Paper: arXiv 2305.16915

Authors: Victor Le Coz, Iacopo Mastromatteo, Damien Challet, Michael Benzaquen

Abstract

Trading pressure from one asset can move the price of another, a phenomenon referred to as cross impact. Using tick-by-tick data spanning 5 years for 500 assets listed in the United States, we identify the features that make cross-impact relevant to explain the variance of price returns. We show that price formation occurs endogenously within highly liquid assets. Then, trades in these assets influence the prices of less liquid correlated products, with an impact velocity constrained by their minimum trading frequency. We investigate the implications of such a multidimensional price formation mechanism on interest rate markets. We find that the 10-year bond future serves as the primary liquidity reservoir, influencing the prices of cash bonds and futures contracts within the interest rate curve. Such behaviour challenges the validity of the theory in Financial Economics that regards long-term rates as agents anticipations of future short term rates.

Complexity vs Empirical Score

  • Math Complexity: 7.0/10
  • Empirical Rigor: 8.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced linear algebra (matrix calculus, covariance structures) to model cross-impact and derives optimal time scales, indicating high mathematical complexity. It is heavily empirical, utilizing 5 years of tick-by-tick data for 500 assets, backtesting models against return variance, and validating findings in interest rate markets.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>When is cross-impact relevant?"] --> B["Data:<br>5yr tick-by-tick data<br>500 US assets"]
  B --> C["Methodology:<br>Feature Identification<br>for variance explanation"]
  C --> D{"Computational Process:<br>Endogenous price formation?"}
  D -- Highly Liquid Assets --> E["Liquidity Reservoir Effect:<br>Trading pressure moves<br>correlated asset prices"]
  D -- Low Liquidity Assets --> F["Constraint:<br>Impact velocity limited by<br>minimum trading frequency"]
  E --> G["Interest Rate Implications:<br>10yr bond future = primary<br>liquidity reservoir"]
  F --> G
  G --> H["Key Finding:<br>Challenges theory that<br>long-term rates = expected<br>future short-term rates"]

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 to identify the features that make cross‑impact relevant for explaining price return variance and shows that the 10‑year bond future acts as the primary liquidity reservoir influencing other interest‑rate instruments.

Source sentence
Using tick-by-tick data spanning 5 years for 500 assets listed in the United States, we identify the features that make cross‑impact relevant to explain the variance of price returns. ... We

Assumes 6

  • Price formation occurs endogenously within highly liquid assets. model
    Source sentence
    We show that price formation occurs endogenously within highly liquid assets.
  • Trades in highly liquid assets influence less liquid correlated products with an impact velocity constrained by their minimum trading frequency. market
    Source sentence
    Then, trades in these assets influence the prices of less liquid correlated products, with an impact velocity constrained by their minimum trading frequency.
  • Stability property: it is impossible to manipulate the price of liquid products using illiquid instruments. model
    Source sentence
    Stability properties aim at ensuring the impossibility to manipulate the price of liquid products using illiquid instruments.
  • The sample covers the 2017‑2022 period with tick‑by‑tick trades and quotes for 500 US assets. data
    Source sentence
    Our sample includes stocks, bonds, futures on bonds and futures on stock indexes. Unless otherwise specified, our data set covers the 2017‑2022 period.
Show 2 more
  • Pairs are selected uniformly across correlation buckets to ensure coverage of existing correlations. statistical
    Source sentence
    aiming for a uniform coverage of existing correlations. Specifically, we categorize all potential pairs based on their correlation levels into 50,000 equally‑sized correlation buckets.
  • A minimum of 10 to 20 trades in both assets is required to reach the optimal time scale of a two‑dimensional cross‑impact model. execution
    Source sentence
    Consistently with the single asset case, we find that a minimum of 10 to 20 trades in both assets is required to reach the optimal time scale of a two‑dimensional cross‑impact model.

Data

  • Tick‑by‑tick trades and quotes for 500 US assets order-book unknown
    2017‑2022 · 500 assets (stocks, bonds, futures on bonds, futures on stock indexes)
    Source sentence
    We use tick‑by‑tick trades and quotes for 500 assets quoted in limit order books in the United States. Our sample includes stocks, bonds, futures on bonds and futures on stock indexes. Unless otherwise specified, our data set covers the 2017‑2022 period.

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • The 10‑year bond future is identified as the primary liquidity reservoir that influences cash bonds and futures across the interest‑rate curve. theoretical
    Source sentence
    We find that the 10‑year bond future serves as the primary liquidity reservoir, influencing the prices of cash bonds and futures contracts within the interest rate curve.
  • The optimal cross‑impact time scale τ* is driven by the minimum trading frequency of the asset pair, while the maximum frequency has little effect. statistical
    Source sentence
    τ* maximizing the added accuracy is affected by the minimum of the trading frequencies of the assets pair. In contrast, the maximum of these two frequencies has little effect on this time scale.
  • Higher trading frequency positively impacts the maximal accuracy R2* of the cross‑impact model. statistical
    Source sentence
    observed across the tested bin sizes. Indeed, Fig. 7 shows that an increase in the trading frequency improves the mean R2*
  • Added accuracy from cross‑sectional information ∆R2* increases with correlation, rising from 0 to above 5% across the sample. statistical
    Source sentence
    As expected, we observe a positive and monotonous relationship between the added accuracy from the cross sectional information ∆R2*
  • A minimum of 10‑20 trades per asset is required to achieve the optimal two‑dimensional cross‑impact time scale. statistical
    Source sentence
    Consistently with the single asset case, we find that a minimum of 10 to 20 trades in both assets is required to reach the optimal time scale of a two‑dimensional cross‑impact model.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 37 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 4 papers in the archive (2 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.