Auto deleveraging rules balance fairness and platform risk effectively

Path-Independent Deleveraging Rules

Computer Science and Game Theory

Summary

When financial platforms automatically reduce risky positions, how they distribute losses matters a lot. The authors study rules that split these losses in ways that don’t depend on the order of events and that handle account changes fairly. They find a set of rules that trade off between spreading risk evenly and being fair to all accounts. Testing their ideas on a real platform’s past event shows better results in sharing burden and reducing risk compared to what was actually used.

What this means in practice

  • For cryptocurrency exchange risk teams: Improve auto-deleveraging methods by choosing rules that better balance fairness and platform risk based on historic episode replay.
  • For financial risk managers: Implement parametrized deleveraging rules to control risk exposure while managing fairness across client accounts during downturns.

Tested on one dataset.

Authors

Christoph Schlegel, Akaki Mamageishvili

Abstract

We characterise path-independent auto-deleveraging (ADL) rules as an extended log-exponential family. Immunity to account splitting, merging and reopening further restrict the class. Homogeneity (essentially) restricts the class to a one-parameter family ordered by both fairness and platform risk. Its split-proof members continuously interpolate between proportional division and water-filling. Among these rules, decreasing the parameter reduces risk at the cost of increasing inequality in how accounts are treated. This suggests selecting the parameter through an ex-ante calibration to a risk bound on historical data to choose a maximally fair rule subject to that bound. A counterfactual replay of Hyperliquid's ADL episode on 10 October 2025 finds that a range of intermediate parameters significantly improves on the observed allocation in both burden dispersion and subsequent risk statistics.