Pith. sign in

REVIEW 2 major objections 5 minor 24 references

Pricing and Calibration of VIX Derivatives in Mixed Bergomi Models via Quantisation

T0 review · 2 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Vector quantisation makes daily VIX Bergomi calibration feasible.

desk verdict Solid quantisation speed-up for Bergomi VIX pricing, but the corridor-only calibration objective leaves parameters unidentified, so the stability claims need to be reframed. read the letter →

arxiv 2506.23409 v1 pith:OGLN7HSU submitted 2025-06-29 q-fin.PR

Nelson Kyakutwika , Mesias Alfeus , Erik Schlögl This is my paper · ORCID
classification q-fin.PR MSC 91G2091G60
keywords MixedBergomimodelsvectorquantisationVIXfuturesoptionsjointcalibrationparameterstabilityforwardvariancecurvemodel
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper argues that vector quantisation—replacing continuous Gaussian drivers by precomputed finite grids—is accurate and fast enough to price and calibrate mixed one- and two-factor Bergomi models to daily VIX futures and options. The authors report relative pricing errors below $0.01\%$ in the one-factor model and below $2\%$ in the two-factor model against exact quadrature, with a two-fold speedup in the one-factor model and roughly a 120-fold speedup in the two-factor model. This speed makes feasible a joint slice-by-slice calibration over 105 trading days and 1,213 maturity slices of VIX market data, which the paper then uses to assess in-sample fit and parameter stability. Both models fit the observed VIX futures and calls nearly perfectly, and the mixed one-factor model appears sufficient if VIX derivatives are the only target. The practical consequence is that daily recalibration of mixed Bergomi models to VIX derivatives is computationally within reach.

What carries the argument

The load-bearing object is the $L^2$-optimal $N$-point Gaussian quantiser: a fixed grid $\{y_1,\ldots,y_N\}$ with Voronoi probabilities $p_j$ that replaces a continuous Gaussian random variable by a discrete proxy. In the one-factor model, the Ornstein-Uhlenbeck state $X_t$ is scaled from a 1,000-point univariate normal quantiser; in the two-factor model, a 1,450-point bivariate normal quantiser is transformed by Cholesky decomposition to match the correlation of the two OU drivers. Because $V IX^2_{T_i}$ is a deterministic function of the quantised state, the VIX futures price becomes $\sum_j \sqrt{V^j_i}\,p_j$ and call/put prices become analogous sums of payoffs, with a 20-node Gauss-Legendre rule handling the time integral. These precomputed grids and simple weighted sums, rather than multi-dimensional numerical integration, are what deliver the speed.

What would settle it

Take one daily maturity slice and rerun the same calibration from many random starting parameter vectors, recording the final objective value and the returned $(\gamma_T, \omega_T^1, \omega_T^2)$. If the objective remains at the same near-zero level while the parameters scatter widely within the bid-ask corridor, the point estimates and their time-series stability are not identified, independent of how fast quantisation makes the pricing.

Watch

Extended reading notes

Core claim

The central claim is that vector quantisation can replace exact quadrature as the pricing engine for mixed Bergomi models without sacrificing accuracy. Using a 1,000-point univariate normal grid for the one-factor model and a 1,450-point bivariate normal grid for the two-factor model, the authors compute VIX futures, calls, and puts as weighted sums over the grid; the resulting prices agree with exact quadrature to within $0.001\%$, $0.007\%$, and $0.00035\%$ (one-factor) and $0.1\%$, $1.5\%$, and $0.4\%$ (two-factor) for futures, calls, and puts, respectively, while running twice as fast in the one-factor model and about 120 times as fast in the two-factor model. On this basis the paper jointly calibrates the models, maturity slice by maturity slice, to 105 daily VIX futures and call surfaces with maturities from one week to nine months. The calibration errors are tiny—mean relative futures errors around $3.5\times 10^{-5}$ and mean relative bid-ask errors for calls around $3\times 10^{-5}$—and the calibrated parameters show reasonable, though not perfect, stability over time.

Load-bearing premise

The calibration objective in Section 4.2 penalises only model call prices that fall outside the bid-ask corridor, so the reported point estimates of $\gamma_T$, $\omega_T^1$, and $\omega_T^2$ are not uniquely identified from the data; the apparent stability of the parameter paths could be an artifact of the optimizer's arbitrary choice inside that flat region.

