Brooks Automation: A Chipmaking Gem

This semiconductor-equipment maker is in a prime spot to cash in on the changeover to more advanced wafer-fabrication plants

By Richard Tortoriello

Semiconductor end-markets are seeing many signs of improvement, including higher memory-chip prices, increased wafer output, and high capacity utilization for leading-edge chip technologies. One small-cap chip-equipment maker that stands to benefit from this rebound is Brooks Automation (BRKS ).

The Chelmsford (Mass.)-based company competes in the rapidly growing market for semiconductor factory-automation hardware and software. Brooks started out as a supplier of robotic arms used to feed silicon wafers into and out of vacuum process chambers (many steps in chipmaking are done in a vacuum). In 1998, Brooks diversified through a series of acquisitions, adding factory-automation software to its product line.

More recently, the company agreed to acquire PRI Automation (PRIA ), a maker of factory-automation systems. Whereas Brooks focuses on automating the tools that make semiconductors and providing the interface between the tool and the factory floor, PRI is the leading provider of transport equipment used to move wafers from one tool to another within the factory. In addition, PRI makes wafer stockers, used to store wafers within the factory, and it's the leading provider of "atmospheric robots," which operate outside a vacuum. The PRI acquisition, valued at about $380 million at the time of the announcement, is expected to close in April.


  The Brooks-PRI combination will offer customers a complete suite of wafer-fabrication automation hardware, something competitors can't duplicate right now. In addition, Brooks is the only company that provides a complete, integrated line of factory software. S&P believes this lineup gives Brooks a competitive advantage that will enable it to keep gaining market share.

Brooks is experienced at acquisitions and should be able to effectively integrate PRI, increase efficiency there, and improve PRI's gross profit margins, which have lagged behind Brooks's over the past three years by a 15% margin.

Brooks divides its business into three segments. Its tool-automation division -- which represented 44% of revenues in fiscal 2001, ended in September -- includes vacuum and atmospheric robots, integrated tool front ends, and complete tool-automation systems. In 2001, Brooks was No. 1 in the market. S&P expects this business to continue growing at the historical average rate for wafer-fab equipment of about 18% to 20% annually.


  The second segment is factory automation, which contributed 26% of fiscal 2001 revenues. It primarily consists of modules used to move wafers from a factory floor into a processing tool. The two most common modules are "standard mechanical interface facilities" and "front-opening uniform pods." SMIF is the standard for 200-mm (8-inch) wafers, while FOUP is the standard for new 300-mm (12-inch) wafers. With the September, 1999, acquisition of Infab, Brooks acquired SMIF and FOUP technologies that allow it to compete in each of these markets. S&P expects factory-automation equipment to grow faster than the industry average, as factories become increasingly automated.

Lastly, the factory-software division, which contributed 30% of last year's revenues, includes manufacturing software that focuses on reducing process cycle times and improving quality; planning and logistics; and engineering equipment that monitors equipment performance. Brooks holds the leading position in this market and is the only company that supplies a complete range of software applications. We believe factory software holds significant growth potential over the long term, as chipmakers begin to see the efficiencies that can be realized by automating feedback and control of all processes throughout the factory.

Market researcher Dataquest estimates that sales of factory-automation equipment will grow at a compounded annual rate of 24% from 2001 to 2004, vs. a 20% rate for wafer-fab equipment in general. The primary force behind this growth will be the conversion from 200-mm to 300-mm wafer fabs. Because cassettes of 300-mm wafers are too heavy for humans to move and significantly more expensive than their 200-mm counterparts, S&P expects 300-mm fabs to be fully automated. With 300-mm fabs costing up to $2.5 billion, Brooks sees its per-fab revenue opportunity ranging from $150 million to $200 million, vs. $50 million in 200-mm fabs.


  The chip downturn's severity has delayed the buildout of 300-mm factories, but S&P thinks it will pick up. Taiwan Semiconductor's recent announcement that it will raise 2002 capital spending by $1 billion, to $2.6 billion -- most of which will be used for 300-mm factories -- is evidence of this trend. Brooks currently expects chipmakers to build eight new 300-mm fabs and expand five fabs in 2002 (mainly later in the year). S&P foresees an additional 13 new fabs in 2003.

Also, the combination of automation hardware and process-control software will allow chipmakers to manage their factories more efficiently and reduce operating costs. In the future, we expect factory-automation software to become an increasing necessity for fabs.

Brooks has a variety of Japanese and U.S. competitors, but few offer the same range of products and degree of integration. Asyst Technologies (ASYT ) has a strong position in SMIF and FOUP factory interfaces and atmospheric robots. Japanese companies Daifuku, Murata Machinery, and Shinko compete with PRI in the transport market.


  S&P projects a loss per share, before goodwill amortization and acquisition-related costs, of $0.96 for fiscal year 2002, followed by a return to profitability in fiscal year 2003, with earnings per share of $1.23. Peak earnings during the last cycle were $1.94 per share (calendar year 2000). Given the favorable market trends and an improved competitive position, which will likely lead to market-share gains, S&P figures Brooks can double its peak earnings during this market cycle.

Brooks shares now sell at 2.7 times sales and 2 times book value, below historical peak valuations of 4 to 5 times sales and 6 times book value. S&P has set a 6- to 12-month target price of $62, or three times our fiscal 2003 sales estimate, a nearly 40% jump from the current price of about $45.

Analyst Tortoriello covers semiconductor capital-equipment stocks for Standard & Poor's

Before it's here, it's on the Bloomberg Terminal.