7/21/98 Jubak/Microsoft Investor article. [ASND references]
investor.msn.com
When to hold 'em, when to
fold 'em
No one wants to sell a high-flying stock
too late -- or too early. Here's how to use
price targets to signal the right move.
By Jim Jubak
How long do you hold a winner?
It's a tough and important question. On
the one hand, you don't want to sell too
soon. Investors who sold Cisco Systems
in July 1997 certainly hadn't done badly --
the stock had returned 48% in the
previous 12 months on top of an 80%
return in the 12 months that began in July
1995. But they did miss the 83% gain the
stock recorded over the next 12 months.
On the other hand, you certainly don't
want to stay with a winner while it gives
back all its gains. In the 12 months that
ended on Feb. 14, 1997, for example,
shares of Ascend Communications
(ASND) gained 50% in price. In the
subsequent 12 months, investors who
held on gave the bulk of those gains back
as shares fell 47%. Some round trip.
Price targets can help get you out of a
stock near its top and prevent you from
selling too soon. They're an invaluable
tool -- but only when they're used
correctly.
Too many investors believe that when a
stock hits its target price, it's sending a
clear and certain signal to sell. That's
simply not true. (And it's not true, either,
that a stock that hasn't yet hit its target
price is a "hold.") An investor is probably
better off not setting target prices at all
than using them in this way. Let me
show you how I think target prices should
be used to figure out what to do with a
winner. I'll use three examples from
Jubak's Picks -- Vitesse Semiconductor
(VTSS), Cisco Systems (CSCO) and
Tellabs (TLAB) -- of stocks that have
either hit their targets recently or look like
they will soon.
I'll begin with Vitesse, because this
example illustrates very clearly that a
price target isn't a sell signal -- but
instead a reminder to re-evaluate a stock.
Last week, I upped my price target for
Vitesse for the second time this year.
(See the Update at the end of my
column, "Why I'm getting out of Intel.")
The first time, on April 17, I ratcheted my
target upward to $34 after the company
beat analyst estimates by 2 cents a
share and the stock punched through my
year-end 1998 target of $27.50 (adjusted
for the stock's 2-for-1 split). Last week I
moved my target up to $40 a share by
year-end after the company again beat
estimates. In the run-up to the earnings
report, Vitesse had broken through my
$34-a-share target.
Details
Company Facts
1-yr Chart
Press Releases
Earnings Estimates
Earnings Surprise
Consensus EPS Trend
Earnings Growth Rates
Profit Margins
Upping my target and holding on instead
of selling was easy in each case
because, just as the stock hit my target
price, the company reported revenue and
earnings that exceeded my projections.
In each case, it seemed to me that the
current news said that the future was
even brighter than I had imagined.
Let's see exactly how much brighter.
Before the latest earnings report,
analysts were expecting that Vitesse
would report 69 cents a share for the 12
months ending in December 1998.
(Vitesse is on a September fiscal year,
so to get this figure I have to add up the
actual earnings for the company's fiscal
quarter and projections for the third and
fourth quarters of fiscal 1998 plus
projections for the first quarter of fiscal
1999.) That would have resulted in
earnings growing at 41% a year from the
49 cents a share recorded in calendar
year 1997.
After the earnings report, projections for
1998 went up as analysts factored in the
company's 50% earnings-per-share
growth in the June quarter (over the
year-earlier period) and the implications
of a 67% increase in revenue in the
period. Over the next week, analysts
upped their estimates for the September
quarter by a penny and added another
penny to the December quarter. Throw in
the penny from the recent surprise and
investors are already looking at an
estimated 72 cents a share in December
rather than 69 cents.
That additional 3 cents is actually low, I
think. Sales at Vitesse have been slowed
by a lack of manufacturing capacity -- the
company simply couldn't make as many
chips as customers wanted. That
constraint began to ease in mid-1998 as
the company's new Colorado Springs
plant came online. That extra capacity
has helped Vitesse increase revenue
growth from 63% in the March quarter to
67% in the June quarter. The Colorado
Springs plant still didn't run at full speed
last quarter, so I think Vitesse could
actually tack on a few more percentage
points in revenue growth. I'd add another
3 cents a share to get my estimate of 75
cents a share for calendar-year 1998
earnings.