Editorial extensions

If this is right

  • Daily recalibration of mixed one- and two-factor Bergomi models to VIX derivatives becomes practical, because quantisation is about twice as fast as exact quadrature in the one-factor model and about 120 times faster in the two-factor model.
  • Both models fit the observed VIX futures and call surfaces almost perfectly in single-day cross-sectional calibration, with mean relative errors on the order of $10^{-5}$.
  • If the objective is only to calibrate VIX futures and options, the mixed one-factor Bergomi model is sufficient; the two-factor model's extra factor buys only a marginal improvement.
  • The term structure of initial forward variance $\xi_0^T$ is the dominant driver of pricing accuracy, so refreshing $\xi_0^T$ daily while keeping the other parameters fixed preserves most of the fit to VIX futures.
  • When calibrated parameters are held fixed, out-of-sample pricing errors stabilise after about six days rather than growing, suggesting the models capture VIX derivative dynamics over at least a month.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper leaves implicit that the reported parameter stability is conditional on the flat bid-ask corridor objective; a strictly convex penalty term would be needed to identify $\gamma_T$, $\omega_T^1$, and $\omega_T^2$ uniquely.
  • Because the Gaussian quantiser grids are precomputed once and reused, the speed advantage should carry over to real-time pricing, risk management, and scenario simulation for VIX derivatives, not just calibration.
  • A natural extension would be to apply the same grid-quantisation machinery to other Markovian forward-variance models, or to use adaptive or recursive quantisation to refine the grid in regions that matter for deep out-of-the-money strikes.
  • The dominance of $\xi_0^T$ suggests a practical workflow: strip the forward variance term structure from VIX futures daily, and recalibrate the smile parameters on a slower cadence such as weekly.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper applies vector quantisation (precomputed optimal Gaussian grids) to price VIX futures and options in mixed one- and two-factor Bergomi models. It benchmarks quantisation against exact quadrature, reports speedups, and then calibrates the per-maturity parameters (γ_T, ω_T^1, ω_T^2, ξ_0^T) slice-by-slice to 105 days of CBOE VIX futures and call data using a corridor-type loss. It evaluates static fit and parameter stability, including four fixed-parameter out-of-sample tests. The authors conclude that both models fit VIX derivatives with near-perfect in-sample accuracy and satisfactory parameter stability, and that ξ_0^T is the dominant driver of pricing accuracy.

Significance. If the numerical claims are correct, the paper offers a practical speedup for recalibrating mixed Bergomi models to VIX derivatives: the accuracy tests are benchmarked against exact quadrature with small relative errors, and the calibration over 1,213 maturity slices is a useful empirical exercise. The proposed one-step parametrisation of the two-factor model is also convenient. The paper is honest about the static-vs-dynamic distinction and provides several out-of-sample tests with fixed parameters, which is a strength. The main weakness is that the corridor-only objective leaves per-slice parameters unidentified, so the stability assessment is not as solid as the pricing methodology. The paper also provides MATLAB code for the accuracy tests, which supports reproducibility of the numerical claims.

major comments (2)
  1. [Section 4.2, objective function] The call-option term is a sum of squared positive deviations outside the bid-ask corridor, so it vanishes identically whenever all model call prices fall inside the corridor. With four free parameters per slice (γ_T, ω_T^1, ω_T^2, ξ_0^T) and only the single VIX futures equation active on that region, the minimizer is generically non-unique; fmincon then returns a point determined by starting values and algorithm details. Consequently, the daily parameter paths in Section 5.1 (Figures 9-12) and the Abstract/Conclusion statements about 'satisfactory stability' describe an arbitrary selection within the flat region rather than identified model parameters. This does not affect the quantisation speed/accuracy claim, but it is load-bearing for the dynamic-stability assessment. Please remedy by reporting corridor widths and multi-start dispersion, using a mid-price or penalty-anchored objective, or substantially qualifying the stability claims.
  2. [Equations (12)-(14) and Tables 2-4] The reported ARBAE and RBAE measure only violations outside the bid-ask corridor, not deviations from mid-market prices. Near-zero values therefore mean that model prices lie within the bid-ask bounds, which is weaker than 'near-perfect fits' to market prices. Since the Conclusion states that both models achieve near-perfect fits to VIX futures and calls, the corridor-only metrics overstate static performance. Please supplement with mid-price errors, such as root mean squared relative error against mid quotes, or at least report the fraction of quotes priced inside the corridor.
minor comments (5)
  1. [Figure 8] The vertical axis label '9T0' appears to be a rendering error; it should read ξ_0^T.
  2. [Section 3.1] The word 'calender' is a typo and should be 'calendar'.
  3. [Section 1] The phrase 'In the literature:' contains a stray colon and should be 'In the literature,'.
  4. [Reference [16]] The reference is listed as 'Eduardo Abi Jaber et al.' without full author details; please complete the bibliographic entry.
  5. [Section 3.4] To make the 2x and 120x speedups reproducible, please report the hardware, MATLAB version, quadrature tolerance settings, and number of timing repetitions.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the pricing benchmark is independent exact quadrature, and the stability claims rest on out-of-sample fixed-parameter tests.

