arxiv: 2605.05173 · v1 · submitted 2026-05-06 · 🧮 math.ST · stat.TH

Recognition: unknown

Concordance, symmetrization and non-exchangeability for bivariate copulas

\'Avaro Rodr\'iguez-Garc\'ia, Manuel \'Ubeda-Flores

Pith reviewed 2026-05-08 15:24 UTC · model grok-4.3

classification 🧮 math.ST stat.TH

keywords bivariate copulasnon-exchangeabilitysymmetrizationSpearman's rhoSchweizer-Wolff measureBlomqvist's betadependence measures

0 comments

The pith

Symmetrization preserves Spearman's rho but annihilates non-exchangeability in bivariate copulas.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper connects non-exchangeability measures for bivariate copulas to standard dependence and concordance functionals. It shows that the symmetrization operator, which averages a copula with its transpose, leaves Spearman's rho unchanged while driving all non-exchangeability measures to zero. Blomqvist's beta is found to carry no information about the degree of asymmetry. The central result is the sharp inequality that the Schweizer-Wolff measure sigma is bounded below by six times the first non-exchangeability measure mu_1, which directly implies that any asymmetry forces positive dependence. Explicit formulas for Kendall's tau, Spearman's rho, and tail coefficients are derived for the family of maximally non-exchangeable copulas as concrete illustrations.

Core claim

The symmetrization C maps to (C + C transpose)/2 preserves Spearman's rho while annihilating the non-exchangeability measures mu_p for all p. Blomqvist's beta is independent of the degree of non-exchangeability. The sharp lower bound sigma(C) greater than or equal to 6 mu_1(C) holds for the Schweizer-Wolff measure sigma, establishing that asymmetry implies dependence. Closed-form expressions for tau, rho, and the tail-dependence coefficients are obtained for the family of maximally non-exchangeable copulas.

What carries the argument

The symmetrization operator that replaces any copula C by the average (C + C transpose)/2, acting together with the family of non-exchangeability measures mu_p and the Schweizer-Wolff dependence measure sigma.

If this is right

Spearman's rho is invariant under symmetrization for every bivariate copula.
Every non-exchangeability measure mu_p drops to zero after symmetrization.
The inequality sigma greater than or equal to 6 mu_1 is attained for some copulas and therefore sharp.
Blomqvist's beta cannot detect or quantify non-exchangeability.
Explicit formulas for dependence and tail coefficients exist for the entire family of maximally non-exchangeable copulas.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The bound suggests that any dependence measure that vanishes on independent copulas must also vanish on symmetric copulas.
Numerical checks on randomly generated copulas could confirm how often the bound is close to equality.
The preservation property may allow construction of asymmetric copulas with prescribed rho values by starting from symmetric ones and adding controlled asymmetry.

Load-bearing premise

The non-exchangeability measures mu_p defined in prior work are well-defined for every bivariate copula and behave as expected under the standard symmetrization operator.

What would settle it

A single bivariate copula C for which sigma(C) is strictly less than 6 mu_1(C), or for which Spearman's rho changes after symmetrization.

read the original abstract

We study the relationship between measures of non-exchangeability $\mu_p$ ($p\in[1,+\infty]$), in the sense of Durante et al. (2010), and classical dependence functionals for bivariate copulas. We show that the symmetrization $C\mapsto(C+C^t)/2$ preserves Spearman's $\rho$ while annihilating $\mu_p$, and that Blomqvist's $\beta$ carries no information about the degree of non-exchangeability. We also establish the sharp lower bound $\sigma(C)\ge 6\,\mu_1(C)$, where $\sigma$ is the Schweizer-Wolff dependence measure, showing that asymmetry implies dependence. Closed-form expressions for $\tau$, $\rho$, and the tail-dependence coefficients of the maximally non-exchangeable family $\{M_\theta\}$ are derived as illustrations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper proves a sharp bound σ(C) ≥ 6 μ_1(C) for bivariate copulas and supplies closed forms for the M_θ family, but stays inside a narrow slice of copula theory.

read the letter

The central new piece is the inequality σ(C) ≥ 6 μ_1(C), obtained by comparing the L1 integrals that define the two measures and using the elementary relation |C(u,v) − uv| ≥ (1/2)|C(u,v) − C(v,u)| after symmetrization. They also show that symmetrization leaves Spearman's ρ unchanged while setting every μ_p to zero, and that Blomqvist's β is blind to asymmetry. For the family of maximally non-exchangeable copulas they give explicit formulas for τ, ρ, and the tail coefficients. These calculations are direct and use only standard copula properties, so the results look reproducible from the definitions alone. The bound is presented as sharp, and the argument given in the stress-test note is internally consistent with no circular steps or hidden assumptions beyond the usual normalization of the measures. The paper does not claim anything outside the bivariate setting or try to link the inequality to estimation or simulation, which keeps the scope honest. The main limitation is exactly that narrow scope: everything is bivariate, the functionals are the specific ones from Durante et al., and there is no discussion of how the bound might behave under margins or in higher dimensions. A reader who already works with non-exchangeable copulas will find the inequality and the closed forms useful as reference points; someone outside that subfield will not need the paper. The math is grounded and the claims are modest, so the work is worth a referee's time even if revisions are needed to tighten the sharpness verification or add a short numerical check.

Referee Report

0 major / 3 minor

Summary. The manuscript explores the connections between non-exchangeability measures μ_p (p ∈ [1, +∞]) for bivariate copulas, as defined by Durante et al. (2010), and standard dependence measures such as Spearman's ρ, Blomqvist's β, and the Schweizer-Wolff measure σ. It demonstrates that symmetrization of a copula preserves Spearman's ρ but eliminates non-exchangeability, that Blomqvist's β provides no information on the degree of non-exchangeability, establishes the sharp lower bound σ(C) ≥ 6 μ_1(C) implying that asymmetry forces dependence, and provides closed-form expressions for Kendall's τ, Spearman's ρ, and tail-dependence coefficients for the family of maximally non-exchangeable copulas {M_θ}.

