By Thomas W. Smith With mounting evidence of slowing business conditions, semiconductor stocks fell pretty hard in late 2000 until the Federal Reserve started easing interest rates in January. The Fed's easier monetary policy is giving some assurance that the economy-- and the chip sector-- might strengthen by the second half of 2001.
Coming off low levels at year-end, the semiconductor sector as tracked by Standard & Poor's jumped 18.6% in January 2001, versus a 3.5% gain for the S&P 500 Index, while the semiconductor equipment sector rose 32.5%.
That's a nice beginning to a comeback for the group. But the stocks remain well off their 52-week highs. So if you think the economy is going to rebound by late 2001, now is the time to consider investing in the better positioned semiconductor companies.
Before we list recommendations, investors should keep in mind that it will be impossible for semiconductor companies to top growth seen last year. In fact, the semiconductor industry had its biggest boom year ever in 2000. Worldwide chip sales grew 37% to $204 billion, according to the Semiconductor Industry Association (SIA).
The economic slowdown contributed to inventory build ups for most types of chips in late 2000, prompting most semiconductor makers to revise their revenue guidance downward for 2001, particularly the first half. The SIA had been forecasting growth of 22% in 2001, but on February 5, 2001, SIA president George Scalise confirmed in a press release that the association's 22% projection was unlikely to be achieved. Mr. Scalise warned that the inventory overhang might drag growth down as low as 10%, according to The New York Times. Ouch!
As usual, private forecasters have been ahead of the SIA in changing forecasts. In January, Bill McLean of IC Insights forecast 7% growth for chips in 2001. At the same time, Dataquest said it is looking for 5% to 10% growth, down from its earlier forecast of 15% to 20% growth. VLSI Research estimates 5% chip industry growth.
While growth has been impaired for the chip makers, the semiconductor capital equipment suppliers may suffer even more this year. Chip equipment makers' estimated worldwide sales in 2000 rose 83% to $46.7 billion, according to the trade association Semiconductor Equipment and Materials International (SEMI). Orders for semiconductor equipment were front-loaded into the first half of 2000, and many suppliers were already seeing orders decline in the summer. This year, capital spending on chip equipment is expected to decline 4%, according to IC Insights. The other research shops estimate spending to be near break-even to down 6%.
BAD NEWS BEHIND US? The chip stocks are likely to remain subdued through February and March 2001 as more bad news about order cancellations and pushouts are revealed. For example, Cisco Systems (CSCO), a prime buyer of many high-performance communications chips, announced weaker than expected results on February 7, 2001, sending the sector down for the day. The negative announcements cause concern that the widespread chip inventory congestion problems might not be resolved by summer, as most observers appear to believe.
The good news is we anticipate further interest rate easing in March and later in 2001. Also, some investors may argue that Cisco's announcement is the high-water mark for actual bad news, so investors should just assume the first half of 2001 is more-or-less ruined and begin pricing the stocks on estimates of business conditions in late 2001 and early 2002. Barring negative surprises in the global economy, growth in late 2001 and early 2002 should be a lot better.
OUR FAVORITE STOCKS. Between the brightness of lower interest rates and the darkness of weak earnings, I am slowly turning from neutral to bullish on the chip sector. I think shares of chip makers will do better than those of chip equipment makers because the chip cycle should play out longer. That is, the wafer fabrication plants (fabs) must already be full of new equipment before the plants can run so hard that chip pricing conditions deteriorate.
Chip makers that are doing fairly well in the inventory correction include Atmel (ATML
, 5 STARS (buy)), which gets about one-fourth of its revenue from flash memory chips, which remain in strong demand. Another one I like is Microchip Technology (MCHP
, 5 STARS), which saw some slowdown in orders, but appears ready to get beyond its inventory problem by March 2001, about one quarter ahead of many rivals. Both ATML and MCHP sport modest valuations, relative to many other chip makers.
Two higher fliers that I am sticking with as buy recommendations despite the inventory woes are Integrated Device Technology (IDTI
, 5 STARS) and Vitesse Semiconductor (VTSS
, 5 STARS). Both of these appear to offer good long-term growth prospects at reasonable valuations for high-growth companies. That said, they have suffered as Cisco's problems grew. I am in the camp that believes that the worst for Cisco is being announced now, and that conditions for chipmakers serving the fast-growth communications equipment niches will look a lot better in six to twelve months.
, 5 STARS) is a company with superior growth potential that is trading at a reasonable valuation, albeit at a premium to the market's valuation. Xilinx is the leader in programmable logic devices (PLD chips). These chips can receive final programming by the customer, who is typically a communications original equipment manufacturer. This offers considerable time-to-market advantages to the customer, who is willing to pay a premium for PLDs over traditional chips. PLDs have a chance to take market share from the alternative application-specific integrated circuits (ASIC chips) as PLDs are offered with higher performance at lower cost.
In the chip equipment area, I have marked most of the companies on my list down to hold. That includes quality front-end equipment makers such as Applied Materials (AMAT
, 3 STARS (hold)), Novellus Systems (NVLS
, 3 STARS), and Lam Research (LRCX
, 3 STARS). I am watching some of the back-end equipment makers for signs of life, because their order flow turned down first in the group in summer 2000, and they might turn up first too. Higher orders for back-end equipment generally indicates an industrywide intention to produce higher volume runs of chips, indicating a broad recovery. (Front-end orders are made for more diverse reasons. They might be made to expand capacity, but they might just be for plant modernization without an intention to increase capacity.)
On the January earnings calls, however, business conditions still looked tough for two leading back-end equipment suppliers: Kulicke & Soffa (KLIC
, 1 STARS (sell)) and Teradyne (TER
, 2 STARS (avoid)). We downgraded Kulicke & Soffa shares on Feb. 14 to sell from hold after the company announced it would cut 7% of its workforce (300 employees) amid weak orders.
On the other hand, some equipment suppliers have greatly expanded market opportunities now that the next generation 300mm wafer fabs are being built. These include Brooks Automation (BRKS
, 4 STARS (accumulate)) and PRI Automation (PRIA
, 3 STARS). Their type of factory automation equipment is generally required for the heavier, more valuable 300mm wafers, whereas it was optional at the 200mm wafer fabs. Companies that have a high percentage of such "technology buys" in their order flow are doing better in the present downturn than those that are more dependent on "capacity buys." Among the front-end equipment companies, an example of a company with a lot of must-have technology for the 300mm fabs is KLA-Tencor (KLAC
, 4 STARS), a maker of yield management equipment.
The companies listed above with accumulate or buy rankings are certainly worth considering as the sector starts to get some wind under its sails. Smith is an equity analyst at Standard & Poor's