Introducing High-frequency Market-making with Inventory Constraints and Directional Bets for after-work quants

High-frequency market-making is not only “capture the spread.” Inventory limits bind, and sometimes the desk also holds a directional view. How do those pieces interact in a formal model?

High-frequency Market-making with Inventory Constraints and Directional Bets (Fodra, Labadie, 2012) studies quoting under inventory constraints, with room for directional tilts—still as a controlled risk problem, not a tip sheet.

Fit for learners: Open-PDF classic/working paper on microstructure — useful for after-work Python quant learners building intuition before coding.

This Hub note is a plain-language intro. Not investment advice.

The intuition (without the full math)

• Pure spread capture ignores how fast inventory can hit a hard limit.
• Constraints change admissible quotes: when you are long and near the cap, you shade to encourage reducing risk.
• “Directional bets” here means the model can incorporate a view—educationally useful, easy to misuse if you treat a view as certainty.
• Read the paper for the precise control formulation and results; keep the evening takeaway qualitative.

Do not confuse a 2012 working-paper policy with a modern exchange-ready strategy.

Why after-work learners should care

Many blog posts show symmetric quotes around mid. Real learning starts when limits appear. This open PDF sits next to Avellaneda–Stoikov / Guéant–Lehalle in the same reading path: reservation ideas → inventory risk → constraints and views.

Healthy practice: add a hard inventory cap to a toy simulator and watch quote asymmetry appear—label “learning only.”

One-evening reading checklist

1. Restate inventory constraints in one sentence.
2. Skim how directional views enter (and how they could go wrong).
3. Compare mentally to Avellaneda–Stoikov: what feels similar?
4. Write a short note: “view ≠ edge; constraint ≠ stop-loss magic.”

Download the PDF

Download PDF

Landing: arXiv:1206.4810

Important disclaimer

Not investment advice. Academic market-making models are stylised. They are not instructions to quote on a live venue.

What this is not

• Not CEF / subsidy framing
• Not Quant or Macro course CTA
• Not prop-desk recruiting content

Next step

Single course CTA: SoloQuant Academy’s Introduction to Python Programming—language fundamentals first; stochastic control papers come later.