The ongoing shift to more efficient technologies will boost capital expenditures by chipmakers 20% to 30% next year. Sound like yet another argument to buy shares of red-hot semiconductor-equipment makers? Not necessarily.
Simply put, cyclical investments in technology upgrades don't reflect underlying demand for semiconductors and the products that utilize them. And without a massive surge in end-user buying, shares of equipment makers will remain overvalued. That's the thesis of Banc of America Securities analyst Mark FitzGerald , who stood nearly alone on Monday against a
flurry of sell-side enthusiasm for the sector.
"We don't expect this technology investment cycle to broaden into a capacity investment cycle unless there is a 1990s type of recovery in the end markets," FitzGerald said in a note to clients, adding that a new boom is "not currently our view of the world. It is difficult to imagine that industry earnings in a 'technology-only' cycle would justify current evaluations."
Since the beginning of the year,
, the sector's 800-pound gorilla, has appreciated by 61%, while the Philadelphia Stock Exchange Semiconductor Index is up 56%.
"Applied Materials' broad product portfolio provides extra leverage to technology and capacity buys early in the cycle," Merrill analyst Brett Hodess wrote. "Combined with cost-cutting, this should provide stronger-than-group-average earnings growth in the first part of the cycle."
Likewise, Hodess praised cost-cutting and innovation at much smaller
Kulicke & Soffa
, which has soared 108% this year, with much of the appreciation in the last two months. Like AMAT, it also was upgraded by Merrill. Merrill is seeking investment banking business with both companies.
Also upgraded by other brokerages recently were
Varian Semiconductor Equipment
Although Applied Materials, like other equipment makers, has faced three years of declining revenue as chipmakers cut spending, the recent signs of recovery noted by
and others have encouraged investors to jump back into the stock. "I'm not saying there has not been improvement, but investors seem to think it's the 1990s again," FitzGerald said. (BofA does not have a current banking relationship with Applied Materials, but may seek one.)
Monday's upgrades prompted a caustic comment from hedge fund manager Bill Fleckenstein, president of Fleckenstein Capital, and a
contributor. "These valuations are insane. Strip out the mania years, and this company has never performed the way the sell side predicts," he said.
At Monday's close of $20.58, Applied is trading at 158 times its estimated 2003 earnings (its fiscal year ends in October). That multiple shrinks to a more reasonable 23 in 2005 -- if Applied earns the 89 cents on $7.7 billion in revenue that is the analysts' consensus gathered by Thomson First Call.
That would require results approaching those of the company's best during the boom of the late '90s. In 2000, AMAT earned $1.21 on sales of $9.5 billion. In its previous fiscal year, it earned only 45 cents on $5.1 billion.
Expectations for 2004 are a bit more modest -- but still high. Wall Street expects a 46-cent profit on sales of $5.7 billion. Analysts expect the company to make 13 cents on $4.4 billion in revenue this year.
One other bearish point: Memory makers and Intel account for 40% to 50% of semiconductor-equipment spending.
Should DRAM stall in 2004, choked by excess capacity, as FitzGerald expects, and if Intel's capex for 2004 is flat to down, the sector could take a major hit.