The honest answer is that most OEMs have moved from surprised to managing but haven't fully made the structural changes needed. Longer term contracts, building buffer inventory, qualifying secondary sources. All of that is happening but more slowly than the urgency of 2021 suggested it should. Quarterly earnings pressure makes it very hard to hold expensive inventory as a hedge.

From our side the OEMs who established long-term volume commitments after 2021 are in a much better position than the ones who went back to transactional buying as soon as capacity eased. The institutional memory in this industry is remarkably short for how painful the lesson was.

The tariff on Chinese-origin components creates a second layer of semiconductor risk that I don't think has been fully priced into most people's planning. A lot of the cost-competitive chips in the market have Chinese origin in their supply chain at some level. Fully mapping that exposure is a multi-year project most companies haven't completed.

The long-term commitment point is right but it cuts both ways. We locked in volume agreements with two OEM customers post-2021. One of them came back 18 months later asking to renegotiate downward because EV volumes missed projections and their overall chip pull was lower than modeled. So now we are holding capacity we committed to on their behalf and absorbing the imbalance. The OEMs who are most disciplined at the procurement level are sometimes the least reliable when their own demand assumptions fall apart.