Thursday the market fell like a rock, especially the Qs. I’m not here to say the world is ending, and I’m not here to say we’re ripping back to the moon. I just want to walk through the data.
This was a tech and AI selloff. The Qs fell out of bed, semiconductors got hit, and everything else got pulled down with them. Equal-weight large caps had a bad day too, but the monthly chart didn’t do lasting damage. I recorded this Friday in the pre-market, with the Qs gapping up about 1.5%.
A 4.5% drop on the S&P 500 in one session is rare, and even rarer above the 200-day moving average. Moves like this usually happen below it, when you’re already in a bear market. This one felt like a rug pull.
It was the 45th-biggest down day since 1950, and only the seventh 4% down day that still closed above the 200-day. We’d also gone about 110 days without a 2% move in SPY. This session was about six and a half times a normal day.
The theme is simple: rarely the low, and rarely the top.
One week later, it’s a coin flip. Further out — 60 days, 120 days, and beyond — the market has done much better. Of the other six times this happened from above the 200-day, almost all were higher about six months later, except 1989. It was rarely a V-bottom. Usually you got some continuation, then a month or so of chop, then we worked our way back up.
In 19 out of 20 cases, there was a lower close than the crash day. That usually wasn’t another crash — more of a bounce that failed. On average it took about six weeks to get back to the post-election high. Buying the panic has generally been better than shorting into it, though that’s not a system I would trade myself.
The hedge-fund blow-up talk is still unconfirmed. The concentrated selling gives it a little credence: tech was down huge, while other sectors were down but not world-ending. If this is a real shift in AI sentiment, it may take months to repair. If a fund or two blew up and that gets cleaned up, it can rebound much faster. I don’t know which one this is, and I don’t need to. We’ll keep trading our systems.
If someone told you they called this, where were they on Wednesday? The data says this is likely not the best buy right now, and also not the best short. If you don’t have a system for tape like this, sit back for a couple of weeks and let the market digest it.
We took a small trade into Thursday’s close because the data supported a bounce. By Friday morning that was about a half-percent gain. Quick hits like that help you stay ready for whenever this repairs.
Have a good weekend. Get away from the screens.