Big Numbers – The Semiconductor Supply Chain

…To make sense of the big picture, one needs to follow the money and then head to China.

Ed Pausa the primary author of PricewaterhouseCooper’s (PwC) recently published report “Continued Growth: China’s Impact on the Semiconductor Industry – 2011 Update” provided an overview at this month’s MEPTEC luncheon. His presentation was a helpful tour to start digesting this impressive report, now it its seventh annual update. The report runs 112 pages in length and is packed with figures, data and most importantly analysis. Building a cohesive picture from many disparate data sources is a major undertaking and PwC should be applauded for making available this excellent work.

After listening to this presentation and reading the report, I find two items that really stand out as primary market forces. Unraveling the convoluted web of the semiconductor supply chain to examine these items will lead to greater understanding of the industry. They are, for lack of better terms, “end consumption” and “self-sufficiency“. A semiconductor device is often manufactured in country X, then shipped to country Y for assembly into a product at which point it is “consumed”. The product then may be shipped to country Z for use by the end user (customer or business). Market data tracks a semiconductor up to the point of consumption (in country Y) whereas the “end consumption” is really in country Z. It is the market trends in country Z that flow through the entire supply chain. “Self-sufficiency” is the proportion of a country’s demand which is satisfied by it’s domestic semiconductor manufacturing.

That Asia Pacific dominates consumption of semiconductors in products (both for domestic use and export) with $160 billion of the $298 billion (2010 figures / 2011 will be similar) market total should not be a surprise to those who follow the World Semiconductor Trade Statistics. However, the graph shared by Ed (shown above, from page 11 of the PwC report) which splits the data out differently shows that Asia Pacific consumption isn’t really the Four Asian Tigers of Hong Kong, Singapore, Taiwan, and Korea. In 2010, China was 40.5% of the market (PRC not including Hong Kong) at $121 B. When China is separated out of the statistics for Asia Pacific, both the size and growth rate of the Chinese market is staggering.

Even with comparable (to same order of magnitude) gross domestic products (GDPs) of $10 trillion for China versus $14.7 trillion for the US doesn’t explain China having twice the market share for semiconductor sales as all the Americas (North, South, and Central) – 18% versus 40.5% for China. Furthermore, with a GDP per person (PPP) of $7,600 versus $47,200 for the US it is hard to believe that per person spending on products with semiconductors is on the same level and certainly not 2x. This is why understanding the supply chain to determine the point of “end consumption” is necessary. The consumption discrepancy is a result of 65% of the semiconductors used in China are either exported or assembled into products which are exported. (Graph to right from PwC page 25.)

The remaining $46.2 B of semiconductors is used to satisfy Chinese end consumption. This is significant demand when compared to $54 B of the Americas’ total demand (including semiconductors used in products domestically and exported). Unfortunately, this is where the trail of data runs cold. There are statistics about sales by company headquarter location but no readily available numbers for production or end consumption by country (other than the PwC data for China). After the presentation, Ed also indicated they were not aware of this type of data.

Using statistics from the US Census Bureau we find that the US was a net exporter of $18 B of semiconductors in 2010. And even though US semiconductor companies (based upon headquarter location) are estimated to produce half the world’s semiconductors (by value), without additional data it is hard to determine how many semiconductors are produced or end consumed in the US preventing direct comparisons with other markets such as China. Without solid data it is hard to determine which end markets really drive semiconductor demand.

In discussing overall China market growth with $108.6 B of consumption and only $21.3 B of production in 2010, Ed highlighted this $85.3 B gap. In terms of “self-sufficiency”, China can only meet 46% of its own end demand and just 20% of their total demand which enables the export of high technology products. PwC predicts these gaps will remain large (shown in the figure to the right from PwC page 94) with overall market (consumption) predicted to have compounded annual growth rate (CAGR) of 11.6% while the increase of capacity CAGR is estimated to be in the range of 10-18% depending on scenario. Both growth of consumption and production will exceed the world-wide average.

I predict that just as countries desire energy independence, China will attempt to increase its “self-sufficiency” to meet its domestic demand for semiconductors for both “end consumption” and export. It will be interesting to see how quickly they invest in production capability and technology to close this gap. With market numbers like these, China is the biggest market and will be the center of the action as they improve upgrade their technology capabilities to produce advanced semiconductors. Additional growth is likely as their PPP (GDP per capita) increases and their consumers have more to spend on electronics. Strategies that include China are essential for long term success in the semiconductor supply chain.

%d bloggers like this: