Paper: arXiv 2506.19255

Authors: Jianyong Fang, Sitong Wu, Junfan Tong

Abstract

We propose a novel two-stage framework to detect lead-lag relationships in the Chinese A-share market. First, long-term coupling between stocks is measured via daily data using correlation, dynamic time warping, and rank-based metrics. Then, high-frequency data (1-, 5-, and 15-minute) is used to detect statistically significant lead-lag patterns via cross-correlation, Granger causality, and regression models. Our low-coupling modular system supports scalable data processing and improves reproducibility. Results show that strongly coupled stock pairs often exhibit lead-lag effects, especially at finer time scales. These findings provide insights into market microstructure and quantitative trading opportunities.

Complexity vs Empirical Score

  • Math Complexity: 4.5/10
  • Empirical Rigor: 6.0/10
  • Quadrant: Street Traders — empirical and applied, lower math complexity

Why this score: The paper introduces a practical two-stage screening method with a low-coupling system design, featuring empirical testing on real A-share data across multiple time scales, but relies on standard statistical metrics (correlation, Granger causality) rather than advanced mathematics.

Research Flowchart

  flowchart TD
  A["Research Goal: Detect Lead-Lag<br>Relationships in Chinese A-Share Market"] --> B["Stage 1: Long-Term Coupling<br>(Daily Data)"]
  A --> C["Stage 2: High-Frequency<br>Lead-Lag Detection<br>(1/5/15-min Data)"]
  B --> D{"Coupling Analysis:<br>Corr, DTW, Rank Metrics"}
  C --> E{"Statistical Analysis:<br>CCF, Granger Causality, Reg"}
  D --> F["Identify Strongly<br>Coupled Stock Pairs"]
  E --> G["Detect Significant<br>Lead-Lag Patterns"]
  F --> G
  G --> H["Key Outcomes:<br>Strong Coupling -> Lead-Lag Effects<br>Finer Scales = Stronger Signals<br>Low-Coupling System Design"]

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 novel two-stage framework that first screens for strongly coupled stock pairs using long‑term daily data and then verifies lead‑lag relationships with high‑frequency data.

Source sentence
We propose a novel two-stage framework to detect lead-lag relationships in the Chinese A-share market. First, long-term coupling between stocks is measured via daily data using correlation, dynamic time warping, and rank-based metrics. Then, high-frequency data (1-, 5-, and 15-minute) is used to detect statistically si

Assumes 2

  • Stocks with strong long‑term coupling are more likely to exhibit meaningful lead‑lag effects. market
    Source sentence
    This two-stage design is founded on the hypothesis that stocks with fundamental economic relationships (as evidenced by their long-term coupling) are more likely to exhibit meaningful lead-lag effects, rather than spurious correlations.
  • Only stock pairs exceeding predefined coupling thresholds are considered for lead‑lag detection. data
    Source sentence
    Only stock pairs exceeding predefined coupling thresholds proceed to the second stage, where we apply the lead-lag detection methods described in the next section.

Data

  • A-share market data · Akshare API market-prices public
    2019-2024 · 1,283 A-share stocks (excluding listings after Jan‑1‑2019 and extended halts)
    Source sentence
    Data was obtained via the Akshare API, which provides access to OHLCV (Open, High, Low, Close, Volume) data from the East Money database.

Recognised sources: AKShare

Costs

Not applicable The paper reports no trading or execution result.

Finds 4

  • Lead‑lag effects are strongest at the 1‑minute granularity and weaken progressively at coarser granularities, disappearing at the daily level. statistical
    Source sentence
    Lead-lag effects are strongest at the 1-minute granularity and progressively weaken at coarser granularities … At the
  • Long‑term coupling between stocks is a reliable predictor of short‑term lead‑lag relationships; coupled pairs have a substantially higher probability of significant lead‑lag effects. statistical
    Source sentence
    Long-term coupling between stocks serves as a reliable predictor of short-term lead-lag relationships, with coupled stock pairs showing a substantially higher probability of exhibiting significant lead-lag effects.
  • Within‑industry and economically connected stock pairs exhibit the strongest lead‑lag relationships; large‑cap stocks tend to lead smaller industry peers. statistical
    Source sentence
    These relationships are strongest within industry groups and between companies with economic connections
  • The two‑stage “coupling screening followed by lag verification” methodology improves detection efficiency and reduces false positives. performance
    Source sentence
    Our two-stage screening and verification methodology significantly improves the efficiency of lead‑lag detection while reducing false positives.

Checked by others

Contradicted by

  • A Decadal Analysis of the Lead-Lag Effect in the NYSE (2023)
    On the question: Do lead‑lag relationships in stock markets align with industry or economic connections?
    Compare the findings

    This paper: Within‑industry and economically connected stock pairs exhibit the strongest lead‑lag relationships; large‑cap stocks tend to lead smaller industry peers.

    That paper: Lead‑lag pairs often form outside of industry lines, contradicting previous literature.

    A (NYSE) finds lead‑lag pairs often outside industry lines, while B (A‑Share market) finds the strongest lead‑lag links within industry and economic ties, showing a contradiction.

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.