Sell, hold and buy
decisions shouldn't
be based on
whether a stock
price is still a few
dollars short of the
target, but whether
the projected return
is high enough to
justify holding the
stock.
Also, Vitesse has key products in some
of the fastest-growing sectors in the
telecom industry. Its gallium arsenide
chips are uniquely well suited for the
highest-speed OC48 fiber-optic networks
now being built. The company has scored
close to a clean sweep with the
companies supplying gear to stack more
optical signals inside each fiber-optic
cable. Wave division multiplexing leaders
Lucent Technologies (LU) and Ciena
(CIEN), as well as the still-private
newcomers Tellium and Novatel, are all
Vitesse customers.
What's that 75 cents a share worth? The
stock's current price-to-earnings ratio is
55.6, a slight discount to the growth rate
in the most recent quarter. Analysts are
projecting that the company's earnings
growth rate will fall to 25% in fiscal 1999.
While I don't think it's reasonable to
project 60% growth in 1999, I believe
25% represents too severe a drop. My
guess is that in December, investors will
be looking ahead to great but slightly
lower growth. That should translate to a
lower price-to-earnings ratio -- say, about
5% lower, or 53. Multiply 75 cents a
share by 53 and you get $39.75 -- I round
it up and get my end-of-the year target of
$40 a share.
I don't want to hold on just because the
December target is higher than the
current stock price. I still need to figure
out whether the projected gain is worth
waiting for. With the stock trading at $34
on July 17, I'm projecting a $6-a-share
gain, or almost 18% over the next
five-plus months. That's well above the
25% annual return I require for investing
in a risky small-cap technology stock. So
I'll hold Vitesse for awhile yet.
I'm projecting a
$6-a-share gain, or
almost 18% over the
next five-plus
months. So I'll hold
Vitesse for awhile
yet.
I'd like to stress a key point here. The
most important number in this analysis
isn't the target price itself, but the rate of
return that the target price implies you
can expect from now until the target date.
Sell, hold and buy decisions shouldn't be
based on whether a stock price is still a
few dollars short of the target, but
whether the projected return is high
enough to justify holding the stock.
A stock that has had a fast run-up may
still be way short of the target price, for
instance. But the gain in the recent
period may actually represent such a big
piece of the projected return from owning
the stock for a longer period that it's no
longer worth holding the stock to the
target. Let me show you what I mean,
using Tellabs as an example.
On June 2, I set a target price of $93 a
share for Tellabs for June 1999 ("It pays
to be imprecise"). That was about 41%
above the stock's June 2 price of $65.78.
Trouble is, here we are, not even two
months later and the stock has already
hit $86.69. According to my original
target price, I'm looking at a potential
gain of $6.31 over the next 10 months.
That's a potential return of just 7.3% --
well short of the 25% annual return I
require for a stock with Tellabs' risk.
Unless something has changed in the
last two months to radically improve
Tellabs' prospects, I should clearly sell
the stock -- even though it's still better
than $6 below my target.
Details
Company Facts
1-yr Chart
Press Releases
Earnings Estimates
Earnings Surprise
Consensus EPS Trend
Earnings Growth Rates
Profit Margins
Plenty has changed at Tellabs, however,
and that's exactly what's turned on the
after-burners in the stock. There have
been major changes in both projected
earnings per share and projected
price-to-earnings ratio -- the factors I use
to calculate a target price.
First, the earnings picture has turned out
to be even brighter than I expected. On
June 2, I projected that earnings for the
next 12 months would come in at $1.97.
That already looks too low.
Tellabs surprised Wall Street when it
reported earnings of 46 cents a share on
July 17 -- 4 cents, or almost 10%, higher
than expected. Earnings per share grew
by 44% in the quarter (over the same
period in 1997) and sales climbed 34%.
Importantly for future earnings, sales for
the company's flagship Titan 5500
platform for Sonet-based telecom
networks soared a better-than-expected
51%.
