Like the roller coaster ride that is the semiconductor industry, the SEMI Industry Strategy Symposium (ISS) 2012 had its share of ups, downs, twists, and turns. Semiconductor Equipment and Materials International – better known as SEMI – as the industry association of suppliers to semiconductor manufacturers has held this annual conference in early January for thirty five years to provide updates on business conditions and technology roadmaps to enable SEMI members to plan for the coming year. The conference was packed with senior management paying close attention to the industry leaders, analysts, and customer presenters. All of the presentations, even the most poorly disguised sales pitch or infomercial, contained several valuable insights.
In his keynote presentation “Technology Law Still Delivers“, William Holt (Senior Vice President; General Manager, Technology & Manufacturing Group, Intel Corporation) opened the conference with much optimism based upon the rate of transistor demand growth of 15x in 5 years (from 2005 to 2010 and again from 2010 to 2015). This growth far exceeds the reduction of cost predicted by Moore’s Law. [I calculate a 5.7x reduction = 2^(5 year period / 2 year Moore’s Law cycle).] Even though there is substantial benefit by moving faster, Intel continues to scale to the next technology node every two years, simply because they can’t go any faster. And though we are closer to the limits of scaling we are not close enough to predict when it will end. “Perceived barriers” to scaling will be “surmounted, circumvented, or tunneled through”. Future performance improvements will come from transistor structure, materials, and integration (including new interconnects and packaging beyond current 3D under development).
The mood then turned gloomy in Session 1 – Geo-Economic Trends. Duncan Meldrum (Senior Director, Center for Forecast & Modeling, IHS) in “The Global Outlook in a Post-Financial Crisis World“ predicted slower gross domestic product (GDP) growth of 2.7% world-wide in 2012. US GDP is predicted to grow only 2% while China’s growth will slow to 7.8% and Brazil’s growth will heat up to 3.2%. This translates to a forecasted baseline of weak semiconductor consumption (measured in total wafer area) growth of 1.7% for 2012 (year over year) and 6.0% for 2103 (over 2012). All the potential risk scenarios Duncan discussed were negative, if any are partially true they would lead to even lower consumption than forecasted.
David Townes (Managing Director & Co-Founder, Needham & Company) discussed his outlook of being short term bullish and long term bearish. He reviewed data of long term trends such as the dollar having lost 99% percent of its value over the last 100 years and the average unadjusted real rate of inflation being closer to 8% than 4% as published. (Data shown in chart above.) If the real inflation rate is 8%, the real return of many investments including treasury bills (T-Bills) is negative. This chart and many of the charts presented were from Shadow Government Statistics which publishes data with the political adjustments removed. In terms of debt, he is concerned about the possibility of default leading to an accelerated debasement of the currency. He highly recommends Reinhardt & Rogoff’s book This Time is Different: Eight Centuries of Financial Folly which discusses the debt crisis. In terms of coping, at both a corporate and personal level, he suggests global diversification, holding cash of countries with net exports of raw materials (such as Canada, Australia, and Norway), keeping capital structures simple, owning real (tangible) things including equities (which is the focus of Needham’s business), and maintaining liquidity.
Lastly, in “Economic Outlook” Robert Fry (Senior Economist, DuPont) forecasted GDP growth of 1.9% for 2012 with a possibility of higher growth in 2013 however he wouldn’t take his forecast of 3.1% seriously (yet). His leading indicators including oil prices (he prefers Louisiana Light Sweet since it more closely correlation to both gasoline prices and Brent crude from the North Sea than the oft quoted West Texas Intermediate which measures oil flowing to Oklahoma where there is a glut) points to a recession in Europe and continued slow growth in the US and elsewhere. In previous years, the semiconductor market was not correlated to general industrial production. However now that the semiconductor market has reached maturity the correlation has greatly improved with semiconductors being a slight leading indicator of general industrial production.
Robert had some good news based upon the decline in natural gas prices due to the abundance of shale rock production. This should lead to increased manufacturing in the US since most US chemicals and industrial heating uses natural gas while much of the rest of the world use oil. The lower energy costs of US natural gas should be advantageous for many industries.
His largest concerns are the risk of default in both Europe and China and US future growth. US growth is troubling since outstanding government debt is close to 100% of GDP and Reinhardt & Rogoff’s analysis shows that growth slows significantly when this ratio exceeds 90%. If an economy is growing at 4% or higher then shocks (such as defaults) can be managed. If an economy is growing at 1% or less these shocks can lead to a recession. What isn’t clear is what happens between 1% and 4% growth. The last major risk to Robert’s forecast is the quality of China’s reported economic data since he doesn’t always believe the released data and he has not found any clear leading indicators.
Following these sets of gloomy forecasts, the next session provided much needed positive news…. I will post additional summaries shortly.