Due to recent developments 'm no longer convinced ASML has alpha as benchmarked against other high quality semi conductor names.

    1. DeepSeek's performance

    Writers for the MIT Technological Review recently discussed DeepSeek's performance and were incredibly impressed, putting it only marginally behind U.S. models. While I think this is bullish for AI due to Jevons Paradox, it's bearish for ASML. DeepSeek was largely trained and ran on Chinese chips that were made using only DUV.

    1. The Match Act

    The Match Act is a bill with Bipartisan support that would effectively end all ASML revenue from China if passed. China was roughly 1/3 of ASML revenue the last two years. I assumed ASML would lose most China revenue by 2030, but this would be a much faster time line.

    1. TSM delaying high NA EUV.

    TSM said it won't buy ASML's high NA EUV until 2029 or 2030. ASML still has some customers for it, but this is reduced demand as compared to what investors were expecting.

    Up until this month, I was extremely confident in ASML, and while these headwinds aren't a reason for panic, it has be reevaluating whether or not they should be my second biggest holding right now.

    Should investors be concerned about ASML?
    byu/Designer_Respect4285 instocks



    Posted by Designer_Respect4285

    3 Comments

    1. You are wrong on the fundamentals – Deepseek breakthroughs were trained on Nvidia H800s.

      2.788M Nvidia GPU hours.

      They are trying to shift but not even close.

      You are right on the other two points.

    Leave A Reply
    Share via
    Share via