Paper: arXiv 2410.04487

Authors: Xiaoyu Shen, Fang Fang, Chengguang Liu

Abstract

We provide a rigorous convergence proof demonstrating that the well-known semi-analytical Fourier cosine (COS) formula for the inverse Fourier transform of continuous probability distributions can be extended to discrete probability distributions, with the help of spectral filters. We establish general convergence rates for these filters and further show that several classical spectral filters achieve convergence rates one order faster than previously recognized in the literature on the Gibbs phenomenon. Our numerical experiments corroborate the theoretical convergence results. Additionally, we illustrate the computational speed and accuracy of the discrete COS method with applications in computational statistics and quantitative finance. The theoretical and numerical results highlight the method’s potential for solving problems involving discrete distributions, particularly when the characteristic function is known, allowing the discrete Fourier transform (DFT) to be bypassed.

Complexity vs Empirical Score

  • Math Complexity: 8.5/10
  • Empirical Rigor: 4.0/10
  • Quadrant: Lab Rats — high math complexity, limited empirical testing

Why this score: The paper features advanced mathematical proofs involving spectral filters, Gibbs phenomenon, and convergence rates, indicating high math complexity. While it includes numerical experiments and applications in quantitative finance, the emphasis is on theoretical convergence proofs rather than extensive backtesting or real-world data implementation, resulting in moderate empirical rigor.

Research Flowchart

  flowchart TD
  A["Research Goal: Extend COS Method to Discrete Distributions"] --> B["Methodology: Extend with Spectral Filters"]
  B --> C["Input: Characteristic Function of Discrete Distribution"]
  C --> D["Computational Process: Discrete COS Method"]
  D --> E["Outcome 1: Rigorous Convergence Proof & Fast Rates"]
  D --> F["Outcome 2: Bypassing DFT for Speed"]
  E --> G["Validation: Numerical Experiments & Applications"]
  F --> G

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 provides a rigorous convergence proof extending the semi‑analytical Fourier cosine (COS) formula to discrete probability distributions using spectral filters and establishes faster convergence rates for several classical filters.

Source sentence
We provide a rigorous convergence proof demonstrating that the well-known semi-analytical Fourier cosine (COS) formula for the inverse Fourier transform of continuous probability distributions can be extended to discrete probability distributions, with the help of spectral filters. We establish general convergence rates for these filters and further show that several classical spectral filters achieve convergence

Assumes 2

  • The discrete distribution has finite support. model
    Source sentence
    Throughout this paper, we assume that the discrete distribution has finite support.
  • The CDF is right‑continuous at 0 and left‑continuous at π, implying no probability mass at the interval endpoints. statistical
    Source sentence
    In Theorem 2.3, we assume that the CDF FX is right-continuous at 0 and left-continuous at π.

Data

  • Poisson‑binomial distribution synthetic unknown
    Source sentence
    In a one-step calculation, we use COS to perform discrete Fourier inversion, and successfully recover the CDF of Poisson-binomial (PB) and generalized Poisson-binomial (GPB) distributions and a Hawkes process.
  • Generalized Poisson‑binomial distribution synthetic unknown
    Source sentence
    In a one-step calculation, we use COS to perform discrete Fourier inversion, and successfully recover the CDF of Poisson-binomial (PB) and generalized Poisson-binomial (GPB) distributions and a Hawkes process.
  • Hawkes process synthetic unknown
    Source sentence
    In a one-step calculation, we use COS to perform discrete Fourier inversion, and successfully recover the CDF of Poisson-binomial (PB) and generalized Poisson-binomial (GPB) distributions and a Hawkes process.

