Diary of a Pro Quant

My shot in the dark NFP prediction this month:



All models suggest a much better number than the 205k forecast.

CPI... Same models as last time so they're low confidence, but I'd like to document how they do none-the-less.

M/M - Matches expectations.




Y/Y - Below expectations.

I've been quieter than usual lately so I thought I'd provide an update... I've come to find manually running a lot of these scripts takes too long and it's inefficient. Therefore, I decided to move some of my scripts to the cloud where my charts will simply update automatically. This way, as I keep building out new charts I can just deploy them on the cloud and won't have to manually rerun them all the time.

This project is done. My seasonality charts are now running on the cloud, and the framework is in place to add more. Technically I made a website, but it currently has no domain since I don't intend to share anything on there.

Hello,


this question may seem stupid, but I would be happy to hear from you.
Does a quant trader focus on developing one or two strategies and focusing on them? Or is he always concerned with detecting new patterns with statistical anomalies in the market, putting together a series of these patterns to try to predict the probability and then place a trade?

What do I mean by statistical anomalies? If you buy and sell randomly with proportional TP/SL at the end of the day you will have a 50% assertiveness with small fluctuations, larger fluctuations (e.g. 56% win in 1000 trades) would be a statistical anomaly...

~Sorry, I know the focus here is ML and AI.