dollars, six times annualized (loss making)
revenues.
One of the reasons for the massive glut in semiconductors has
been that domestic producers and Asian nations pursued a
game of chicken. They continued to move ahead with capital
expenditures even after the fundamentals peaked out. Everyone
seemed to believe they could out-spend each other and be better
positioned, which only exacerbated the problem. In doing so,
they kept business going for the capital equipment manufacturers
at least a year longer than they should have, as they continued to
create capacity which increased the glut. The semiconductor
equipment makers are now paying the price, as we are starting
to see massive cancellations. The business will now be worse
for a longer period due to the lack of cutbacks when they should
have occurred. The big equipment stocks - Applied Materials,
KLA-Tencor, Novellus Systems, etc. - still sell at prices that they
sold at near the peak of the semiconductor boom in 1995. When
they sell at valuations that are typical for cycle troughs, the stock
prices will be at least 60% lower.
The insatiable need for bandwidth is the battle cry used to justify
any and all networking investments. It is true we do need more
bandwidth, but that doesn't mean that typical commodity-type
pressures aren't at work today in the businesses. The facts
speak for themselves. Growth in the networking sector collapsed
from 48% in 1996 to 16% in 1997.
Investors tend to forget that this is a very GDP-sensitive business
and its days of uninterrupted growth are over for now. In reality,
the demand for bandwidth is far exceeded by current and future
supply as evidenced by product price erosion and deteriorating
operating results for all the major networking companies.
However, you would hardly know this was happening by looking
at current valuations.
The Internet is the ultimate imagination machine. The Internet will
be part of the future, it's just that we don't know quite how. That
hasn't stopped today's speculators from deciding that they not
only know who will win, they know what it's worth. Yahoo!,
Amazon.com, and America Online, with barely any earnings
collectively, have a combined market capitalization of $20 billion,
equal to approximately half of the entire market capitalization of
the Thailand stock market. Never mind that there's no barrier to
entry and the supply of product nearly infinitely expandable. In
today's bull market it is the perfect industry: no fundamentals,
therefore no disappointments.
The dreams and imaginations of technology stock buyers
occasionally produce valuations that are wildly absurd. The last
few years have comprised one of those periods. Investors have
placed huge multiples on peak earnings, not understanding that
recent profit margins are not only outside the norm and
unsustainable, but also starting to shrink. Instead of lower
valuations to discount this unusual growth period, the boom has
been extrapolated. So if you put a summation sign in front of this,
what do you have? An inherently tough business with extreme
valuations where fundamentals are deteriorating. The fact that
these problems have all occurred during a boom illustrates the
degree of overcapacity that exists and is a harbinger for massive
problems in the next recession. This in and of itself is a disaster
in the making. If we have a reversion to the mean in the stock
market and/or a recession on top of this, we will see a real
debacle in the sector with prices down 80-90% from the peak.
So, what is an investor to do with technology stocks? If you are
bold, short them. If not, have patience, it won't be that long until
people hate them once again. The question is, will you have the
courage to buy them when they are cheap. My bet? Most won't.
William A. Fleckenstein <fleckenstein@go2net.com>, special to StockSite
-MMV |
