The computer industry has been riding on the premises of Moore’s Law for the past 20 or more years. Moore’s Law states that the processing power of a chip doubles every 18 months as prices decline by approximately 50%. To put this in perspective, during the 2000-03 downturn average selling prices (ASPs) declined 88% and 2007-08 ASPs declined 77% according to the Gartner Group. Intel’s latest chip code-named Tukwila has 2 billion transistors and IBM Power7 chip is 4 times faster than the previous version. At this pace of price decline and enhanced capabilities it is very difficult for most chip manufactures to maintain profitability in a silicon cycle. For example, between 2001-2007 the top 8 DRAM manufacturers had operating profits of $10.4B. However, excluding Samsung operating profit was -$1.6B. This means that only one chipmaker was profitable in the DRAM business over a semiconductor cycle.
Therefore, going forward profitability will be the key driver for consolidation and creating an oligopoly. However, consolidation in the industry is not a new phonomomena. Consolidation started when Intel phased out its DRAM business in 1995 and Texas Instruments sold its DRAM business to Micron in 1998. The next phase of consolidation, will create an oligopoly. In memory – Samsung and Toshiba/Sandisk, Logic – Intel and AMD, Networking – CSCO, Huwai and HPQ; Cell phones – AAPL, RIMM and NOK; Hardware- DELL, IBM and HPQ; Software- ORCL, MSFT and SAP; Semi Cap equipment- AMAT, ASML and LRCX; Foundry – TSMC and STMicro. Tier 2 and 3 companies will have to compete with these giants at lower ASPs. The financial powress of these companies give them the ability to make countercyclical investments (i.e. buy leading edge technologies in a downturn), which position them for profitable growth and market share gain in a recovery.
This lack of profitability has also lead to fewer Fabs being built. Only the Tier 1 companies will be able to afford to build new Fabs going forward. In the 1990’s the approximate cost to build a Fab was $2B. Today it cost roughly $4B to build a Fab. Due to this prohibitive cost there are few Fabs being built. According to AMAT’s management there are only 9 Fabs being built between 2010 and 2011. With fewer Fabs being built this means the oligopoly can control capacity and thus prices.
How does a Tier 2 and 3 company compete in this oligopoly world?
First, Tier 2 and 3 companies should focus on strengthening their financial base and embrace new technologies and products. Second, Tier 2 and 3 companies purchasing leading edge technologies should evaluate the total addressable market and geography. The addressable market must be large – at least $3B. These markets include PC, cell phones, video games. TV, set-top boxes etc (see table below). Also, based on the competitive nature and the structure of these markets the companies’ ability to penetrate these markets is very important. Most companies are not profitable in these markets unless they can attain a 15-25% market share penetration. The choice for geographic penetration is also important. For example, the PC industry is very large in the U.S. However, the market is well served and penetrated. Therefore, the incremental buyer is not in the U.S. but rather in China and India due to low market penetration.
Therefore, for technology investors it is imperative that the focus on consolidation not be displaced with the derailment of Moore’s Law because the industry is going through its natural phase of consolidation based on the dynamics and structure of the industry. In the Technology sector the strong becomes stronger and the weak becomes weaker in the absence of large addressable markets and innovation. This leads to the formation of an oligopoly in various industries that can be very disruptive to others in the food chain.
Written by
Roselia St. Louis
No comments:
Post a Comment