full rationale

The paper's derivation chain is self-contained. Quantisation accuracy is assessed against exact quadrature, which is an external numerical benchmark independent of the calibrated parameters, and the reported speed gains are direct computational measurements rather than consequences of the model equations. The in-sample calibration fit is explicitly presented as calibration accuracy, not as an out-of-sample prediction, so using the same relative-error metrics in the objective and in the summary statistics is standard fitting behaviour rather than a reduction by construction. The parameter-stability claims are supported by Tests 1 to 4, which fix the calibrated parameters for up to 30 trading days and price subsequent daily surfaces, with only the initial forward variance refreshed either by the model-independent replication formula (15) or by recalibration; this is genuine out-of-sample pricing relative to the fitted parameters. The use of equation (15) to strip the initial forward variance from VIX futures and options is a market-data input, and the model futures price is then computed from the full distribution of VIX, so the futures match is not forced by construction. The self-citations to Bergomi's mixed-model constructions and to similar objective functions in the literature are background modelling choices, not load-bearing justifications that import the paper's conclusions. The flatness of the calibration objective inside the bid-ask corridor noted in the reader's take is an identifiability concern about non-unique optima, not a circularity: it does not make any derived quantity equivalent to its own input. No equation in the paper is defined in terms of the quantity it is used to predict, and no fitted parameter is renamed as an independent prediction.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claims rest on standard risk-neutral forward variance machinery plus per-slice calibrated parameters. The fixed model parameters (k, k1, k2, theta, rho) are inputs from prior literature, not derived in this paper. No new theoretical entities, forces, or particles are introduced.

free parameters (6)
  • gamma_T (mixing weight, one- and two-factor) = e.g. 0.9154 for one-factor, T=7 days, 09 Apr 2024; varies by slice and day
    Calibrated per maturity slice in Section 4.2; sets the convex combination in Eq. (3) and Eq. (5).
  • omega_T^1 (vol-of-vol, first component) = e.g. 17.98 for one-factor, T=7 days, 09 Apr 2024
    Calibrated per slice; drives smile steepness.
  • omega_T^2 (vol-of-vol, second component) = e.g. 1.28 for one-factor, T=7 days, 09 Apr 2024
    Calibrated per slice; controls smile convexity.
  • xi_0^T (initial forward variance) = e.g. 0.0234 for one-factor, T=7 days, 09 Apr 2024; also stripped from market via Eq. (15)
    Calibrated per slice in Section 4, or replicated from VIX market in Section 5; dominant pricing driver.
  • k (one-factor mean reversion) = 1
    Fixed a priori following Bourgey et al.; not calibrated here.
  • k1, k2, theta, rho (two-factor structure) = 7.54, 0.24, 0.23, 0.7 (Set III from Bergomi)
    Fixed a priori; not calibrated to VIX derivatives in this study. The fit quality depends on this choice.
assumptions (5)
  • domain assumption The forward variance curve xi_t^T is a martingale in t under the risk-neutral measure Q.
    Invoked when writing dxi_t^T with zero drift in Sections 2.1 and 2.2; standard Bergomi model construction.
  • domain assumption VIX squared is the integral of forward variances over the next 30 days, VIX_t^2 = (100^2 / Delta) * integral from t to t+Delta of xi_t^T dT.
    Eq. (8) and (9); connects the model to the market VIX and relies on the model-free replication formula.
  • domain assumption The optimal quantisers for standard normal and bivariate normal distributions downloaded from the Quantisation Website are accurate enough for pricing.
    Used in Section 3.4; accuracy is verified numerically against quadrature, so this is supported by the paper's tests.
  • domain assumption Parameters gamma_T, omega_T^1, omega_T^2, and xi_0^T are constant on the interval [T_i, T_i + Delta] for each calibration slice.
    Section 4.2; needed to compute VIX at maturity T_i from forward variances over the 30-day window.
  • domain assumption Market option prices satisfy put-call parity, so the risk-free rate r and futures price F can be extracted by least squares.
    Section 4.1; used to obtain rates and futures prices from bid and ask quotes.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Pricing and Calibration of VIX Derivatives in Mixed Bergomi Models via Quantisation." pith.science (2026). https://pith.science/paper/OGLN7HSU

