Breaks, Myths And Hyper-Scalers

I’d say investors and traders were back at the desk on Thursday, but that’d be untrue, at best. A lie, at worst. (And the Donald Trump supporters among you better hope there’s a meaningful difference between untruths and lies, because if not, there’s a stooge living in your bathroom mirror.)

Europe was still on break for the most part, and to the extent anyone was working Thursday in the US, they were “working” in Chris Farley air quotes.

As a quick, but important, aside, it’s absurd that we can agree, as a society, to take a break from everything to honor and celebrate what most of us in the West, even the “believers,” understand is little more than a fairy tale, but when the break’s over, we go right back to being willfully (pathologically) disagreeable over issues that are very real, deadly serious and plainly killing us as a body politic and even as a species.

I say that every Christmas. It’s manifestly insane to call a societal truce on an arbitrary day in the name of a purported man-god who lived two millennia ago, then rekindle the acrimony that’s dissolving the social fabric the very next day as if we didn’t just prove to ourselves that amity can be had in something as simple as a folk tale.

Anyway, what can you do? Buy US equities, I guess. Society may be falling apart, but at least you’ll enjoy the hyper-scaler windfall. Have a look at the chart below.

That’s just the S&P for 2024 against global equities excluding US shares. As you can see, the disparity’s vast.

Specifically, Wall Street’s on track to outperform by a stunning 25ppt, the largest margin since 1997, during the dot-com melt-up. That’s apt. After all, many contend we’re witnessing a repeat of the late-90s bonanza complete with the dawn of a new tech epoch and a Fed-engineered soft landing.

Of course, the dot-com boom ended in one of modern history’s most spectacular busts, so here’s hoping “this time’s different” in that regard, at least.

Apropos, Wells Fargo’s Chris Harvey suggested skeptics could be surprised in 2025 by a second leg of the AI trade. “xAI’s Grok 3 and Meta’s Llama 4 LLMs, trained on record 100k GPU clusters [may] trounce previous LLM performance and put the scaling law debate to rest, kicking off the next leg of the AI-trade,” he wrote, previewing “a renewed arms race among hyper-scalers to build even larger training clusters.”


 

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

One thought on “Breaks, Myths And Hyper-Scalers

  1. I owe Chris ?F?a?r?l?e?y? Harvey an apology (or I would if he both knew I existed and cared what I thought, neither of which is the case). In the comments here, I said something to the effect that his briefly-street-highest ’24 EOY S&P forecast of 5600 sounded a lot like ringing a bell at the top. Turns out he was undershooting, and it seems the bell remains unrung.

    To atone, I’ve hired Bennett Brauer to apologize on my behalf: https://www.youtube.com/watch?v=AdkkTV3pIa0

Create a free account or log in

Gain access to read this article

Yes, I would like to receive new content and updates.

10th Anniversary Boutique

Coming Soon