The Risk Of Two Supply Chains

Ever since the Trump administration weaponized trade restrictions against individual companies — first ZTE, then Huawei — China has begun developing a second supply chain for electronics.

Inside of China, this is viewed as a necessary step for survival. In April 2018, the U.S. government banned ZTE from sourcing U.S. components for seven years, nearly putting that company out of business. At the time, ZTE sourced roughly 40% of its components from U.S. companies. That was followed up last year by a ban on U.S. companies doing business with Huawei (which has been at least partially rescinded since then).

China is a huge end market in its own right, and that market is expanding. The country has the financial clout, internal expertise and manufacturing prowess to build its own electronics industry. And with its expansive “One Belt, One Road” plan, the total market opportunity extends well beyond its own political borders.

Still, from a global manufacturing standpoint, redundancy in manufacturing and multi-sourcing is inherently inefficient. While redundancy can provide a buffer against geopolitical issues and what the insurance industry calls acts of God — viruses, fires, earthquakes and floods — it also can cause serious economic problems that are difficult to contain once they get going. And in technology, those problems snowball faster than in other markets.

The dynamics for “techonomics” tend to follow a consistent pattern. Startups flood into a market whenever investors sense a new opportunity. Those companies fight for market dominance, creating a handful of giants that can play off one supplier against another until the field consolidates to the point where that is no longer possible.

There are some notable exceptions to this pattern, of course, at the leading edge of semiconductors. Faced with rising costs of continued node scaling, fewer and fewer companies can afford to compete in critical areas such as lithography (ASML with EUV), memory (SK Hynix and Samsung with HBM), and advanced-node manufacturing (TSMC, Samsung and Intel). While China is building up its memory and manufacturing capabilities, it’s not clear just how competitive those efforts will be given the expense and head start in equipment depreciation.

But China will be much more successful in flexing its muscles back a couple process nodes. It is adding 200mm capacity at a rapid pace. If enough equipment can be found to fill those fabs, then manufacturing capacity will explode across a number of markets that have been constrained in the past. And this is where the dual supply chain begins to have a much bigger impact. With a single supply chain, it is much easier to regulate inventory build-ups, such as the one that shook the global economy following the 2001 dot-com bust. With two supply chains, this is much more difficult.

Controlling inventory is especially important during a time of accelerating technology change, because the shelf life of excess inventory is sharply reduced. The time it takes for a chip to go from leading-edge to obsolete is shrinking, in part because many of the hottest technologies are still being developed, and in part because the end markets that will use those technologies are splintering and evolving. That, in turn, weakens the gray market, which acts as safety net for excess inventory. If the technology evolves too quickly, that safety net collapses.

And this is where things begin to get really ugly. With more fabs and greater competition in a second supply chain, there will be a strong tendency to keep those fabs running. That will lead to overcapacity, undoing decades of efforts to control inventory. By having multiple large economies — the Americas, Japan, Europe and China — all working in the same supply chain, problems in one region are offset by others. With two supply chains, there are no such controls, effectively creating a time bomb that will cause an explosion in inventory.

China’s long reach into new markets, along with new technologies and strong growth in other regions, may provide enough growth potential to keep two supply chains going for a while. But if one of those economic engines falters for any reason, trouble sets in. No economy goes up forever, and that’s where a single supply chain, just-in-time manufacturing and global cooperation suddenly begin to look very attractive again.