Caught in a Global Labyrinth
Yesterday brought a reality check for Trump Administration generals, 5G microwave warriors, and tech strategists who believe that they can exploit the US semiconductor technology lead to bring down China.
Microchips provide a silicon backbone and central nervous system for the world economy. Everything ultimately depends on them — from finance, to medicine, to transportation, to defense. Over the last some thirty years, chip revenues have almost tripled as a share of global GDP.
People talk about 5G (fifth-generation wireless standards). Semiconductors pervade every generation of every new technology that entails sensing, measuring, storing, sending, receiving, accessing and processing information. From biotech, to agriculture, to automobiles, to aerospace, can you think of anything those functions don’t cover?
Everybody agrees that in crucial ways, the US is the global leader, with nearly 50% market share, compared to China’s roughly 3.5% in 2018. In Information Technology infrastructure, the US commands a 59% share. In military microchips, the US has an estimated 52% share.
Let the Asians lead in consumer, autos, and smart phones.
With a mercantilist crunch, so it would seem, the US could stop the Chinese ascent through microchip bans and boycotts, “entities lists” of verboten Chinese companies and garottes of crucial tools and technologies.
Thus, the Trump Administration is currently banning sales to Chinese facial recognition device companies and purchases from Chinese telecom equipment companies.
No Fly Zone
Huawei and ZTE get out of here, and no Intel or Nvidia chips for SenseTime and HikVision. Should eventually stop them in their tracks, right?
But the title tells the story: “How Restrictions to Trade with China Could End US Leadership in Semiconductors.” Written in conjunction with the Boston Consulting Group, the Semiconductor Industry Association (SIA) paper confirms that the US leads in 23 out of 32 categories of chips, from personal computer devices, to industrial, to information technology infrastructure, and was actually gaining share until 2018.
Over the last five years, US microchip company shares have produced an annual shareholder return of 14%, roughly four points above the S&P average. US semiconductor industry gross margins stand at 62%, 11 points higher than the rest of the world. With higher returns, the US industry has led the world in R&D with 17% of sales, twice the rest of the world.
Sounds imperial to me, and it is. The US position in the world largely stems from our bastion in microchips and other frontier technologies. We lose this bastion and we will eventually lose our world leadership.
In microchip trade, our microchip bastion gives us a massive surplus. Peter Navarro, Trump trade strategist, should relish it. In his world, a trade deficit confers direct leverage. President Trump has treated our trade gap as a direct loss to the country.
The Administration’s mercantilist model is a zero-sum world where any deviation from trade balance means a loss for the party with the deficit. In the case of China, we were running a deficit near $250 billion in 2018, which represented a massive sluice of US wealth into China.
We design the chips and own them, but most of the time we don’t manufacture the devices that use them. Less than 25% of global semiconductor demand stems from the US domestic market.
Nearly 80% of US microchip revenues come from export markets. In 2018, the Chinese market comprised 58% of global demand for semiconductors and 23% of demand for US chips. Chinese chip suppliers can provide only 14% of the needs of Chinese manufacturers. From the mercantilist perspective, these numbers give the US a tremendous position of power over China.
However, the logic of trade is hostile to all such exploitation. Fabrication of a microchip entails some 1500 exquisitely calibrated steps in some $16 billion plant bringing together complex equipment from around the globe.
It takes decades to establish these supply chains. The Chinese may be paladins at building cities and hospitals and are becoming brilliant at designing chips on computers. In value, they design 14% of the world’s microchips. But they have yet to master the manifold maze of rigorous processes and techniques necessary on the ever-changing frontiers of chip manufacture.
So, the Chinese are caught in this global labyrinth. They are trying to escape, but have yet to succeed. They can issue as many papers as they want, targeting 70% share by 2030, and print as much debt as they need to finance the effort. But they will not change this global technology industry into an autarkic fortress.
However, it is not China, but the United States that imagines that it can convert this global asset into an American redoubt. The US with its dominance would seem to have a better chance. That is where the BCG paper offers a crucial corrective.
Over the last two years of trade war, the US has already started to lose its lead. Since the trade war began, the median year to year growth of the top 25 US chip companies has dropped 90% — from some 10% to just over 1%. After restrictions on sales of chips to Huawei, the top US producers suffered revenue declines of between 4-9%.
The authors of the BCG report, Antonio Varas from the Silicon Valley office and Raj Varadarajan from Dallas, project the effects on the US semiconductor industry of an actual decoupling of the US and Chinese technology sectors. UScompanies, such as Intel, Broadcom and Nvidia would lose 18 percentage points of global share and 37% of their revenues.
These revenue drops, write the BCG team, would force severe cuts in R&D and capital expenditures and the loss of scores of thousands of highly skilled jobs in the US microchip industry alone. These jobs are the foundation for our national defense and prosperity.
But perhaps the warrior strategists can shrug off these losses if they inflict comparable damage on China. But China has alternative possible suppliers in Europe and Asia for perhaps 70% of these chips. Over the last two years, China has begun gaining share while the US loses it.
And if US boycotting actually crippled Chinese progress, they could redress the damage with merely one belligerent step. They could recover Taiwan, a country claimed by China and revealingly absent from the pages of the BCG-SIA report.
Even though the US dominates market share in chip sales, the center of global semiconductor fabrication technology is in Taiwan. Contrary to the assumptions of some strategists, Taiwan is not an island off the coast of California.
In future prophecies, I will explore the implications for investors of upsetting the delicate balance of policies surrounding Taiwan and pushing China toward recapture of this beleaguered island.
Editor, Gilder’s Daily Prophecy