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • The semi‑analytical COS formula converges for discrete distributions, with a proven convergence rate of O(K^{-p}) for several classical spectral filters, which is one order faster than the previously reported O(K^{1-p}). theoretical
    Source sentence
    Specifically, we find that several classical spectral filters, including the Lanczos filter, raised cosine filter, sharpened raised cosine filter, second order exponen-tial filters, exhibit a convergence rate that is one order higher than previously reported in [30]. However, in this paper, we demonstrate that the aforementioned spectral filters actually converge at a rate of O(K−p).
  • Numerical experiments confirm the theoretical convergence rates. statistical
    Source sentence
    Our numerical experiments corroborate the theoretical convergence results.
  • The discrete COS method accurately recovers the CDFs of Poisson‑binomial, generalized Poisson‑binomial, and Hawkes process distributions in one‑step calculations. performance
    Source sentence
    In a one-step calculation, we use COS to perform discrete Fourier inversion, and successfully recover the CDF of Poisson-binomial (PB) and generalized Poisson-binomial (GPB) distributions and a Hawkes process.
  • Spectral filters can be applied analytically to COS coefficients without extra computational cost. theoretical
    Source sentence
    Note that the use of spectral filters only requires modifying the COS coefficients analytically.
  • The method enables efficient semi‑analytical solutions for problems involving discrete probability distributions, such as approximating CDFs and computing moments. theoretical
    Source sentence
    The established theoretical convergence, supported by numerical examples, highlights the potential of using the COS method to find efficient semi-analytical solutions to problems involving discrete probability distributions, such as approx-imating CDFs and computing moments.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 30 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 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.

Cited References (31)

  1. A Novel Pricing Method for European Options Based on Fourier-Cosine Series Expansions (2009) — link
  2. Pricing early-exercise and discrete barrier options by fourier-cosine series expansions (2009) — link
  3. A Fourier-Based Valuation Method for Bermudan and Barrier Options under Heston’s Model (2011) — link
  4. Efficient Pricing of European-Style Asian Options under Exponential Lévy Processes Based on Fourier Cosine Expansions (2013) — link
  5. Reconciling Rough Volatility with Jumps — → 2303.07222
  6. Lifting the Heston model (2019) — link
  7. Deep xVA Solver: A Neural Network–Based Counterparty Credit Risk Management Framework (2023) — link
  8. The rough Hawkes Heston stochastic volatility model (2024) — link
and 23 more
  1. Short-Term Market Risks Implied by Weekly Options (2017) — link
  2. Fourier-Cosine Method for Finite-Time Gerber–Shiu Functions (2021) — link
  3. Valuation of guaranteed minimum maturity benefits under generalised regime-switching models using the Fourier Cosine method (2022) — link
  4. A Fourier Cosine Method for an Efficient Computation of Solutions to BSDEs (2015) — link
  5. Efficient numerical Fourier methods for coupled forward–backward SDEs (2016) — link
  6. On the Fourier cosine series expansion method for stochastic control problems (2013) — link
  7. On the application of spectral filters in a Fourier option pricing technique (2015) — link
  8. The Heston–Queue-Hawkes process: A new self-exciting jump–diffusion model for options pricing, and an extension of the COS method for discrete distributions (2025) — link
  9. Filters, Mollifiers and the Computation of the Gibbs Phenomenon (2007) — link
  10. Family of Spectral Filters for Discontinuous Problems (1991) — link
  11. On the Gibbs Phenomenon and Its Resolution (1997) — link
  12. Chebyshev and Fourier Spectral Methods (2001)
  13. Spectral Methods for Time-Dependent Problems (2007) — link
  14. On the Gibbs phenomenon I: recovering exponential accuracy from the Fourier partial sum of a nonperiodic analytic function (1992) — link
  15. Two-Dimensional Fourier Cosine Series Expansion Method for Pricing Financial Options (2012) — link
  16. Precise option pricing by the COS method—How to choose the truncation range (2022) — link
  17. On the number of terms in the COS method for European option pricing — → 2303.16012
  18. On computing the distribution function for the Poisson binomial distribution (2013)
  19. A simple and fast method for computing the Poisson binomial distribution function (2018)
  20. Closed-Form Expression for the Poisson-Binomial Probability Density Function (2010)
  21. The generalized Poisson-binomial distribution and the computation of its distribution function (2018)
  22. Statistical applications of the Poisson-Binomial and conditional Bernoulli distributions (1997)
  23. Affine Point Processes and Portfolio Credit Risk (2020) — link

2 of 31 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.