Rotation, Reset, and the Next Better Pitch

Last week, my focus was on biotech, healthcare, healthmaxxing, and what I think could become the next major second-order AI theme.

The simple version was this:

AI hardware had already become the obvious trade. Semis, memory, AI infrastructure, and anything attached to the buildout had been where the market’s attention was sitting. But under the surface, biotech and healthcare names were starting to repair. Moderna had already reclaimed important levels, volume was improving across the group, and a lot of these names were starting to act like the market was finally paying attention again.

That call worked.

A lot of the biotech and healthcare names continued higher. MRNA, XBI, IBB, and several of the more speculative bio names kept catching flow and momentum. That is what you want to see when a theme is real. Not one isolated name moving on one headline, but a group starting to confirm itself through price, volume, and relative strength.

But this is also where the game changes a bit since the ideal entry into biotech and healthcare was not after everyone realized the group was acting better. The ideal entry was two or three weeks ago when the charts were quietly climbing and nobody really cared yet.

That does not mean the trade is over. It means the easy entry is behind us for now.

From here, I do not want to chase every green biotech candle just because the theme is finally obvious. The better plan is to let the new uptrend prove itself, identify the names that hold up best on weakness, and wait for the first controlled pullback toward short-term moving averages.

The market gave us the heads-up. Now we need the market to give us a better pitch.

There has been a lot of discussion around the rotation.

Some of it is real. Healthcare has acted better. Biotech has acted better. Robotics had a brief pump. Value and Dow-type names caught a bid while high-beta AI and semi names got hit. If you were only watching the headline index narrative, it looked like everything was fine. The Dow was making record highs and the “market” looked strong.

But if you were watching QQQ, SMH, SOXX, DRAM, memory names, and the more crowded AI momentum basket, the tape looked very different.

That is an important distinction.

Index-level strength can be deceptive. Sometimes the market is not broadly strong. Sometimes it is just hiding the damage by rotating into a different sleeve. That is basically what happened into the holiday break. The headline looked fine, but crowded tech, AI, semis, and memory saw real selling.

That does not automatically mean the AI trade is dead.

It means the first easy phase of the AI trade is probably over, and the market is forcing people to separate the real leaders from the tourists.

This is usually where everyone makes the same mistake in opposite directions.

When AI names are going straight up, everyone thinks they can never go down. When they finally get hit, everyone wants to declare the whole trade dead.

Both reactions are lazy.

The more useful view is that the trade became crowded, levered, and extended. Then it got hit at quarter-end, right as rotation/rebalancing pressure started showing up, and right as a lot of people were probably leaning the same way across the same names.

That creates damage, but it can also create opportunity.

Sentiment around AI momentum names has turned very bearish very quickly.

A lot of semi and memory charts are technically damaged now. There is no need to pretend otherwise. Some of these names broke short-term structure, lost key moving averages, and printed ugly high-volume selling. When that happens after a major run, you do not just blindly buy the first red candle and assume everything is fine.

But this is also where short-term mean reversion becomes possible.

When a group goes from loved to hated in two sessions, especially after a massive run, you have to start looking for where the sellers might exhaust. Not because the chart is perfect, but because the positioning may have reset enough to create a tradable bounce.

I think we need to seriously consider that the selloff may have been more of an end-of-quarter and early-July rebalance event than some clean fundamental death blow to AI.

A lot of these names have been crowded and hyper levered. They were owned through single names, ETFs, options, 3x products, and probably by a lot of funds leaning the same way across their books.

Once that starts to unwind, it can look dramatic fast.

You also have the cross-market piece now. The AI/semi trade is not just a U.S. trade anymore. It has become tied into Asia, especially Korea and Japan, because of memory, hardware, supply chain, and global tech leadership. So when the unwind starts, it can become broader than just a few U.S. tickers.

That is why the next couple of weeks are interesting.

Not because the charts are all clean. They are not.

They are interesting because a real flush in a leadership group can create the next better long opportunity if price starts stabilizing where it should.

For semis, I would be watching the major ETFs and leaders closely.

SOXX, SMH, SOX, DRAM, and the large-cap semi/memory names are the cleaner places to track the group. I am less interested in trying to buy the weakest names just because they are down the most. Weak names can stay weak. Broken charts can stay broken.

The better trade is usually in the names that wash out, hold important areas, and then show demand again before the rest of the group repairs.

That is the difference between buying damage and buying a reset.

If we get an undercut, sweep, and reclaim around important short-term support, that is where I become more interested. I would be looking for things like: