Thread by @TerribleQuant

Entirely unprompted here, but please check out @FadingRallies. Also @choffstein's Liquidity Cascades paper (link below). The flow between MMs, passive funds, ELS, and generally the effects of reflexive dealer hedging are key to understanding this regime!

thinknewfound.com/liquidity-casc…

Even if you aren't a trader (I certainly am not, although I try to keep up with it all) it is still super important to understand the regime and how it all fits in from a risk perspective. You CANNOT just take the models as your risk! Eigenportfolios decay, I would know I work

with them all the time so that isn't even the perfect metric (although I do love them). Statistical models will capture some risk but at the end of the day, you choose the parameters and the distribution you feed in is key. Knowing fat tails exist is incredibly important for this

Not just from a NNT pov where you say ahhh the end is nye, but from a conditional perspective where you have to play around with models like CVaR. Especially, if you are doing anything remotely long vol/skew/kurt then it will all be event-driven. Risk models aren't just a

peripheral model to include after the fact. Risk signals and alpha signals are distinctly important features to engineer. Risk signals are usually event driven, but can be statistical as well. VRP, CRP, SRP (vol,corr,skew risk premiums) are all great. VIX-VOLI too!

Your NNs only learn based on the data you feed them. One of the worst mistakes in quant for juniors is trusting your model. RenTech almost blew up, but Jim knew not to trust the models, TWICE, and that is a key reason they are still here today. Even the best funds are wrong!

Just don't take it personally, your model isn't wrong, it did it's best with the data. YOU WHERE WRONG, because you trusted the data. Data is non-ergodicit. The assumptions you make your money on (day-to-day) are very different than what you manage to keep your money based on.