Last week I talked about the numerous plans to build semiconductor factories. The key word is “plans”. This enthusiasm to spend billions, much of it government money, may not generate the results we imagine.
For example, think back about three years ago. There was little mention of a chip shortage, so supply and demand were probably matched then. Semiconductors were getting better and cheaper, which contributed to low inflation for many consumer goods.
In other words, it you ignored which companies were making profits, and how much, the average consumer was not in a bad place. So the only big problem I see our chipmaking plans solving is to avoid doing business with China. Only, as I’ve tried to show, the behemoths in this market are South Korea and Taiwan, nations friendly to the U.S. and valued trading and geopolitical partners.
While demand for semiconductors doubtless will grow, the manufacturing capacity on the drawing board just in the U.S. could flood the market in a few years, especially during recessionary periods. This has caught the attention of industry leaders and more importantly investors.
Another problem is when you have such concentration in any industry like we have with TSMC and Samsung, they have immense power to simple undercut competitor prices, making recovery of expensive new plant costs very difficult – the Saudi Arabia business plan. Both these companies are planning fabs in the US in anticipation of some type of import restrictions.
Oddly we seem to have forgotten the similarity with Japanese auto manufacturing a few decades ago. If semiconductors follow the same pattern, U.S. companies will have more competition at home, and as auto manufacturers found out, lower pricing power. New U.S. plants will still require critical imports from chipmaking equipment to rare earths like neodymium. It takes a globe to make a chip.
What does this mean for farmers? I can’t find much downside. I wouldn’t bet on all these fabs getting built, and it doesn’t eliminate the complex global supply chain required for those new plants, but it should keep the flow of new technology for our machinery and consumer goods more reliable, relentlessly improving and relatively cheap.
