Thread by @KrisAbdelmessih

A very humble thought on math in trading. I say humble because I risk straw-manning quant (look, I have yet to meet an IYI type quant that Taleb would caricature. I've been blown away by the curiousity and brains of every quant I've worked with). So with that caveat...

I look at backtests. But their useful domain feels really narrow to me. I'm not a quant so maybe it's just fear of what's over my head. I'll give an example of the type of idea that diverts my eyes from backtesty work.

I'm more attracted to ideas that are upstream of past moves. Understanding flows is an example of that.

How does it interact with backtests?

At the meta level. When I see a backtest and long histories etc my first meta question is about is N really N?

If the same dynamic governs a long time period that dynamic is driving actions. Markets are theater. There's a meta game of beliefs and narrative that cause large accounts to hit buy and sell. If the narratives and players preside over a large period is N possibly much smaller?

I'd call it statistical regime shrinkage for lack of an educated term. Stickiness may be reinforcing N but what is reinforcing the stickiness? What mandates and rules are behind the money and what makes them change.

I like learning about portfolio construction math and I like playing with simple simulations. Simple simulations are limited bc they are generated by well-behaved distributions. But they are handy for intuition.

Backtests may offer more realistic distributions but they have their own baggage in the way they can limit your view. There's a siren song of understanding in them that is easily oversold if you open yourself to it.

If I'm making a type II error it's the error I prefer

So while I certainly like to indulge looking at backtests I'm almost never getting excited about them (at least any non-short term stuff).

I'm much more interested in what holds the theater together and activates flow or a new behavior. Something durable. With boundaries.

Things that you can see change, traceable back to humans giving a directive to convert money into risk or vice versa.

Ok, done w/ my poor man's @VitruviusCurve imitation

1 of my fave things written about markets didn't know it was about markets.
ncase.me/attractors/