Semiconductors – China and subsidies

Micron – Collateral damage

• In an attempt to overturn the ban that China has placed on its chips, Micron has committed to invest $602m in its chip-packaging plant in China in a move that I suspect is not going to change very much.
• I have previously argued (see here) that China’s move to ban Micron is a sign of just how weak China’s position is when it comes to semiconductors.
• This is because if China really wanted to hurt the USA it could block the sale of Apple products in China or the use of Qualcomm and Nvidia chips.
• The problem here is that this would hurt China just as much, if not more in some cases and so it went with an option where it could easily source the same product from somewhere else.
• Hence, I suspect that this move by Micron (which could be one it was going to make anyway) is unlikely to sway the Chinese regulators very much and so Micron chips are likely to remain blocked.
• I think that it will take far more than this to change the position of Chinese regulators and this is extremely unlikely as regulations are getting tighter, not looser at the moment.
• Hence, the outlook remains difficult for Micron, especially in this environment of weak demand and high inventory and I see no reason to pick this up.

Intel – About face.

• A week is a very long time in politics as the German government appears to have done an about-face and will now increase the subsidies it is offering to Intel from $7.2bn to $11bn.
• Intel makes a reasonably good case for further subsidies as there are far more on offer in Asia and inflation will have increased the build cost meaningfully from when it was first announced.
• The problem that all of these companies face when building leading-edge fabs in Europe and the USA is that they will really struggle to make them economically competitive with those built in Taiwan and South Korea.
• This means that customers are going to have to pay higher prices for the silicon chips that they buy with the trade-off being that they will be resistant to supply shocks related to geopolitical events in China’s backyard.
• I think that the current economic environment is likely to cause a slowdown in fab builds as there is not enough demand in the market (outside of AI training chips) to absorb a lot more production capacity without an economic recovery.
• I am not expecting this much before mid-2024 and so I suspect that we will see further delays of leading-edge fabs in Europe and the USA which make life easier for Intel which is in a uniquely difficult position at the moment.
• I continue to think that the risk-reward trade-off of Intel vs TSMC or MediaTek or Qualcomm is sub-optimal and would need to see Intel a lot lower before considering getting involved.