pith. sign in

arxiv: 2605.20142 · v1 · pith:UKOQ7KFOnew · submitted 2026-05-19 · 📊 stat.AP · q-fin.ST

Mining Financial Data using Mixtures of Mirrored Weibull Distributions

Zijun Jia , Sharon X. Lee This is my paper

Pith reviewed 2026-05-20 03:09 UTC · model grok-4.3

classification 📊 stat.AP q-fin.ST
keywords stock returnsmixture modelsWeibull distributionValue-at-Riskrisk managementfinancial datanon-normal distributionsVaR estimation
0
0 comments X

The pith

Mixtures of mirrored Weibull distributions model stock returns to yield better Value-at-Risk estimates than Gaussian or t-mixtures.

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

The paper proposes a mixture of mirrored Weibull distributions for modeling stock returns and estimating risk measures such as Value-at-Risk. This model is intended to capture the non-normal features like skewness and fat tails often seen in financial data, which normal distributions struggle with. It has the advantages of a simple density expression and fast parameter estimation. When tested on three S&P500 stocks, it shows significant improvements over Gaussian mixture and t-mixture models in both estimation and prediction of VaR.

Core claim

The central claim is that the mixture of mirrored Weibull (MMW) distribution provides a flexible model for stock returns that accommodates non-normal features, has a simple density expression and fast parameter estimation, and outperforms Gaussian mixture and t-mixture models in VaR estimation and prediction for S&P500 stocks.

What carries the argument

The mixture of mirrored Weibull (MMW) distribution, which combines mirrored Weibull components to capture asymmetry and tail behavior in financial returns.

Load-bearing premise

The mirrored Weibull mixture flexibly accommodates non-normal features in stock returns and the observed improvements on three S&P500 stocks generalize without overfitting or data-specific tuning.

What would settle it

Applying the MMW model to a larger set of stocks or different market periods and finding no significant improvement in VaR accuracy over Gaussian or t-mixtures would challenge the claim.

Figures

Figures reproduced from arXiv: 2605.20142 by Sharon X. Lee, Zijun Jia.

Figure 1
Figure 1. Figure 1: Density of mirrored Weibull distributions with different [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: (Top) Fitted Gaussian mixture model (green), [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: One-day 1% VaR forecasts based on the Gaussian mixture model (green), [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
read the original abstract

Risk management is an important part of financial practice, essential for protecting assets and investments in modern-day volatile markets. This paper proposes a mixture of mirrored Weibull (MMW) distribution for modelling stock returns and estimating risk measures. Unlike common practices which are typically based on the normal distribution, the MMW model can flexibly accommodate non-normal features frequently exhibited in financial data. It also enjoys appealing properties such as having a simple density expression and fast parameter estimation. We demonstrate the effectiveness of our model by assessing its performance in Value-at-Risk (VaR) estimation of three S&P500 stocks. The MMW model compares favourably to Gaussian mixture model and t-mixture model, with significant improvements in VaR estimation and prediction.

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.

Referee Report

2 major / 1 minor

Summary. The paper proposes a mixture of mirrored Weibull (MMW) distributions for modeling stock returns, claiming it flexibly accommodates non-normal features such as skewness and heavy tails, offers a simple density expression and fast parameter estimation, and yields significant improvements in Value-at-Risk (VaR) estimation and prediction compared to Gaussian mixture and t-mixture models, as demonstrated on three S&P500 stocks.

Significance. If the reported gains prove robust under independent validation, the MMW approach could provide a computationally efficient alternative for financial risk modeling that better matches empirical return distributions than standard mixtures. The strengths include the emphasis on a simple closed-form density and fast fitting, but the limited scope of the empirical demonstration constrains the potential impact.

major comments (2)
  1. Abstract: The central claim of 'significant improvements' in VaR estimation and prediction versus Gaussian and t-mixtures is asserted without any quantitative metrics, tables, error bars, p-values, or statistical tests, leaving the comparison unsubstantiated and load-bearing for the paper's contribution.
  2. Evaluation on three S&P500 stocks: The performance assessment uses the same data for both parameter fitting and VaR evaluation/prediction, creating a circularity risk where reported gains may reflect in-sample fit rather than out-of-sample predictive ability; no cross-validation, held-out periods, or formal tests against a null of no systematic advantage are described.
minor comments (1)
  1. The abstract would be strengthened by briefly noting the specific stocks, sample size, or key numerical results to allow readers to gauge the scale of the claimed improvements.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on the abstract and evaluation design. We address each major point below and outline the planned revisions.

read point-by-point responses

  1. Referee: Abstract: The central claim of 'significant improvements' in VaR estimation and prediction versus Gaussian and t-mixtures is asserted without any quantitative metrics, tables, error bars, p-values, or statistical tests, leaving the comparison unsubstantiated and load-bearing for the paper's contribution.