In reaching my earnings projection of
$1.97 for the four quarters ending in
March 1999, I'd already given Tellabs
credit for more growth than Wall Street
expected. So the great earnings and
revenue growth reported by Tellabs this
quarter ups my existing estimate only
modestly to $2.08 a share.
But that's only half of the Tellabs story. In
February, the company agreed to acquire
Coherent Communication Systems
(CCSC), a maker of echo-cancellation
systems that optimize voice levels in
local, long-distance and wireless phone
systems. Then in June, Tellabs agreed to
acquire Ciena, the top maker of wave
division multiplexing equipment.
Together, the two acquisitions make
Tellabs a candidate for entry into the top
tier of communications-network
equipment suppliers. With a product for
the hot asynchronous transfer mode, or
ATM, market on the way this year and
plans to build an optical switch well
advanced, Tellabs could, perhaps, be the
next Cisco or Lucent. (ATM is, for the
moment, the protocol of choice for
managing mixed voice and data
networks.)
Tellabs' earnings
picture has turned
out to be even
brighter than I
expected. On June
2, I projected
earnings for the
next 12 months
would come in at
$1.97. That already
looks too low.
A lot has to go right for that to happen.
The optical switch has to reach the
market and work well at a reasonable
cost. Tellabs' sales force will have to do a
better job of marketing Ciena's gear to
the regional Bell phone companies than
Ciena has. And Tellabs will have to
merge three companies with relatively
little time-consuming turmoil. But the
market has decided that this scenario
deserves a higher price-to-earnings
multiple than a simpler, less ambitious
Tellabs did. The stock's P/E ratio has
climbed from 47 in early June to 49.3 on
July 17. And I think that the market will
keep closing the gap between Cisco's
P/E, which was 51 in early June and 65
on July 17. (Remember that to get an
accurate P/E ratio for Cisco, you have to
recalculate the company's earnings to
exclude the effect of one-time charges.)
So in setting my Tellabs price target for
next June, I'll use a P/E ratio of 52. That
gives me a target price of $108.16 -- a
25% projected return for holding the
stock until next June. (I agree with
management's calculation that the
merger with Ciena will not dilute earnings
per share, and could actually add to
them.)
So far I've dealt with two stocks that have
cooperated mightily with an investor
trying to link fundamentals and future
stock price. In each case, the stocks
have reported improved earnings and
revenue growth just as an investor
needed to decide whether to sell or hold.
Details
Company Facts
1-yr Chart
Press Releases
Earnings Estimates
Earnings Surprise
Consensus EPS Trend
Earnings Growth Rates
Profit Margins
But what do you do when the stock
climbs so fast that it busts through the
price target before the company reports
any substantial new information? In other
words, what do you do about Cisco?
My price target for Cisco is $95 by the
end of December. That seems quaint.
The stock broke through that level on July
15 -- and kept on going. It closed above
$103 on July 20. Cisco shares are up
more than 35% since May 5 -- the date it
last reported earnings. The company isn't
due to report any new numbers for about
another month. So, because Cisco has
climbed so far so fast, I now have a stock
in Jubak's Picks that is over my price
target a month before I have any new
numbers from the company that might
justify a higher target price.
I'm not completely without new
information, however. Some of Cisco's
competitors have reported. Ascend
Communications (ASND) beat Wall
Street estimates by a penny a share, for
example. Bay Networks (BAY) is being
acquired by Northern Telecom (NT), and
Cisco historically has made hay when
competitors merge. See
Ascend/Cascade and 3Com/U.S.
Robotics.
Is that enough for me to make a new
earnings projection and then calculate a
new target price? Not a chance. On the
fundamentals in hand, Cisco is a sell.
Anyone who holds on for the next month
until the stock reports is doing so on faith
and faith alone. Doing that doesn't make
me comfortable. But I'm sticking with the
stock at least until the numbers come in.
Cisco's management has a record of
delivering, and I think that makes a little
faith justifiable.
Besides, I think selling a stock like Cisco
in a market like this is just plain wrong.
I'll explain what I mean by that -- and give
you my theory on what's driving this
market -- in my next column on Friday. |
