Herding and Liquidity in Order-Book Markets. II. Fundamental Anchoring and the Resilience of Liquidity

By Jan Novotny

Rating

1862
Battle Count: 74

Relevance

6/10
Highly relevant for understanding order-book resilience, liquidity stress dynamics, and the limits of cross-market contagion. Directly informs market-making strategies (inventory risk, funding constraints), pairs trading (limits of arbitrage as a contagion channel), and risk management (distinguishing intrinsic fragility from contagion). Less directly applicable to signal generation or alpha capture, but critical for understanding when and why liquidity evaporates in stressed markets. The finding that fundamental anchoring is the stabiliser rather than market-maker presence has implications for exchange design and liquidity provision strategies.

Implementation Complexity

7/10
The agent-based model involves discrete-event simulation of coupled limit order books with multiple agent types (anchored providers, momentum herders, arbitrageurs, market makers with inventory control, leveraged holders with margin calls). The six-channel hierarchy, anchor dial, matched-control protocol, and paired statistical tests add significant complexity. However, the model is fully specified with clear parameter ranges and the code is publicly available. The main complexity lies in correctly implementing the funding-constrained maker population, the leverage spiral mechanics, and the rigorous stressed-versus-matched-control design.

Reproducibility

5/5
Full code, simulation, and analysis code available on GitHub (https://github.com/hanssmail/cross-market-liquidity-abm). Event-driven simulation with no integrator step to converge. All runs checked for non-finite values. Foundation gates verify zero-coupling reduction and sign consistency. Paired seed design (seeds 1000-1019) supports stronger statistical tests. Companion single-market paper (Novotny, 2026a) provides the baseline crossover.

About this paper

Methodology: Agent-Based Model with Causal Dial Experiment. Problem types: Market Making, Risk Management, Causal Inference, Systemic Risk Assessment, Liquidity Stress Modelling, Cross-Market Contagion Analysis.

The interactive Everscope explorer (charts, battles, favorites) loads below.