Significance. The results offer a clear quantitative relationship between non-exchangeability and dependence through the sharp bound, which is derived from basic integral inequalities and normalization constants. The preservation properties under symmetrization and the explicit calculations for {M_θ} add value for theoretical understanding and practical modeling of asymmetric dependence structures in copulas. These findings are grounded in standard copula theory and could influence how dependence is assessed in the presence of asymmetry.

minor comments (3)

The definition of the symmetrization operator C ↦ (C + C^t)/2 should be stated explicitly at the beginning of the relevant section for clarity.
A brief discussion on the conditions under which the inequality |C(u,v) − uv| ≥ (1/2)|C(u,v) − C(v,u)| holds would enhance the proof's transparency.
Ensure that the range of the parameter θ is clearly specified when introducing the family {M_θ}.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive assessment of our manuscript and the recommendation of minor revision. The referee's summary correctly reflects the paper's contributions on the relationships between non-exchangeability measures and classical dependence measures for bivariate copulas.

Circularity Check

0 steps flagged

No significant circularity; derivations are self-contained from definitions

full rationale

The paper's central results, including the sharp bound σ(C) ≥ 6 μ_1(C) and properties of symmetrization, are obtained by direct integral comparisons and standard copula inequalities such as |C(u,v) − uv| ≥ (1/2)|C(u,v) − C(v,u)|. The μ_p measures are imported from the external reference Durante et al. (2010) without self-citation chains or fitted inputs. No step reduces a claimed prediction to a definition or prior result by construction; all derivations remain independent of the target quantities.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work relies on prior definitions of non-exchangeability measures and standard copula properties without introducing new free parameters or invented entities.

axioms (1)

domain assumption Bivariate copulas satisfy the standard properties and the non-exchangeability measures μ_p are as defined in Durante et al. (2010)
All results build directly on these established definitions for the functionals and symmetrization.

pith-pipeline@v0.9.0 · 5444 in / 1238 out tokens · 44383 ms · 2026-05-08T15:24:58.224765+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

16 extracted references

[1]

and Dette, H

Bücher, A. and Dette, H. (2010). A note on bootstrap approximations for the empirical copula process.Statistics & Probability Letters,80(23–24), 1925–1932

2010
[2]

and Teicher, H

Chow, Y.S. and Teicher, H. (1978).Probability Theory: Independence, Interchangeability, Mar- tingales. Springer, New York

1978
[3]

and Fernández-Sánchez, J

Durante, F. and Fernández-Sánchez, J. (2010). Multivariate shuffles and approximation of cop- ulas.Statistics & Probability Letters,80(23–24), 1827–1834

2010
[4]

and Fernández-Sánchez, J

Durante, F. and Fernández-Sánchez, J. (2012). On the approximation of copulas via shuffles of Min.Statistics & Probability Letters,82(10), 1761–1767

2012
[5]

and Sempi, C

Durante, F. and Sempi, C. (2016).Principles of Copula Theory. Chapman & Hall/CRC, Boca Raton

2016
[6]

Durante, F., Klement, E.P., Sempi, C., and Úbeda-Flores, M. (2010). Measures of non- exchangeability for bivariate random vectors.Statistical Papers,51(3), 687–699. 7

2010
[7]

and Rivest, L.-P

Genest, C. and Rivest, L.-P. (1993). Statistical inference procedures for bivariate Archimedean copulas.Journal of the American Statistical Association,88(423), 1034–1043

1993
[8]

Hua, L., Polansky, A., and Pramanik, P. (2019). Assessing bivariate tail non-exchangeable de- pendence.Statistics & Probability Letters,155, Article 108556

2019
[9]

(1997).Multivariate Models and Dependence Concepts

Joe, H. (1997).Multivariate Models and Dependence Concepts. Chapman & Hall, London

1997
[10]

and Mesiar, R

Klement, E.P. and Mesiar, R. (2006). How non-symmetric can a copula be?Commentationes Mathematicae Universitatis Carolinae,47(1), 141–148. Mikusiński, P., Sherwood, H., and Taylor, M.D. (1992). Shuffles of Min.Stochastica,13(1), 61–74

2006
[11]

(2006).An Introduction to Copulas, 2nd ed

Nelsen, R.B. (2006).An Introduction to Copulas, 2nd ed. Springer, New York

2006
[12]

Nelsen, R.B. (2007). Extremes of nonexchangeability.Statistical Papers,48(2), 329–336. Pérez, A. and Prieto-Alaiz, M. (2016). A note on nonparametric estimation of copula-based multivariate extensions of Spearman’s rho.Statistics & Probability Letters,112, 41–50

2007
[13]

Scarsini, M. (1984). On measures of concordance.Stochastica,8(3), 201–218

1984
[14]

and Schmidt, R

Schmid, F. and Schmidt, R. (2007). Multivariate extensions of Spearman’s rho and related statis- tics.Statistics & Probability Letters,77(4), 407–416

2007
[15]

and Wolff, E.F

Schweizer, B. and Wolff, E.F. (1981). On nonparametric measures of dependence for random variables.The Annals of Statistics,9(4), 879–885

1981
[16]

Sklar, A.(1959).Fonctionsderépartitionàndimensionsetleursmarges.Publications de l’Institut de Statistique de l’Université de Paris,8, 229–231. 8

1959