Spurious correlations, from A. Humbert
Lots of them! And with AI-generated explanations.
Asindu Drileba responds:
Wow. As much as the explanations may be wrong, they logically make sense. LLMs are really getting good. I didn't know they could do this.
Gyve Bones writes:
You can do similar things with astrological cycles and events, I came to realize when I built an ephemeris for the Market Information Machine and coded macros for ways to use the data. You can curve-fit any data to some combination of sinusoidal cycloids and get a 99% correlation, which then falls apart going forward because it was indeed a spurious back-fit with no reasonable causal linkage.
M. Humbert suggests:
If you track which countries where I’ve lived, those markets always experienced a bear market. I’m currently in the US. Advise strongly that you bet the farm, take out a 2nd mortgage, and buy out of the money put options the Magnificent 7.