Research notes
Conclusions, not claims.
Everything the lab learns about market microstructure, written up as it happens. Each note names the sample, names the label, and reports the result. Pass or fail.
Start hereGuide
Getting started with QuantDev.ai
How this site is organized, what to read first, and how to use it if you're serious about building trading systems.
Latest
Newest notes across every series.
Memory in microstructure: how discrete events become a signal
Raw ticks are noise. Memory is the bridge to a backtestable alpha.
Locking the first three raw atoms
Three atoms locked. One dropped on monotonicity. The data picked the seeds.
Four trade-aggression atoms — and three of them collapsed
Queue-normalised size. Sweep depth. Consumed-fraction. Through-ticks. Only one is its own axis.
Foundations
The machinery below the chart, and the habits that keep research honest.
Getting started with QuantDev.ai
Start here. The five-minute orientation.
Market microstructure, explained like you'll build something with it
Charts hide the machine. Learn what's actually happening.
Why your backtest lies to you (and how to make it stop)
A good backtest is an argument, not a result. Here's how to make the argument honest.
Latency isn't what you think it is
The distribution matters. The tail matters more.
Signals
Order-book signals walked through carefully, including where they break.
Method
How a research pass is structured, from raw messages to a cost-aware backtest.
Raw atoms
Evidence passes on candidate microstructure features. What locked, what collapsed.
Memory
Turning discrete events into a signal that survives to a tradeable horizon.