Semiconductor Scenarios Diverge In 2020; Who's Best Positioned?

In a report published Tuesday, Pacific Crest analyst Weston Twigg discussed how semiconductor scaling challenges threaten to slow or limit leading-edge node migrations, especially among foundry producers.

According to Twigg, the rate of equipment demand growth depends on two node migration scenarios through 2020. First, most new capacity is gained through leading-edge additions; second, significant new capacity is added to older nodes.

Under the first scenario, capital expenditure accelerates "meaningfully" in 2019 to 2020 as 10 nm matures and 7 nm ramps. Under the second scenario, capital expenditure growth is "moderate" although several new 28 nm fabs would need to be built.

"In our view, if scaling slows, the equipment segment is still very investable," Twigg wrote. "Slow and steady investment at various nodes, instead of racing to add leading-edge capacity, could help optimize supply and demand and limit periods of overcapacity."

Twigg said that scaling challenges do exist and the high costs related to node migrations will likely slow node transitions and limit the number of companies that will operate at the new leading edge. The analyst also noted that he would be "surprised" if any foundries ramp 5 nm technology by 2020, which would match the conventional node cadence.

Who Is Best Positioned?

According to Twigg, Applied Materials, Inc. AMAT and Lam Research Corporation LRCX are "best positioned" for a scenario where scaling slows (scenario two). In addition, both companies have good exposure to strong leading-edge spending (scenario one), but should also benefit on a relative scale versus lithography and process control if leading-edge spending slows and older 28 nm or 45 nm spending improves.

The analyst also added that Teradyne, Inc. TER "should perform well" under both scenarios, but chip complexity would likely help the company in scenario one.

Finally, ASML Holding NV (ADR) ASML, Nanometrics Incorporated NANO, FEI Company FEIC and KLA-TEncor Corp KLAC all have the potential to outperform if 10 nm and 7 nm progress as planned in scenario one. However, the analyst did caution that he views slower scaling as more likely.