Hey Traders,
Wednesday, October 7th, was a lousy day to own a normal stock.
The equal-weight S&P 500 closed that day at its worst level against the regular index since the equal-weight fund launched back in 2003.
The S&P was up 14% on the year.
The average stock in it was up about 10%, and a whole lot of those were up nothing.
If you held a sensible mix of companies that make things people actually use, you spent the last few weeks of September watching Nvidia, Microsoft and Meta run off without you.
On that same Wednesday, a data-center builder called Applied Digital (APLD) held its earnings call, and an analyst asked management a question every APLD holder should be chewing on. I’ll get to it.
What happened over the next ten days told me which side of the AI trade I want to be on right now, and it sounds a lot like a 1989 Soul II Soul record.
Back to Life (RSP)
You remember the kid who always got picked last for kickball.
For about three years, the average S&P stock has been that kid.
RSP is the equal-weight S&P 500. It holds the same 500 companies, except Nvidia gets the same slice as a regional bank or a mid-sized industrial. You end up owning the rest of the economy, you know, the “real stuff” you and I eat and wear while we ponder the future of an AI-driven world.
For a long stretch, owning it felt like a punishment. Goldman says breadth is the weakest it’s been since the dot-com bubble, and a basket of 42 AI stocks has produced about two-thirds of the S&P’s price gain since the start of 2024. Everybody else sat on the bench.
Then October 7th came and went. Over the following week, the 10-year backed off from north of 5.2% to right around 5%, the chip stocks cooled, and RSP started outrunning the semis.
A couple of weeks doesn’t make a trend, and I’m not calling one yet. It does fit what I’ve expected for a while. I think the narrow trade starts to reverse.
Not violently… but it turns.
The AI builders keep taking heat on whether the data centers get built on schedule, where the power comes from, and how efficient all that spending really is. The questions keep piling up on the guys writing the checks.
If you look out to the horizon, I’ve always liked the users better than the builders anyway. I’m talking about companies that take cheap and cheerful AI off the shelf and let it quietly juice their productivity. That benefit should start showing up across the rest of the index, well beyond the handful of names everyone already owns.
RSP is how you own that rotation. When breadth comes back, equal weight is where I want to be sitting. One to love.
Back to Reality (APLD)
Back to October 7th…
Applied Digital reported a wider loss as it poured more money into construction. Then an analyst on the call described the bond market as “challenging to say the least” and asked management where future funding comes from.
That question sits at the center of this whole company. Applied Digital builds AI data centers, and it borrows to do it. It ended August with $6.4 billion of debt against $2.9 billion of cash, and the debt runs to about 94% of the company’s market value. At the pace of the past year, its own cash couldn’t fund twelve months of building.
We’ve flagged this name more than once, and that debt is why.
Start with rates. Yields pulled back a touch this week, sure. The 10-year still sits around 5%, and I don’t think sticky is done. The AI buildout keeps steady upward pressure on rates, and that pressure won’t evaporate on one soft print.
If yields stay elevated, money gets picky. Investors stop paying up for the story and start digging into the balance sheet. They want strong operating cash flow. They want a company that can fund its own AI buildout out of pocket, rather than borrowing every dollar to get there.
Management will tell you more than 80% of the debt doesn’t come due until fiscal 2031 or later, and that’s fair.
A lending pause doesn’t need a maturity wall to hurt, though. It just slows the building, and the building is the whole story here. With that build getting questioned on power, cost and efficiency, I can’t think of a shakier place to stand than a heavily indebted pure-play builder without enough operating cash flow to ride out a rough stretch.
When the market narrows down to the highest-quality names, APLD gets left outside.
I still like the AI theme. I just don’t want to own it through a balance sheet like this with rates where they are. One to leave.
So that’s where October 7th left me.
The kid who got picked last is coming back to life, and APLD’s own earnings call gave the builders who borrow every dollar their back-to-reality moment.
Over the next few weeks, I’m watching whether RSP keeps beating the chip stocks and where the 10-year settles. If both keep moving the way they did the first half of October, I like where these two calls sit.
But you gotta wonder, can my “one to love one to leave” picks just keep becoming the Ticker Highlight Show’s trade of the week?
I mean listen, the last one I chose, just last week, closed up 118% a day after getting into it.
Perhaps I’m biased, but I think these two choices have the same kind of promise.
That said, the other traders have chosen some solid plays to fight over Monday.
Here for a good time… AND a long time,
Hans