Executive Summary

• The Trump administration is reportedly considering a 50% self-sufficiency target for US semiconductor needs as it nears the conclusion of its Section 232 investigation; this is significantly more aggressive than prior commentary suggesting that merely building in the US would exempt companies from anticipated sectoral tariffs of 100%.

• The 50% self-sufficiency target seems feasible assuming a ~5 year time horizon, mostly on the back of TSMC Arizona’s 6-fab campus buildout; we estimate that ~10% of TSMC’s total output and 20-30% of its leading-edge production could be supported by the Arizona project, though timelines will likely need to be accelerated.

• Our base case assumption is that there will be a multi-year grace period as the industry ramps up to the 50% self-sufficiency target, but customers will feel pressured to book domestic capacity as soon as it becomes available.

• This development is incrementally negative for the overall tech sector (especially leading-edge fabless vendors and hardware OEMs) on higher production costs and potential tariffs; semi fabrication companies with large US footprints such as Intel and Texas Instruments could stand to benefit.