<Lehman Bros. on the Big Picture>
Headline: Semiconductors: Thesis in a Nutshell - UPDATE
Author: Michael A. Gumport, CFA 1(212)526-5368
Company: AMD,ALSC,AVT,CYRX,IDTI,INTC,IRF,LSI,MU,NSM,PIXT,RMTR,STM,TXN,XLNX
Industry: SEMICO
Ticker :
Today's Date : 10/21/97
------------------------------------------------------------------------------
* 12/95 THESIS (RECAP): 1) 1978, NOT 1984; 2) DEMAND, NOT SUPPLY; 3)
OPPORTUNITY; 4) DON'T GET STEAMROLLED, BUT BAD NEWS TO BOTTOM 1Q96, 5) MAJOR
REBOUND IN EARLY 1996.
* 10/97 THESIS (UPDATE): 1) HAVE SEEN A 35% SNAPBACK IN CHIP UNIT VOLUME
SINCE MID 1996 AS INVENTORY CORRECTION HAS COME TO AN END; 2) ONGOING SOLID
GROWTH IN INDUSTRIAL PRODUCTION APPEARS LIKELY TO SUPPORT ONGOING STRONG CHIP
DEMAND GROWTH; 3) ONLY THE MEMORY SECTOR IS STILL SUFFERING SIGNIFICANT
CAPACITY EXCESSES, 4) EXPECT STRONG CHIP FUNDAMENTALS AND STOCK PERFORMANCE.
* CURRENT SITUATION: INDUSTRY DECELERATING BECAUSE 1) 35% UNIT VOLUME GROWTH
UNSUSTAINABLE, AND 2) SUMMER IS NORMALLY SLOW. DECELERATION CAUSING SLOPPY
EPS RESULTS AND LIMITING VISIBILITY. SEASONAL STRENGTH ACCELERATES NOVEMBER-FEBRUARY
(PEAKS IN APRIL). STRONG UNDERLYING FUNDAMENTAL DEMAND DRIVERS
SHOULD MAKE FOR HEALTHY ORDERS GROWTH AND RENEWED CHIP STOCK STRENGTH OVER THE
NEXT 6 MONTHS. RECENT CHIP STOCK WEAKNESS A BUYING OPPORTUNITY.
----------------------------------------------------------------------------
FAVORITE STOCKS: LARGE CAP. Among large cap issues, we continue to focus on
most improved companies with substantial profit margin upside including TXN-$
121 7/8-1 and STM-$78-1. We also continue to believe the long term
consolidation story within the chip distribution sector is intact and that
AVT-$61 15/16-1 will be a key beneficiary. The biggest change in our position
during the past 3-6 months has been our increased concern that INTC's-$84 1/16
-2 market share is in decline and that the stock's multiple will shrink. INTC
is doing everything right (accelerating product cycles, for instance), but, at
the moment, its reception in the market seems more related to success/failure
of clone makers than its own strong new product offerings.
FAVORITE STOCKS: MID/SMALL CAP. Among smaller cap issues, we continue to see
best stories at LSCC-$58 3/4-1 (the largest market share gainer in PLDs; will
double in size in 3 years, profits more than double, and multiples expand) and
VTSS-$44 7/16-1 (unique, high integration GaAs technology fuels 40% annual
growth). IRF's-14 7/8-1 Sept. was disappointing, but the big break in the
stock looks like a clear opportunity and we expect a quick pickup in demand.
XLNX-$37 7/8-1 has limited current momentum, but we continue to see them as
the leader in the long term most attractive segment of the PLD market, FPGAs.
LSI-$24 1/2-1 near term is still facing difficult EPS prospects. And memory
makers MU-$33 3/16-1, IDTI-$12 11/16-1, and ALSC-$9 7/16-1, too, all face poor
short term prospects.
FAVORITE STOCKS: VENTURE/NEAR VENTURE. Among the venture or near-venture
situations we cover, we continue to see RMTR-$7 9/16-1 on a jagged path
towards a $14 1 year price target: Both its FRAM and EDRAM technologies seem
to be making good albeit erratic progress. PIXT's-$3 5/8-V3 flat panel
effort, meanwhile, still appears to be taking dramatically longer than we
anticipated, and the real test of its commercial viability will not occur
until mid 1998 when its foundry partner is scheduled to begin to ramp
production.
Chart 1: U.S. Chipmakers' Domestic Orders versus Industrial Production
(Chart here for LB sales)
CHIP ANALYST RULE #1: INDUSTRIAL PRODUCTION TIMES 10. The correlation
between industrial production and chip orders is by far the best of any
macroeconomic variable. Chip orders are no longer published. But we are
confident that chip orders would now be reported at more than 50% above a year
ago based on the healthy ongoing growth of industrial production.
Chart 2: Worldwide IC Unit Volume
(Chart here for LB sales)
DECELERATION: The 1995-1996 chip industry setback reflected an inventory
correction. The U.S. economy decelerated in late 1995, and that triggered a
chip inventory correction. The economy reaccelerated quickly, but the
inventory correction continued until August 1996. At that point, chip unit
volumes were 7% below the prior year even though end demand was up 20% (and
chip inventories at electronics manufacturers had been trimmed by 1 month,
roughly 50%). Now unit volume has snapped back to match end demand: In July
1997, IC unit volume was 35% above the prior year. That growth largely
reflects catch-up after a period of underbuying and is not sustainable. We
are now decelerating towards a sustainable 20% growth path (August was up
just 29% vs. last year). Normal summer seasonal are accentuating the
deceleration.
Chart 3: Industrial Production versus Real M3
(Chart here for LB sales)
WE SEE NO RECESSION. In fact, the economy has been quite robust, and that is
good news for chip makers. Ultimately, we defer to economists. Still, we
track a variety of variables which in the past have been good indicators of
industrial production (and, hence, chip orders). In that regard we note that
real M3 (which, despite a gap in 1993-1994, has among the best long term
records as an indicator of industrial production) is very solidly positive.
Higher inflation or higher interest rates (and tighter money) could change
this, but, for the moment, it would be hard to have this indicator more
positive than it is today. ONGOING GOOD INDUSTRIAL PRODUCTION GROWTH (0.3%
PER MONTH) SHOULD PUT THE CHIP INDUSTRY ON A 20% GROWTH PATH; EXPECT TO FEEL
STRONG REACCELERATION IN THE NOVEMBER-FEBRUARY TIMEFRAME AS NORMAL SEASONALS
AUGMENT THE UNDERLYING MACROECONOMIC DYNAMICS.
Chart 4: Industrial Production versus Real Retail Sales
(Chart here for LB sales)
A SECOND, GOOD MACROEONCOMIC INDICATOR OF INDUSTRIAL PRODUCTION TRENDS (AND,
HENCE, CHIP ORDER TRENDS) HAS BEEN THE RATIO OF REAL RETAIL SALES TO
INDUSTRIAL PRODUCTION. We note that this ratio, despite a slight
deterioration in recent months, continues to trend up. That suggests
production is largely selling through into end markets and that production can
be sustained at current levels and move higher as consumption expands.] |
