NQ Research Lab
retained text
Been working on some really cool stuff lately. What initially began as a detailed analysis of my historical trading activity (to develop an automated algo) has sprawled over weeks into a multidimensional market state research framework. The end goal is to produce a profitable trading system, but if that doesn’t come to fruition, building a market state engine. So if I can’t develop a legitimate trading edge, instead of predicting NQ, imagine an engine that answers in real time - what type of environment are we trading right now and which behaviors historically work in this environment? That could be super useful to someone with a discretionary edge - think about looking at the chart and thinking about an entry - then you check the market state engine and see that the current behavior is continuation friendly, amplified, and persistent. Suddenly that mean reversion trade might not be the best choice. This could allow you to increase the probability of profitable trades with more targeted entries. First I built an NQ “observer” in GPT a couple weeks ago that collects second by second data related to my different trading strategies and tags the data so that I can analyze market state before and after certain events transpire. It is a lot of data. I will begin analyzing after a few weeks worth of RTH sessions. In addition to the observer data, i have tick level NQ price data from databento to aid with constructing the NQ internal state. I am starting analysis on market breadth state using tick-NQ data. After that is 0 DTE QQQ options data to analyze directional flow stack studies and then separately gamma/hedging mechanisms. After the individual states are analyzed independently, I am going to do a multidimensional interaction including all states. I’m really excited about where this could go and would love any thoughts and feedback. Also curious if anyone else has done similar research and analysis. FYI I have been a discretionary NQ trader for 15 years.
document fetched 2026-10-01 00:48 · canonical https://reddit.com/r/algotrading/comments/1wubkbk/nq_research_lab