@misc{pith2026250623409,
  author       = {Pith},
  title        = {Pith review of: Pricing and Calibration of VIX Derivatives in Mixed Bergomi Models via Quantisation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OGLN7HSU}},
  note         = {Machine review of arXiv:2506.23409}
}
read the original abstract

We apply vector quantisation within mixed one- and two-factor Bergomi models to implement a fast and efficient approach for option pricing in these models. This allows us to calibrate such models to market data of VIX futures and options. Our numerical tests confirm the efficacy of vector quantisation, making calibration feasible over daily data covering several months. This permits us to evaluate the calibration accuracy and the stability of the calibrated parameters, and we provide a comprehensive assessment of the two models. Both models show excellent performance in fitting VIX derivatives, and their parameters show satisfactory stability over time.

Figures

Figures reproduced from arXiv: 2506.23409 by the authors.

Figure 1
Figure 1. Top: VIX futures term structure in mixed Bergomi models, exact quadrature (blue) versus quantisation (red). Bottom: Relative errors between exact quadrature and quantisation. k1 = 7.54, k2 = 0.24, ρ = 0.7, θ = 0.23, γ T = 0.60, ω T 1 = 9.12, and ω T 2 = 1.10. These parameters are used in all the Figures 1, 2, and 3. As shown in Figures 1, 2, and 3, quantisation achieves high accuracy levels. For the one-factor model… view at source ↗
Figure 2
Figure 2. Top: VIX call prices in mixed Bergomi models, exact quadrature (blue) versus quantisation (red). Bottom: Relative errors between exact quadrature and quantisation. the VIX future and call prices in [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 3
Figure 3. Top: VIX put prices in mixed Bergomi models, exact quadrature (blue) versus quantisation (red). Bottom: Relative errors between exact quadrature and quantisation. 4 Joint calibration of VIX futures and options 4.1 Dataset The dataset consists of the daily bid and ask quotes on VIX call and put options over 105 trading days, from 2 January 2024 to 31 May 2024. We apply some of the standard exclusion filters: we remov… view at source ↗
Figures from the paper (11 more)
Figure 4
Figure 4. Figure 4: Futures (vertical lines) and smiles from the joint calibration to VIX futures and [PITH_FULL_IMAGE:figures/full_fig_p016_4.png]
Figure 5
Figure 5. Figure 5: Futures (vertical lines) and smiles from the joint calibration to VIX futures and [PITH_FULL_IMAGE:figures/full_fig_p016_5.png]
Figure 6
Figure 6. Figure 6: VIX futures term structure from the joint calibration to VIX futures and calls [PITH_FULL_IMAGE:figures/full_fig_p017_6.png]
Figure 7
Figure 7. Figure 7: Evolution of the 30-day moving average calibration error for the joint calibration to VIX futures and calls. Left: VIX futures. Right: VIX calls. purposes of pricing VIX futures and calls, we replicate ξ T 0 using the VIX market. For constant ξ T 0 over the interval [T…
Figure 8
Figure 8. Figure 8: Term structure of initial forward variances as of [PITH_FULL_IMAGE:figures/full_fig_p020_8.png]
Figure 9
Figure 9. Figure 9: Term structures of calibrated parameters in the one-factor model. [PITH_FULL_IMAGE:figures/full_fig_p022_9.png]
Figure 10
Figure 10. Figure 10: Term structures of calibrated parameters in the two-factor model. [PITH_FULL_IMAGE:figures/full_fig_p022_10.png]
Figure 11
Figure 11. Figure 11: Evolution of calibrated parameters at different maturities in the one-factor [PITH_FULL_IMAGE:figures/full_fig_p022_11.png]
Figure 12
Figure 12. Figure 12: Evolution of calibrated parameters at different maturities in the two-factor [PITH_FULL_IMAGE:figures/full_fig_p023_12.png]
Figure 13
Figure 13. Figure 13: Boxplots illustrating the daily average relative error distributions for one [PITH_FULL_IMAGE:figures/full_fig_p024_13.png]
Figure 14
Figure 14. Figure 14: Boxplots illustrating the daily average relative error distributions for one [PITH_FULL_IMAGE:figures/full_fig_p025_14.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 23 canonical work pages