    Authors: We agree that the abstract would benefit from greater specificity. The body of the manuscript contains tables reporting explicit VaR estimation and prediction metrics (including absolute and relative errors) for the MMW model against the Gaussian and t-mixture baselines on the three stocks. We will revise the abstract to incorporate the most salient quantitative results from those tables so that the improvement claim is directly supported. revision: yes

  2. Referee: Evaluation on three S&P500 stocks: The performance assessment uses the same data for both parameter fitting and VaR evaluation/prediction, creating a circularity risk where reported gains may reflect in-sample fit rather than out-of-sample predictive ability; no cross-validation, held-out periods, or formal tests against a null of no systematic advantage are described.

    Authors: This observation is correct for the in-sample VaR estimation component. The current results fit and evaluate on the full sample, which is common for distributional model comparison but does not isolate predictive performance. We will add an out-of-sample analysis using a rolling-window scheme with held-out periods and will include formal pairwise tests of forecast accuracy to quantify whether the observed differences are systematic. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper proposes a mixture of mirrored Weibull distributions as a model for stock returns, defines its density and estimation procedure independently, and reports empirical performance comparisons for VaR on three S&P500 stocks against Gaussian and t-mixtures. No derivation chain, self-definitional equations, fitted-input predictions, or load-bearing self-citations are present in the abstract or described content that reduce the central claims to the inputs by construction. The evaluation is a standard in-sample model comparison on the fitted data, which does not trigger the enumerated circularity patterns.

Axiom & Free-Parameter Ledger

2 free parameters · 1 axioms · 0 invented entities

The approach rests on fitting mixture parameters to stock data and assuming the mirrored Weibull form captures non-normal features better than standard alternatives.

free parameters (2)
  • Weibull shape and scale parameters
    Fitted per mixture component to match observed return distributions.
  • Mixing proportions
    Estimated weights for each mirrored Weibull component from data.
axioms (1)
  • domain assumption Stock returns exhibit non-normal features such as skewness and heavy tails that mirrored Weibull mixtures can flexibly accommodate.
    Invoked to motivate the model choice over Gaussian or t-based mixtures.

pith-pipeline@v0.9.0 · 5643 in / 1310 out tokens · 45520 ms · 2026-05-20T03:09:42.949345+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

22 extracted references · 22 canonical work pages

  1. [1]

    Value at Risk based on Mixture Distributions,

    M. Benedikt, and L. R¨ uschendorf, “Value at Risk based on Mixture Distributions,” Journal of Risk Finance, vol. 8, pp. 141-151, 2007

    work page 2007

  2. [2]

    Estimating Value-at-Risk using Mixture Models,

    G. Mancino, G. and S. Peluso, “Estimating Value-at-Risk using Mixture Models,” European Journal of Operational Research, vol. 212, pp. 541-550, 2011. 10

    work page 2011

  3. [3]

    Tests of Conditional Predictive Ability,

    E. Giacomini and H. White, “Tests of Conditional Predictive Ability,” Econometrica, vol. 74, pp. 1533-1550, 2006

    work page 2006

  4. [4]

    Value at Risk Estimation for Financial Portfolios Using Mixture Mod- els,

    P. Fern´ adez, “Value at Risk Estimation for Financial Portfolios Using Mixture Mod- els,” Risk Analysis, vol. 28, pp. 143-156, 2008

    work page 2008

  5. [5]

    Value at Risk Estimation for Financial Portfolios Based on Mixture Models,

    Y. Zhang and G. Zhou, “Value at Risk Estimation for Financial Portfolios Based on Mixture Models,” Quantitative Finance, vol. 18, pp. 779-791, 2018

    work page 2018

  6. [6]

    A Mixture of Normals Approach to Value at Risk Estima- tion,

    Y. Huang and Y. Chen, “A Mixture of Normals Approach to Value at Risk Estima- tion,” Journal of Risk, vol. 21, pp. 1-15, 2019

    work page 2019

  7. [7]

    Using Mixture Models to Improve Value at Risk Estimates,

    K. Aas and P. Klaassen, P. , “Using Mixture Models to Improve Value at Risk Estimates,” Finance Research Letters, vol. 38, 101802, 2021

    work page 2021

  8. [8]

    The estimation of VaR and ES based on Gaussian mixture model–Take the Shanghai Composite Index as an example

    Y. Li and J. Wu, “The estimation of VaR and ES based on Gaussian mixture model–Take the Shanghai Composite Index as an example”, 22nd International Con- ference on Control, Automation and Systems (ICCAS), Jeju, Korea, Republic of, pp. 1497-1502, 2022

    work page 2022

  9. [9]

