One defining characteristic of the Hardware sector is that it has historically been very cyclical. Despite this, long term growth has been solid - the semiconductor industry booked a 17% annual growth rate from 1975 to 2000. Looking ahead, two potential changes are in store for this sector.
The most populous industry was Semiconductor Equipment. This consists of firms that supply the equipment needed to build the microprocessors and integrated circuits that are the fundamental building blocks of electronics. Stocks here ranged from industry behemoth Applied Materials (AMAT), to specialty supplier KLA-Tencor (KLAC), down to small niche firms like FormFactor (FORM).
This is a fairly attractive industry for investment. Technology continues to advance. As chip sizes get smaller and density of transistors gets higher, these firms benefit from constant equipment upgrades from semi manufacturers like Intel (INTC). Since semiconductor manufacturing is such a complex and multi-step process, many firms benefit from being the specialist provider dedicated to a certain phase of the process. However, do your homework. This constant change also creates ample opportunity for new entries to introduce better technology to take business from an established player. Look for firms in this sector that supply equipment to all the big chip makers, and have a history of maintaining their competitive position. These companies gain the trust of their clients and usually have a leg up winning new business contracts.
These are the companies that actually build and/or design the integrated circuits that power today's electronics. In this industry, the Magic Formula dug up companies like NVIDIA (NVDA), Silicon Image (SIMG), and Texas Instruments (TXN).
It's rather difficult for semiconductor manufacturers to build lasting moats. Competitive advantage in this industry is mainly related to having better technology than the other guy, and unfortunately this is one of the least durable competitive advantages. If a competitor comes along with better features or a cheaper design, sales can go down the drain in a hurry. Look for a diverse line of products and customers and a low debt balance sheet. The semiconductor history books are littered with one hit wonders that had no act two (to wit: PortalPlayer with it's Apple (AAPL) iPod contract).
Electronic equipment is a tough business. First of all, competition is rife and cutthroat. Both industries are filled with dozens of well run companies all competing to build devices with more features at lower prices. This limits the pricing power for products, while at the same time requiring continuous research & development expenditures, making it very difficult to maintain high profit margins. Secondly, product features and design are constantly changing, so that today's hot product is quickly replaced with another one with better design, smaller size, or bigger feature set. Clearly, building a moat in this industry is extremely difficult.
The primary way to build competitive advantage in this business is to become a low price leader. Dell in the 1990's and early 2000's was a good example of this, but their vaunted business model has been successfully copied by several competitors, mitigating Dell's advantage today. Apple has had success by consistently beating the competition in design and features, but historically no company has been able to maintain a long term competitive advantage in this manner. How long Apple's current success can be maintained remains to be seen (although the iPhone does appear to be years ahead of the competition).
Data networking firms build equipment that is used to build and control data networks. Products are devices such as switches and routers. Stocks here include NETGEAR (NTGR) and Foundry Networks (FDRY) The major player in this industry is Cisco Systems (CSCO).
Realistically, this segment really doesn't differ all that much from wireless and computer equipment. Data network protocols and interfaces are very standardized and widely accepted, so locking customers in with a proprietary system is not a possibility. This means that all firms selling data networking products are competing on price primarily, and features secondarily. And as always in a commodity business, only the low cost provider truly has a durable competitive moat. I