Paper: arXiv 2306.13378
Authors: Yuki Sato, Kiyoshi Kanazawa
Abstract
The Lillo-Mike-Farmer (LMF) model is an established econophysics model describing the order-splitting behaviour of institutional investors in financial markets. In the original article (LMF, Physical Review E 71, 066122 (2005)), LMF assumed the homogeneity of the traders’ order-splitting strategy and derived a power-law asymptotic solution to the order-sign autocorrelation function (ACF) based on several heuristic reasonings. This report proposes a generalised LMF model by incorporating the heterogeneity of traders’ order-splitting behaviour that is exactly solved without heuristics. We find that the power-law exponent in the order-sign ACF is robust for arbitrary heterogeneous intensity distributions. On the other hand, the prefactor in the ACF is very sensitive to heterogeneity in trading strategies and is shown to be systematically underestimated in the original homogeneous LMF model. Our work highlights that the ACF prefactor should be more carefully interpreted than the ACF power-law exponent in data analyses.
Complexity vs Empirical Score
- Math Complexity: 8.5/10
- Empirical Rigor: 3.0/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper is mathematically dense, presenting an exact analytical solution for a generalized econophysics model using advanced stochastic process theory and statistical mechanics derivations (Section 3 and appendices). It lacks implementation details, backtests, or real trading metrics, focusing instead on theoretical robustness of parameters (prefactor vs. exponent) in an idealized closed system.
Research Flowchart
flowchart TD
A["Research Goal"] --> B["Generalize LMF Model"]
B --> C{"Incorporate Heterogeneity"}
C --> D["Formulate Exact Solution"]
D --> E["Analyze Order-Sign ACF"]
E --> F["Compare to Original LMF"]
F --> G["Key Findings"]
subgraph B ["Methodology"]
direction LR
C
D
end
subgraph E ["Computational Process"]
direction LR
Analyze["Analyze ACF Structure"]
Analyze --> Derive["Derive Power-Law Exponent & Prefactor"]
end
subgraph F ["Comparison"]
direction LR
Homogeneous["Original Homogeneous Model"]
Heterogeneous["New Heterogeneous Model"]
Compare["Compare Exponents & Prefactors"]
end
G --> H["Power-law exponent robust to heterogeneity"]
G --> I["Prefactor sensitive to heterogeneity"]
G --> J["Original model underestimates prefactor"]
style A fill:#e1f5fe
style B fill:#f3e5f5
style G fill:#e8f5e8
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a generalized Lillo-Mike-Farmer (LMF) model that incorporates heterogeneous order‑splitting strategies and provides an exact solution.
Source sentence
This report proposes a generalised LMF model by incorporating the heterogeneity of traders' order-splitting behaviour that is exactly solved without heuristics.
Assumes 5
- The trading system is closed with a constant number of traders M. model
Source sentence
Section 2 describes our model and mathematical notation with the assumption of the closed system.
- Each trader i has a submission rate λ(i) and a metaorder‑length distribution ρ(i)(L). model
Source sentence
We incorporate the heterogeneity of trading strategies into our model, and the characteristic parameters of the ith trader are given by the submission rate λ(i) and metaorder-length (or run-length) distribution ρ(i)(L).
- Executed volume size is always the minimum unit of transactions. execution
Source sentence
For simplicity, we assume that the executed volume size is always the minimum unit of transactions.
- Metaorder length follows a Pareto‑type power‑law distribution asymptotically (ρ≥(L) ≃ cρ L^{‑α} for large L). statistical
Source sentence
Rather, it is a more realistic assumption that the power-law holds only asymptotically: ρ≥(L) ≃cρL−α for large L. (58) Actually, we validated this weaker assumption in our microscopic datasets of the TSE market (see Ref. [20, 21]).
Show 1 more
- Intensity distribution {λ(i)} may be heterogeneous across traders. model
Source sentence
We incorporate the heterogeneity of trading strategies into our model, and the characteristic parameters of the ith trader are given by the submission rate λ(i) and metaorder-length (or run-length) distribution ρ(i)(L). 5.2 Non-robust prefactor formula On the other hand, the prefactor cSK 0 is very sensitive to the heterogeneity of the intensity distribution {λ(i)}i∈ΩPT.
Data
- TSE market microscopic datasets order-book unknown
Source sentence
Rather, it is a more realistic assumption that the power-law holds only asymptotically: ρ≥(L) ≃cρL−α for large L. (58) Actually, we validated this weaker assumption in our microscopic datasets of the TSE market (see Ref. [20, 21]).
Costs
Not applicable
Finds 3
- The power‑law exponent formula γ = α − 1 for the order‑sign autocorrelation function remains robust under arbitrary heterogeneous intensity distributions. theoretical
Source sentence
We find that the power-law exponent in the order-sign ACF is robust for arbitrary heterogeneous intensity distributions. Our theoretical formu-las imply that (i) the power-law exponent formula γ = α −1 robustly holds even in the presence of the heterogeneous intensity distributions.
- The prefactor of the autocorrelation function is highly sensitive to the heterogeneity of trading strategies and is systematically underestimated by the original homogeneous LMF model. theoretical
Source sentence
On the other hand, the prefactor in the ACF is very sensitive to heterogeneity in trading strategies and is shown to be systematically underestimated in the original homogeneous LMF model. (iii) Furthermore, the classical prefactor formula (3b) systematically underestimates the actual prefactor in the presence of heterogeneity in agents.
- Empirical analysis on TSE data confirms that the heterogeneous prefactor formula better matches observed autocorrelations than the homogeneous formula. statistical
Source sentence
Rather, it is a more realistic assumption that the power-law holds only asymptotically: ρ≥(L) ≃cρL−α for large L. (58) Actually, we validated this weaker assumption in our microscopic datasets of the TSE market (see Ref. [20, 21]).
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 13 related papers). Coverage is partial, so this is neither replication nor refutation.
2 papers in the archive : Switching Frictions, Heterogeneous Trading Horizons, and …
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.