Beliefs, Perceptions and Reality-Semiconductor Capital Equipment

Growth has a seductive charm. There is a widespread belief that momentum drives growth and that a succession of knocking down challenges and consequent victories is what characterizes a successful business. But sustained growth is seldom straightforward. Sustained growth frequently occurs as a result of changing course, breaking rules, and changing the rules of the game. I think some of the rules have changed for the semiconductor capital equipment companies.

Years ago, within the semiconductor industry, there was a strategy that merging companies would result in a reduction in the historic overcapacity issue. This trend certainly developed as Texas Instruments sold its DRAM capability to Micron and Hyundai and LG merged their operations. Small players would be wiped out or have to find niches to survive.


Last month, the Semiconductor Industry Association (SIA) reported that global sales of semiconductors grew last year by 3.2%. It really is an amazing figure given the fact that cell phone unit shipments grew 20%, that laptops grew 32.2%, that LCD TVs grew 50%, and consumer appetite for electronics seems unabated globally.

Total bit shipments for DRAMs nearly doubled in 2007, but total revenues declined by 7.4 percent due to a decline of more than 39 percent in ASPs. NAND flash revenues were up 26 percent but unit shipments grew even faster at nearly 46 percent, while ASPs declined by 13.7 percent.

Increased concentration in the industry simply has fanned the flames of competition and price cutting rather than quell them. To quote Steve Pelayo of HSBC who recently was interviewed in The Wall Street Transcript (subscription required):

"DRAM manufacturing has been in a state of oversupply, which resulted in greater than 75% average selling price (ASP) declines last year. As a result of the significant ASP pressure, many DRAM players today are now reporting operating margins that are significantly in the red, as much as negative 50% operating margins. So clearly too much excess supply in DRAMs and a lack of profitability is causing a massive contraction in their capital spending plans this year."

Post tech bubble, semiconductor capital spending did a face plant, down 40% in 2001 and another 30% in 2002. The equipment companies responded by diversifying their revenue sources into other segments such as solar equipment and flat panel displays. In past cycles, whenever semiconductor demand slowed or caught a sniffle, pneumonia ensued for their capital equipment suppliers. This time may be different!

Growth has slowed but in fact these businesses have improved. Let's look at the capital intensity of some of the semiconductor companies versus that of their equipment suppliers:

Capital spending as a percentage of revenues (TTM)

Micron (MU) 59.3 %
Intel (INTC) 11.25 %
Advanced Micro (AMD) 24.75 %

Applied Materials(AMAT) 5.08 %
ASML Holding (ASML) 7.46 %
KLA-Tencor (KLAC) 5.09 %
Novellus (NVLS) 2.12 %

With this lower fixed cost intensity, equipment companies should not see their margins crater and improved what Pelayo calls their "cyclical resiliency."

All of the semiconductor capital equipment companies I have cited above generated free cash flow in the last twelve months. Pelayo adds:

"The cash flow generation capabilities have proven much more improved too, with some consistently generating 20% plus returns on operating cash flow, and in some cases similar returns even on a free cash flow basis (including cash outlays for capital spending). All of this positive cash flow just continues to add to the companies' treasure chests of cash, which has resulted in many of the larger players starting to pay dividends and buy back stock. The dividend yields are still less than 2% or so, but I think they have the opportunity to increase over time. So far, equipment suppliers have been really focused on buying back their stocks. Companies like Applied Materials have bought back a tremendous amount and decreased their shares outstanding by as much as 15% or so."

The solar opportunity for these suppliers seems to receive very little attention by investors. Solar manufacturing has many similarities to semiconductor production and presumably, can provide significant fewer growth for these companies. As I suggest earlier, getting knocked down, getting beaten up by too high a reliance on traditional customers helped the semiconductor capital equipment companies approach other niches. Changing course and changing the rules of the game has made them better businesses.

Disclaimer: Neither I, my family, or clients have a position in the securities mentioned in this post with the exception of Intel.