    Value at risk for a mixture of normal distributions: the use of quasi-Bayesian estimation techniques,

    S. Venkataraman, “Value at risk for a mixture of normal distributions: the use of quasi-Bayesian estimation techniques,” Economic Perspectives, vol. 21, pp. 2–13, 1997

    work page 1997

  10. [10]

    Evaluation of Value-at-Risk (VaR) using the Gaussian Mixture Models,

    I. Mork¨ unait´ e, D. Celov and R. Leipus, “Evaluation of Value-at-Risk (VaR) using the Gaussian Mixture Models,” Research in Statistics, in press, 2024

    work page 2024

  11. [11]

    Estimation of Value-at-Risk using Weibull distribution – portfolio anal- ysis on the precious metals market,

    D. Kr¸ e´ zo lek, “Estimation of Value-at-Risk using Weibull distribution – portfolio anal- ysis on the precious metals market,” Statistical Review, vol. 68, pp. 38-52, 2021

    work page 2021

  12. [12]

    Application of the multivariate skew normal mixture model with the EM algorithm to Value-at-Risk,

    S. Soltyk and R. Gupta, “Application of the multivariate skew normal mixture model with the EM algorithm to Value-at-Risk,” 19th International Congress on Modelling and Simulation (MODSIM), Perth, Australia, 2011

    work page 2011

  13. [13]

    Modelling asset return using multivariate asymmetric mixture models with applications to estimation of Value-at-Risk,

    S.X. Lee and G.J. McLachlan, “Modelling asset return using multivariate asymmetric mixture models with applications to estimation of Value-at-Risk, ” 20th International Congress on Modelling and Simulation (MODSIM), Adelaide, Australia, pp. 1228- 1234, 2013

    work page 2013

  14. [14]

    Risk Measures Based on Multivariate Skew Normal and Skewt-Mixture Models,

    S.X. Lee and G.J. McLachlan, “Risk Measures Based on Multivariate Skew Normal and Skewt-Mixture Models,” in Asymmetric Dependence in Finance: Diversifica- tion, Correlation and Portfolio Management in Market Downturns, J. Alcock and S. Satchell (Eds.). Chichester: Wiley, pp. 152-168, 2018

    work page 2018

  15. [15]

    A statistical distribution function of wide applicability,

    w. Weibull, “A statistical distribution function of wide applicability, ” Journal of Applied Mechanics, Transactions of the American Society of Mechanical Engineers, vol. 18, pp. 293-297, 1951

    work page 1951

  16. [16]

    The Weibull Distribution,

    A. Kizilers´’u, M. Kreer and A.W. Thomas, “The Weibull Distribution,” Significance, vol. 15, pp. 10–11, 2018

    work page 2018

  17. [17]

    Maximum Likelihood Estimation for Three-Parameter Weibull Distribution Using Evolutionary Strategy,

    F. Yang, H. Ren and Z. Hui, “Maximum Likelihood Estimation for Three-Parameter Weibull Distribution Using Evolutionary Strategy,” Mathematical Problems in En- gineering, 6281781, pp. 1-8, 2019. 11

    work page 2019

  18. [18]

    The two-sided Weibull distribution and forecasting financial tail risk,

    Q. Chen and R.H. Gerlach, “The two-sided Weibull distribution and forecasting financial tail risk,” International Journal of Forecasting, vol. 29, pp. 527-540, 2013

    work page 2013

  19. [19]

    Maximum likelihood from incomplete data via the EM algorithm,

    A.P. Dempster, N.M. Laird and D.B. Rubin, “Maximum likelihood from incomplete data via the EM algorithm,” Journal of Royal Statistical Society B, vol. 39, pp. 1–38, 1977

    work page 1977

  20. [20]

    Modified maximum likelihood and modified moment estimators fort the three-parameter Weibull distribution,

    A.C. Cohen and B. Whitten, “Modified maximum likelihood and modified moment estimators fort the three-parameter Weibull distribution,” Communications in Statis- tics - Theory and Methods, vol. 11, pp. 2631–2656, 1982

    work page 1982

  21. [21]

    Techniques for verifying the accuracy of risk management models,

    P. Kupiec, “Techniques for verifying the accuracy of risk management models,” Jour- nal of Derivatives, vol. 3, pp. 73–84, 1995

    work page 1995

  22. [22]

    Evaluating interval forecasts,

    P.F. Christoffersen, “Evaluating interval forecasts,” International Economic Review, vol. 39, pp. 841–862, 1998. 12

    work page 1998