AM CALL: Semiconductor: June SIA Data Tough: Lowering `01 Foreca
2001-08-02 08:29 (New York)
CREDIT SUISSE FIRST BOSTON CORPORATION
Equity Research Americas
U.S./Technology/Semiconductors
Charlie Glavin, CFA 1-415-836-7715 charlie.glavin@csfb.com
Tim Mahon 1-650-614-5040 tim.mahon@csfb.com
Michael Masdea 1-415-836-7779 michael.masdea@csfb.com
Regina Eberhart 1-415-836-7767 regina.eberhart@csfb.com
Caroline Moon 1-415-836-6321 caroline.moon@csfb.com
Tyrone Lee 1-650-614-1143 tyrone.lee@csfb.com
Natalie Wright 1-451-836-6329 natalie.wright@csfb.com
In addition, we remain focused on the ASP profile of the analog industry. In
June, we saw the first month of year-over-year declines in pricing. (down 1.1%
Y/Y) We believe this is the result of the depletion of backlog for high-end
products and the decreased impact of the early-downturn fall off in revenue
from the lower-margin, non-proprietary products. While the stronger than
anticipated pricing strength early in the downturn muted the revenue impact
of this severe downturn, we believe it will result in a dampened upturn when
volumes return (similar to the 1997-1998 cycle) and possibly an extended revenue
downturn.
Microprocessors - Decline Suggests Risk to 3Q and 4Q
Microprocessor revenues were down 3.3% in June versus May and were down 31% Y/Y
and down 9% Q/Q. We believe the sharp month-over-month decline reflects
sustained ASP pressure and a trend unlikely to abate any time soon, given
large price cuts form Intel in August (50%+ on some P4 lines), weak back-to-
school lead orders, and further cuts expected through 4Q.
DSPs-Wireless Firmer, But Continued Weakness in Wireline, PCs
Overall DSP revenues were down 2.5% month-over-month, but down 35% year-over-
year and down 10% Q/Q, as relative strength within the wireless (+5.4% M/M)-
the largest component of DSPs at 53% of the mix-and multipurpose/other (-8.9%)
segments was offset by weakness in wireline (-10.4%) and computer/peripherals
(-17.2%). We expect DSP revenues will continue to decline in light of guidance
from TI, the largest supplier of DSPs, for up to a 20% Q/Q decline in revenue
for 3Q01.
While overall DSP ASPs held up for the third consecutive month, we believe
given that unit declines in this segment didn't begin until December (i.e., 6
months ago), that pricing pressure is imminent. However, should pricing hold
firm, based on previous pricing analysis, we believe this would imply a more
muted recovery for this segment once the upturn begins. As noted before,
CSFB analysts Masdea and Long also have predicted a 3Q unit build-ahead for
the introduction of new models (which may account for some of that June
increase), but the sell through will probably not be enough to significantly
reduce that 30M in excess handsets by year end - meaning that half of the
excess handsets exiting 2000 could still carry over into 2002.
DRAM - Double Digit Revenue Decline in June
On a rolling 3-month basis, DRAM revenues declined 21% in June due in part to
a 13% sequential drop in ASPs which overshadowed a corresponding 18% increase
in units. Recent channel checks with memory suppliers in the Far East
indicate that DRAM pricing has begun to see some signs that the rate of
decline is decreasing. Although this may prove little consolation for most
DRAM manufacturers as we believe market prices are now below manufacturing
costs. Additionally with continued weak end market demand, we believe that
the current inventory glut and oversupply situation will continue, thereby,
delaying any possibility of a near-term recovery in the DRAM market. We once
again highlight our forecast for the worldwide DRAM market to be down 50%
from 2000 levels. We are also introducing our 2002 forecast for DRAM to be up
34% as inventories draw down to normal levels and true end market demand
returns.
Flash - Still Recovering in 2002 From 2001 Hangover
In the month of June, the Flash segment saw its second sequential 14% decline
in worldwide revenues, down a whopping 28% on a year-over-year basis. Drawing
from May's SIA update, we previously highlighted the fact that the Flash
segment, on a three-month rolling average basis, saw its worst sequential M/M
decline since the inception of its tracking by SIA. Recent earnings
announcements from leading Flash memory providers indicates a consensus for
continued near-term pricing erosion coupled with continued excess inventory
for the remainder of the year due in large part to the manufacturing capacity
ramp up from the last boom period. In line with recent grim forecasts from
leading industry players, our revised 2001 forecasts have the Flash segment
down 26% from 2000 levels. In keeping with our view for a long, protracted
recovery in the Flash segment, we are forecasting a 2.6% increase for the
segment in 2002.
PLD - Flat 2002 Forecast
Worldwide PLD revenues saw a slight decline of 1% in the month of June,
however, in comparison to last year's record setting environment was down 58%.
Recent earnings announcements from the top three players highlight the
grim reality that the PLD segment continues to be plagued by a weak
communications end-market from they derive the bulk of their revenues and
mounting days of inventory internally and at distributors.
Having said that we are lowering CSFB's forecast for the 2001 PLD market to
be down roughly 51% from record 2000 levels. Given the industry wide excess
inventory and the unlikelihood for the communications end-market to return in
any meaningful fashion, we are taking this opportunity to introduce our 2002
PLD forecast for a slight decline of 1.1% in 2002.
