Paper: arXiv 2508.07192
Authors: Masato Hisakado, Takuya Kaneko
Abstract
We study the eigenvalue of the Wigner random matrix, which is created from a time series with temporal correlation. We observe the deformation of the semi-circle law which is similar to the eigenvalue distribution of the Wigner-Lèvy matrix. The distribution has a longer tail and a higher peak than the semi-circle law. In the absence of correlation, the eigenvalue distribution of the Wigner random matrix is known as the semi-circle law in the large \(N\) limit. When there is a temporal correlation, the eigenvalue distribution converges to the deformed semi-circle law which has a longer tail and a higher peak than the semi-circle law. When we created the Wigner matrix using financial time series, we test the normal i.i.d. using the Wigner matrix. We observe the difference from the semi-circle law for FX time series. The difference from the semi-circle law is explained by the temporal correlation. Here, we discuss the moments of distribution and convergence to the deformed semi-circle law with a temporal correlation. We discuss the phase transition and compare to the Marchenko-Pastur distribution(MPD) case.
Complexity vs Empirical Score
- Math Complexity: 8.5/10
- Empirical Rigor: 6.5/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper uses advanced random matrix theory, deriving moments and phase transitions via heavy LaTeX formulas, while applying the theory to real FX data, including statistical tests and convergence analysis for practical implementation.
Research Flowchart
flowchart TD A["Research Goal<br>Investigate the effect of temporal correlation<br>on the eigenvalue distribution of Wigner random matrices<br>derived from time series"] --> B["Data & Inputs<br>Financial Time Series (FX)<br>Standard Wigner Matrix (i.i.d.)<br>Wigner-Lévy Matrix (Reference)"] B --> C["Methodology<br>Construct Wigner Random Matrix<br>from time series data"] C --> D["Computational Process<br>Calculate Eigenvalues<br>Compute Moments of Distribution<br>Check Convergence in Large N limit"] D --> E["Key Findings<br>Deformed Semi-Circle Law<br>(Longer tail, Higher peak)"] E --> F["Outcomes<br>1. Correlation explains deviation from semi-circle law<br>2. Phase transition observed vs Marchenko-Pastur distribution<br>3. Validation on FX data"]
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 that temporal correlations in financial time series deform the eigenvalue distribution of Wigner random matrices away from the semi‑circle law, yielding a deformed semi‑circle law with longer tails and higher peaks, and that a phase transition occurs depending on the decay exponen
Source sentence
We study the eigenvalue of the Wigner random matrix, which is created from a time series with temporal correlation. We observe the deformation of the semi-circle law which is similar to the eigenvalue distribution of the Wigner-Lèvy matrix.
Assumes 4
- The time series are normalized to have zero mean and unit variance. statistical
Source sentence
Here, we set the normalization, E(At) = 0, and V(At) = 1.
- Temporal correlations are modeled either as exponential decay or power‑law decay. model
Source sentence
In this study, we consider the exponential decay and power decay cases. Note that T = N 2. A. Exponential decay case We consider the exponential decay case, di = Cov(As, Ai+s) = ri, 0 ≤r ≤1 and calculate the moments. r is the effect at t from the past at t −1.
- For the exponential decay case the correlation parameter r satisfies 0 ≤ r ≤ 1. model
Source sentence
A. Exponential decay case We consider the exponential decay case, di = Cov(As, Ai+s) = ri, 0 ≤r ≤1 and calculate the moments. r is the effect at t from the past at t −1. Here, we set At+1 = rAt + √ 1 −r2ξt, (10) where ξt is i.i.d. and we obtain the exponential decay to create a time series with exponential decay, < At+1, At >= r.
- For the power‑law decay case the correlation follows a power index γ, with a phase transition at γ = 1/2. model
Source sentence
In particular, when the temporal correlation is power decay, we can confirm a phase transition from the finite second moment to the infinite second moment. If γ > 1/2, which is the power index, the second moment of the distribution and the largest eigenvalue are finite.
Data
- Financial return time series (FX, crypto, commodities, bonds, stock indexes) market-prices unknown
Source sentence
In the case of financial data, we use the historical data of the return rt as St. The return is defined using the market price, pt as rt = ln pt −ln pt−1. IV, APPLICATION TO FINANCIAL DATA Here, we calculate the eigenvalue distribution of the Wigner matrix and compare the moments with the semi-cricle law for 25 financial time series, which includes Crypt assets, foreign exchange, commodities, government bonds, and
Costs
Not applicable
Finds 5
- When temporal correlation is present, the eigenvalue distribution of the Wigner matrix deviates from the semi‑circle law, showing a longer tail and a higher central peak. theoretical
Source sentence
We observe the deformation of the semi-circle law which is similar to the eigenvalue distribution of the Wigner-Lèvy matrix. The distribution has a longer tail and a higher peak than the semi-circle law.
- In the exponential decay case the fourth moment remains finite (µ4 = 2 + 4/3·r²/(1‑r²)), confirming deformation but no divergence. theoretical
Source sentence
The fourth moment is µ4 = 2 + 4 3 r2 1 −r2, (11) in the limit of N →∞. Note that the fourth moment is finite, because r < 1.
- In the power‑law decay case, moments become infinite when γ ≤ 1/2, indicating a phase transition from finite to infinite fourth moment and largest eigenvalue. theoretical
Source sentence
If γ > 1/2, which is the power index, the second moment of the distribution and the largest eigenvalue are finite. On the other hand, when γ ≤1/2, the second moment and the largest eigenvalue are infinite.
- Empirically, most financial series fit the semi‑circle law, but several FX pairs (e.g., USD/CHF, USD/CAD, EUR/CHF, EUR/GBP) exhibit higher peaks and fatter tails consistent with temporal correlation effects. other
Source sentence
We can confirm that the moments for the time series fit well with the semi-circle law without FX and some crypt assets. Particularly, No.7 USD/CHF, No.9 USD/CAD, No.12 EUR/CHF, and No.13 EUR/GBP do not fit well.
- The critical Hurst index for the fBm case is Hc = 3/4 (γc = 1/2), above which a phase transition occurs and large eigenvalues appear. theoretical
Source sentence
Hence, the transition point is, Hc = 3/4, which corresponds to γc = 1/2. Above Hc = 3/4, a phase transition occurs and we can observe large eigenvalues.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 16 related papers). Coverage is partial, so this is neither replication nor refutation.
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.