Introducing A Mathematical Approach to Order Book Modeling for after-work quants

Agent-based order-book stories feel microscopic. Diffusions used in pricing feel macroscopic. How do you connect them with math you can actually state?

A Mathematical Approach to Order Book Modeling (Abergel and Jedidi, 2010) studies the limit order book as a multidimensional continuous-time Markov chain, with results in the independent Poisson-arrival case.

This Hub note is a plain-language map for after-work learners. Not investment advice.

Core ideas in plain language

• Model the book as a Markov process (future depends on the present state of the book, given the model’s rules).
• With Poisson-like order flow assumptions, they study existence of a stationary distribution and convergence toward it—cancellation structure matters a lot.
• Via a functional central limit theorem style argument, a rescaled, centered price can look like Brownian motion—a bridge from book dynamics to a familiar macroscopic noise model.
• They illustrate with simulation and market-data comparison (see PDF).

You are not asked to prove the theorems tonight. You are asked to recognise why serious microstructure papers talk about stationarity, cancellations, and scaling.

Why after-work Python learners should care

If you only ever plot OHLC candles, the book is invisible. If you only copy HFT buzzwords, the math is invisible. This paper sits in the middle: it explains what kind of object an order book is in probability terms.

After reading, a healthy mini-project is still humble: simulate a toy Poisson book in Python, plot depth, and stop—no live trading.

One-evening reading checklist

1. Read the motivation: micro agents vs macro SDEs.
2. Skim definitions of the state and arrival events (limit, market, cancel).
3. Note the stationarity / convergence message and the Brownian scaling message.
4. Write three questions for a future you who knows more measure-theoretic probability.

Download the PDF

Download PDF

Landing: arXiv:1010.5136

Important disclaimer

Educational only. Not investment advice. Mathematical models of books are stylised. They do not tell you what to trade.

What this is not

• Not CEF content
• Not Quant/Macro course CTA
• Not an execution-algorithm product pitch

Next step

Need cleaner Python before stochastic models feel friendly? Use Introduction to Python Programming only—the single course link on this page.