Why We Don't Rely on the DOC Book-To-Bill, and Why the SIA Dropped It
Recently, some investors have turned to semiconductor book-to-bill data that
is released monthly by the Department of Commerce (DOC). However, we believe
these investors should take DOC data with a grain of salt before relying too
heavily on it as a leading indicator of a real recovery. We offer two points.
First, recent optimism in the market has been based on a "sustained" upturn
in the DOC order patterns for the past two months of May and June. However,
we note while there has been an uptick in book-to-bill data, this is based on
a single-month, standalone basis, and not on a rolling-3-month average. In
fact, on a 3-month-rolling average basis, book-to-bill is actually down for
the second consecutive month. We stress to investors that they should
examine rolling-3-month data, given the unreliability of single-month
snapshots of book-to-bill and the large volatility of the data month-to-month.
Second, book-to-bill data has not been a good leading indicator of stock
movements-which is presumably why investors would focus on the data. Exhibit
10-adjusted for the impact of the 1-month lag in the book-to-bill data release
-shows that had you relied on book-to-bill data, in September 1999, you would
have sold on the "can't get any better" premise when book-to-bill peaked, and
then would have missed 6 months worth of the rally through April 2000, or a
146% return. Conversely, if you bought into stocks based on a trough in book-
to-bill numbers during 2000, you would have been squeezed by the downturn,
with a 56% stock depreciation between April 2000 and today. Granted, the
data does work on occasion but with such inconsistency, that we examined how it
was collected, too.
We note that the SIA itself stopped tracking bookings due to a quality/
consistency issues with order forecasts given by the same companies its
surveys for billings. What's more, based on feedback from the people running
the SIA data, neither they nor the semiconductor companies it surveys know
themselves where or how the DOC is obtaining its data. We believe gross,
rather than net, bookings are being used-thus rendering the data virtually
useless-and we also understand that if a company does not provide a bookings
number for a certain month, the DOC estimates and extrapolates one-i.e., a shot
in the dark.
The bottom line is that book-to-bill data should not be used by itself as a
reliable metric to determine the health of the industry, particularly now,
when we're dealing with the law of small numbers: if customers have cancelled
all that they could cancel, any bookings would spike the number - but a three
month trend could at least indicate whether it was a sustainable demand-pull
recovery.
Valuations Remain Too High, Use Run-Up to Take Profits
We reiterate that a true recovery is signaled by strong, sustained order
growth, and we caution investors not to be lulled into a false sense of
security from companies reporting slowing rates of cancellations and order
push-outs. Without demand from capital spending programs upstream, a call
that we are seeing sustainable (and not just seasonal) improvements, may be
premature. The vast majority of companies in our universe reported double-
digit sequential declines for 2Q01, and guidance for 3Q varies from another
quarter of double-digit sequential declines (PMCS, TXCC, TXN, VTSS) to modest
sequential declines (AMCC, BRCM, CTLM, INTC) to sequential growth (MLTC, RFMD).
We believe, particularly in light of the recent run in semi stocks, that
valuations continue to assume a robust recovery, with a return to 2000-type
growth rate assumptions embedded into many companies' stock prices (see
Exhibit 11)-a scenario which we believe is highly unlikely for 2002. So far,
the stocks during this downturn have not been supported by a turn in
fundamentals, but a fund flow "long squeeze" - as large institutions have had
to have some weighting in semis, not out of preference, but because there's
limited choices in this economic slowdown.
We continue to believe that we have probably seen the worst in terms of order
declines, but also feel that growth will bump along the bottom. In light of
continued multiple expansions (and stock runups), investors should take
advantage of profit-taking opportunities. Conversely, we believe investors
should not chase stocks for several months, given that over the last three
cycles we have witnessed summer sell-offs in July and August.
N.B.: CREDIT SUISSE FIRST BOSTON CORPORATION may have, within the last three
years, served as a manager or co-manager of a public offering of securities for
or makes a primary market in issues of any or all of the companies mentioned.
Companies mentioned:
Broadcom Centillium Intel Applied Micro Circuits TranSwitch PMC-Sierra Vitesse
Texas Instruments MultiLink RF Micro Devices
CREDIT SUISSE FIRST BOSTON CORPORATION CREDIT SUISSE FIRST BOSTON CORPORATION
Analyst: Glavin, C Telephone: (415) 836-7715 charlie.glavin@csfb.com
CREDIT SUISSE FIRST BOSTON CORPORATION CREDIT SUISSE FIRST BOSTON CORPORATION
Copyright © CREDIT SUISSE FIRST BOSTON, and its subsidiaries and affiliates,
2001. All rights reserved.
CSFB may, to the extent permitted by law, participate or invest in financing
transactions with the issuer(s) of the securities referred to in this report,
perform services for or solicit business from such issuers, and/or have a
position or effect transactions in the securities or options thereon. In
addition, it may make markets in the securities mentioned in the material
presented in this report. VIEW CSEQ MENU TERMS&CONDITIONS FOR IMPORTANT LEGAL
DISCLOSURES.
Provider ID: 10215332
-0- Aug/02/2001 12:29 GMT |
