Paper: arXiv 2308.01112

Authors: Yuki Sato, Kiyoshi Kanazawa

Abstract

In this research, we focus on the order-splitting behavior. The order splitting is a trading strategy to execute their large potential metaorder into small pieces to reduce transaction cost. This strategic behavior is believed to be important because it is a promising candidate for the microscopic origin of the long-range correlation (LRC) in the persistent order flow. Indeed, in 2005, Lillo, Mike, and Farmer (LMF) introduced a microscopic model of the order-splitting traders to predict the asymptotic behavior of the LRC from the microscopic dynamics, even quantitatively. The plausibility of this scenario has been qualitatively investigated by Toth et al. 2015. However, no solid support has been presented yet on the quantitative prediction by the LMF model in the lack of large microscopic datasets. In this report, we have provided the first quantitative statistical analysis of the order-splitting behavior at the level of each trading account. We analyse a large dataset of the Tokyo stock exchange (TSE) market over nine years, including the account data of traders (called virtual servers). The virtual server is a unit of trading accounts in the TSE market, and we can effectively define the trader IDs by an appropriate preprocessing. We apply a strategy clustering to individual traders to identify the order-splitting traders and the random traders. For most of the stocks, we find that the metaorder length distribution obeys power laws with exponent \(α\), such that \(P(L)\propto L^{"-α-1"}\) with the metaorder length \(L\). By analysing the sign correlation \(C(τ)\propto τ^{"-γ"}\), we directly confirmed the LMF prediction \(γ\approx α-1\). Furthermore, we discuss how to estimate the total number of the splitting traders only from public data via the ACF prefactor formula in the LMF model. Our work provides the first quantitative evidence of the LMF model.

Complexity vs Empirical Score

  • Math Complexity: 8.0/10
  • Empirical Rigor: 9.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced statistical physics and econophysics models (LMF model, power-law distributions, autocorrelation functions) with significant mathematical derivation, while demonstrating high empirical rigor through the analysis of a large, nine-year high-frequency dataset, strategy clustering, and quantitative verification of theoretical predictions.

Research Flowchart

  flowchart TD
  Goal["Research Goal:<br>Quantitatively test LMF model predictions<br>on order-splitting in the Japanese market"]

  Data["Input Data:<br>TSE 9-Year Dataset<br>(Virtual Server Account IDs)"]
  
  Step1["Methodology: Preprocessing &<br>Strategy Clustering"]
  Step2["Methodology: Identify<br>Order-Splitting Traders"]
  Step3["Methodology: Analyze<br>Metaorder Length (L)"]
  Step4["Methodology: Analyze<br>Sign Correlation C(τ)"]

  Calc["Computational Process:<br>Verify LMF Prediction: γ ≈ α - 1"]
  
  Outcome["Key Findings:<br>1. Metaorder L follows Power Law P(L) ∝ L⁻α⁻¹<br>2. Confirmed γ ≈ α - 1 (Quantitative Support)<br>3. Proposed method to estimate # of traders"]

  Goal --> Data
  Data --> Step1
  Step1 --> Step2
  Step2 --> Step3
  Step2 --> Step4
  Step3 --> Calc
  Step4 --> Calc
  Calc --> Outcome

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 provides the first quantitative statistical analysis of order‑splitting behavior at the level of individual trading accounts and verifies the LMF model prediction.

Source sentence
In this report, we have provided the first quantitative statistical analysis of the order‑splitting behavior at the level of each trading account.

Assumes 4

  • Metaorder lengths follow a power‑law distribution P(L) ∝ L^{-α‑1} for large L. model
    Source sentence
    consistency with the realistic data analysis, it is a customary to assume the power‑law metaorder distribution: P(L) ∝L^{‑α‑1} for large L with a realistic value
  • Order‑splitting traders submit only market orders; limit‑order submissions are not used in the analysis. execution
    Source sentence
    The volume information on any market order is not used in this report.
  • The dataset’s virtual server IDs reliably identify individual traders except when shuffled on specific dates. data
    Source sentence
    While the virtual server IDs are kept identical for most of the periods, they were shuffled when the arrowhead system was updated on the 24th September 2015 and on the 4th November 2019.
  • A minimum of 0.5 million transactions per stock per year is required for reliable estimation. statistical
    Source sentence
    We studied only stocks whose yearly transaction number is over 0.5 million.

Data

  • Tokyo Stock Exchange (TSE) order‑book dataset · Tokyo Stock Exchange order-book proprietary
    2012‑2020 · All TSE stocks with >0.5 million yearly transactions
    Source sentence
    This dataset covers the nine‑years period from 2012 to 2020 for all the stocks. Remarkably, this dataset includes the virtual server ID, which is a unit of the trader accounts in the TSE platform.

Costs

Not applicable The paper reports no trading or execution result.

Finds 4

  • Metaorder‑length distributions for splitting traders follow power laws with exponent α > 1. statistical
    Source sentence
    As assumed in the LMF model, we confirm that the metaorder length for the STs obeys power‑laws, such that P(L) ∝L^{‑α‑1} for large L with α > 1.
  • The measured sign‑autocorrelation exponent γ satisfies γ ≈ α − 1, confirming the LMF prediction. theoretical
    Source sentence
    By measuring the power‑law exponent γ in the LRC (C(τ) ∝τ^{‑γ}), we provide the scatterplot between α and γ and then directly verify the LMF prediction (3).
  • Random traders (RTs) and splitting traders (STs) are identified for most stocks via binomial‑test clustering. statistical
    Source sentence
    We classify all traders into the random traders (RTs) and the splitting traders (STs) by the binomial test, directly confirming the presense of the STs for most of the stocks.
  • A practical method to estimate the total number of splitting traders from public ACF prefactor is proposed. theoretical
    Source sentence
    Furthermore, we discuss a practical method to estimate the total number of order splitters from the ACF prefactor on the basis of the LMF theory.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 22 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 1 paper in the archive (1 by independent authors)

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.