How to Profit from the End of the AI Trade

When a massive stock market boom starts to slow down, it rarely crashes overnight. 

What I am seeing now is not a crash, but the signs that a slow down is starting to take hold. Given the outsized weightings of artificial intelligence and semiconductor stocks in the benchmark indexes, the broader market is only as good as the AI trade. 

How good, exactly, that AI trade is now is up for debate. 

Why Are AI Stocks Falling?

To make money from a market slowdown, you first need to understand why booms end. It is rarely because the technology fails. Instead, it is usually because the market gets completely flooded with too much supply, of chips in this case. 

When the top chip companies start raising tens of billions of dollars by selling massive amounts of new stock to the public, it’s a sign. When company insiders use their ultra-high stock prices to gather cash and build massive new factories, they openly tell us a supply glut is coming. 

Once those factories open, chip prices will drop, and the massive profit margins that Wall Street fell in love with will shrink. That’s how the bear case could play out. 

Who’s Buying the Dips Here?

This is often when big Wall Street players quietly start selling their shares to lock in profits, while regular investors keep buying the daily dips. If you think the massive AI tech boom is finally running out of steam, you don’t just have to sit there and watch your portfolio take a hit. You can actively prepare for the slowdown that is very much underway. The only question is whether it is a “pause that refreshes” or the start of a plunge that besets semiconductors. 

This chart shows the past three years for the iShares Semiconductor ETF (SOXX), the biggest semiconductor ETF. I’ve marked with arrows the last pair of 30%-35% declines. 

I’ve also marked on the right a rough price range where SOXX would have to go for us to again be talking about a one-third decline. With the percentage price oscillator (PPO) indicator at bottom perched just on top of the zero line, the 20-day moving average having rolled over, and the 50-day moving average about to follow it, I would not bet against a much steeper dive here. Albeit with the obligatory giant bounces along the way. 

Like I said, you don’t have to just sit there and take it. There are inverse ETFs on SOXX, like the Direxion Daily Semiconductor Bear 3X Shares ETF (SOXS), which I’ve shown in this table below, alongside four semiconductor industry ETFs. Now, SOXS is a 3x inverse ETF, so you have to be very careful. 

But the list of ways to “short” semis has burst wide open this year. Not only can we try to profit from the stock prices declining across the board in that industry, we can do so at the stock-specific level.

Are All Semiconductor ETFs the Same?

There’s overlap across several of the top semiconductor ETFs given the concentration in the chip industry. Here’s a quick survey of some of these funds’ top holdings. I show this not only so you can understand what’s in them, but also as a “hit list” for considering being stock-specific, as a bull or bear, as this trade plays out. 

SOXX is very crowded at the top, with about one-third of its assets in the four names I circled here. Still, there’s some spread to another 10-15 names that each have some individual impact. 

This is the VanEck Semiconductor ETF (SMH), the original ETF in this market segment. It is purely capitalization-weighted, so Nvidia (NVDA) stands tall at the top. There’s a lot of overlap between SMH and SOXX, as you can see.

The same can be said about the Invesco Semiconductors ETF (PSI), except that its creator limited the ETF to 30 stocks. It is far from equal-weighted, but it is fairly diverse. 

We can get more granular via ETFs like the Roundhill Memory ETF (DRAM) and the Tema Memory ETF (DISK). DISK leans heavily on just Sandisk (SNDK), Kioxia, and SK Hynix (SKHY). 

DRAM holds a much smaller allocation to SNDK, with focus on Samsung, Micron (MU), and SKHY.  

To me, analyzing these holdings helps me figure out where it makes sense to short via inverse single stock ETFs. But for those looking for a quick and dirty solution with some leverage, SOXS is the most liquid choice. 

And that’s a chart with some massive upside potential. Albeit with big risk attached. Not only due to the possibility that chip stocks do not fall further out of favor, but because inverse ETFs like this one, especially with three times leverage, can work against you quickly. Watch out if you consider this route.

The Bottom Line on the Semiconductor Trade

This is not a “go big or go home” type of trade. It is based in large part on the natural cyclicality of markets. And the simple belief that markets of today get hyper-overvalued. That sets them up for subsequent big drops. 

Rob Isbitts is a semi-retired CIO, former fiduciary investment advisor, and Barchart columnist. Check out his other work at ETFYourself.com (featuring the Fresh Charts weekly trading post), and ROAR.PiTrade.com, helping investors to better-manage their own portfolios.