Paper: arXiv 2307.04953
Authors: Alejandro Rodriguez Dominguez, Om Hari Yadav
Abstract
This paper presents a method to identify causal interactions between two time series. The largest eigenvalue follows a Tracy-Widom distribution, derived from a Coulomb gas model. This defines causal interactions as the pushing and pulling of the gas, measurable by the variability of the largest eigenvalue’s explanatory power. The hypothesis that this setup applies to time series interactions was validated, with causality inferred from time lags. The standard deviation of the largest eigenvalue’s explanatory power in lagged correlation matrices indicated the probability of causal interaction between time series. Contrasting with traditional methods that rely on forecasting or window-based parametric controls, this approach offers a novel definition of causality based on dynamic monitoring of tail events. Experimental validation with controlled trials and historical data shows that this method outperforms Granger’s causality test in detecting structural changes in time series. Applications to stock returns and financial market data show the indicator’s predictive capabilities regarding average stock return and realized volatility. Further validation with brokerage data confirms its effectiveness in inferring causal relationships in liquidity flows, highlighting its potential for market and liquidity risk management.
Complexity vs Empirical Score
- Math Complexity: 7.5/10
- Empirical Rigor: 5.5/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper employs advanced random matrix theory (Tracy-Widom, Coulomb gas) and derives novel causal indicators, demonstrating high mathematical sophistication. It also shows strong empirical validation with synthetic data, historical financial data, and brokerage data, comparing performance to Granger causality and outlining practical applications in risk management.
Research Flowchart
flowchart TD
A["Research Goal: Identify causal interactions between time series<br>using variations in the first eigenvalue"] --> B["Methodology: Lagged Correlation Matrices & Coulomb Gas Model"]
B --> C{"Calculate Largest Eigenvalue λ₁(t) for each lag"}
C --> D["Statistical Analysis: Apply Tracy-Widom Distribution"]
D --> E["Compute Explanatory Power Variability<br>Standard deviation of λ₁"]
E --> F["Define Causal Interaction Indicator<br>Push/Pull dynamics measured by tail event variability"]
F --> G["Key Findings & Outcomes"]
G --> H1["Outperforms Granger's Causality in detecting structural changes"]
G --> H2["Validated on stock returns: Predicts avg return & realized volatility"]
G --> H3["Effective in inferring liquidity flow causality"]
G --> H4["Novel method for market & liquidity risk management"]
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 new indicator based on random matrix theory to detect causal interactions between two time series using the variability of the largest eigenvalue.
Source sentence
A new method to detect causal interactions between two time series using random matrix theory (RMT) and eigenvalues’ distributions being defined from physical experiments is presented, in the form of an indicator.
Assumes 2
- The largest eigenvalue of lagged correlation matrices follows a Tracy‑Widom distribution derived from a Coulomb gas model. statistical
Source sentence
The largest eigenvalue follows a Tracy-Widom distribution, derived from a Coulomb gas model.
- Causality is enforced by the direction of time through lagged correlation matrices. model
Source sentence
the causality component of the interaction is enforced by the direction of time, thanks to lagging the correlation matrices from which the variations of the explanatory power are being measured.
Data
- Financial market daily time series market-prices commercial
Source sentence
The first one uses financial market daily time series data sourced from Bloomberg, spanning the years 2013 to 2023.
- Brokerage business daily time series alternative proprietary
Source sentence
The second set of experiments uses daily time series data from end-of‑day files of a brokerage business containing all clients’ cash and trading activities, flows, and positions. The anonymous dataset was provided by Miralta Finance Bank S.A.
Costs
Not applicable
Finds 5
- The proposed indicator outperforms Granger’s causality test in detecting structural changes in time series. performance
Source sentence
Experimental validation with controlled trials and historical data shows that this method outperforms Granger's causality test in detecting structural changes in time series.
- The indicator σΓ shows dramatic jumps at known structural change points, capturing causal drivers more strongly than non‑causal pairs. statistical
Source sentence
It is clearly seen how the indicator σΓ, computed for all pairs of stock/drivers’ candidates, dramatically jumps.
- The indicator provides predictive capabilities for average stock return and realized volatility. predictive-accuracy
Source sentence
Applications to stock returns and financial market data show the indicator's predictive capabilities regarding average stock return and realized volatility.
- Liquidity risk can be monitored by analyzing interactions between operational flows and cash‑balance increments. other
Source sentence
Liquidity risk can be monitored by analyzing interactions between operational flows (executions and trading activities) and increments in the cash-balance.
- Synthetic i.i.d. series produce much smaller cumulative σΓ values than series with known structural relationships, confirming the indicator captures structural changes. statistical
Source sentence
... the values of the cumulative sum of the indicator σΓ are much smaller than the values for time series with structural relationships and changes.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 26 related papers). Coverage is partial, so this is neither replication nor refutation.
1 paper in the archive
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.