Skip to content
All library documents

Inventory Risk and Avellaneda–Stoikov Market Making

Article Quant Q&A · Author: zichao wang
Read the original

Summary

This short response points to the Avellaneda–Stoikov framework as a practical starting point for coding a market-making strategy. The framework models how a market maker can quote bid and ask prices while managing inventory, helping organize the tradeoff between earning spread and accumulating unwanted exposure. It is suggested in response to questions about order placement, fills, and hedging in a simplified trading simulator.

The response highlights two important limitations of the referenced setup: it uses a basic price process without drift or the market impact of the trader’s own orders, and it assumes unfilled orders are canceled at each time step without cost. These assumptions make the model easier to implement but limit its realism. The document does not give specific quote formulas, hedge rules, parameter choices, or performance evidence, so the reference is a conceptual direction rather than a complete strategy.

Key ideas

  • Avellaneda–Stoikov provides a framework for quoting while accounting for inventory exposure.
  • Market-making decisions balance spread capture against the risk of one-sided fills.
  • The referenced model omits price drift and the market impact of the trader’s orders.
  • Its assumption that unfilled quotes are canceled each time step may not match live trading.

Tags

Market makingMarket microstructureRisk management
Full text
# Market Making Algorithm/ Strategies


# Market Making Algorithm/ Strategies












I have been taking a "Trading Strategies" course, but the experience is awful as the instructor barely provides any learning resources. I have an upcoming evaluation on market making algorithm using VBA ran on a trading simulator against other classmates, the goal is to basically to maximize PnL.

The case brief is here if you are interested in the details. https://people.ucalgary.ca/~sick/RITC/Cases/RIT%20-%20Case%20Brief%20-%20ALGO2%20-%20Algorithmic%20Market%20Making.pdf

I am trying to find the optimal strategies for MM, and I have the following questions

- To earn spread & rebate, I always submit a pair of bid and offer orders. However, they may not be filled, which will lead to inventory risk. What is the best way to hedge against this risk?

- what price of limit orders should I submit at to maximize my chance of order getting filled?

- In general, what's the best strategy for dealing with this simplified version of MM case

## Answer by python_enthusiast (score 5)

https://quant.stackexchange.com/a/44398

Check out Avellaneda and Stoikov (2008)

They model the market maker's problem in a very neat and easy to code way. Some caveats of the model, in case you do decide to use it:

- The price process is the simplest one possible, does not consider drift or market impact of your orders;

- Orders that are not executed are cancelled at the end of each time step (this has no cost to you, but it is something to pay attention to when you are trading, so as to not 'forget' orders in the book.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.