  1. [1]

    http://www.quantize.maths-fi.com/gaussian_database, [Online; accessed April 2024]

    QuantizationWebsite. http://www.quantize.maths-fi.com/gaussian_database, [Online; accessed April 2024]

    work page 2024

  2. [2]

    Joint SPX & VIX calibration with Gaussian polynomial volatility models: Deep pricing with quantization hints

    Eduardo Abi Jaber, Camille Illand, and Shaun Li. Joint SPX & VIX calibration with Gaussian polynomial volatility models: Deep pricing with quantization hints. Mathematical Finance, 2022

    work page 2022

  3. [3]

    A consistent stochastic model of the term structure of interest rates for multiple tenors

    Mesias Alfeus, Martino Grasselli, and Erik Schlögl. A consistent stochastic model of the term structure of interest rates for multiple tenors. Journal of Economic Dynamics and Control, 114:103861, 2020

    work page 2020

  4. [4]

    Smile dynamics II.Risk, pages 67–73, October 2005

    Lorenzo Bergomi. Smile dynamics II.Risk, pages 67–73, October 2005

    work page 2005

  5. [5]

    Smile dynamics III.Risk, pages 90–96, October 2008

    Lorenzo Bergomi. Smile dynamics III.Risk, pages 90–96, October 2008

    work page 2008

  6. [6]

    Stochastic volatility modeling

    Lorenzo Bergomi. Stochastic volatility modeling. CRC press, 2015

    work page 2015

  7. [7]

    Weak approximations and VIX option price expansions in forward variance curve models.Quantitative Finance, 23(9):1259–1283, 2023

    Florian Bourgey, Stefano De Marco, and Emmanuel Gobet. Weak approximations and VIX option price expansions in forward variance curve models.Quantitative Finance, 23(9):1259–1283, 2023

    work page 2023

  8. [8]

    Pricing and calibration in local volatility models via fast quantization.Available at SSRN 2495829, 2014

    Giorgia Callegaro, Lucio Fiorin, and Martino Grasselli. Pricing and calibration in local volatility models via fast quantization.Available at SSRN 2495829, 2014

    work page 2014

Show all 24 references
  1. [9]

    Pricing via recursive quan- tization in stochastic volatility models.Quantitative Finance, 17(6):855–872, 2017

    Giorgia Callegaro, Lucio Fiorin, and Martino Grasselli. Pricing via recursive quan- tization in stochastic volatility models.Quantitative Finance, 17(6):855–872, 2017

    2017

  2. [10]

    Towards a theory of volatility trading.Option Pric- ing, Interest Rates and Risk Management, Handbooks in Mathematical Finance, 22(7):458–476, 2001

    Peter Carr and Dilip Madan. Towards a theory of volatility trading.Option Pric- ing, Interest Rates and Risk Management, Handbooks in Mathematical Finance, 22(7):458–476, 2001

    2001

  3. [11]

    Theinformation contentoftheS&P500indexandVIXoptionsonthedynamicsoftheS&P500index

    San-LinChung, Wei-CheTsai, Yaw-HueiWang, andPei-ShihWeng. Theinformation contentoftheS&P500indexandVIXoptionsonthedynamicsoftheS&P500index. Journal of Futures Markets, 31(12):1170–1201, 2011

    2011

  4. [12]

    Joint cal- ibration to SPX and VIX options with signature-based models.Mathematical Fi- nance, 2023

    Christa Cuchiero, Guido Gazzani, Janka Möller, and Sara Svaluto-Ferro. Joint cal- ibration to SPX and VIX options with signature-based models.Mathematical Fi- nance, 2023

    2023

  5. [13]

    The VIX future in Bergomi models: Fast approximation formulas and joint calibration with S&P 500 skew.SIAM Journal on Financial Mathematics, 13(4):1418–1485, 2022

    Julien Guyon. The VIX future in Bergomi models: Fast approximation formulas and joint calibration with S&P 500 skew.SIAM Journal on Financial Mathematics, 13(4):1418–1485, 2022. 27

    2022

  6. [14]

    Dispersion-constrainedmartingaleschrödingerproblemsandtheexact joint S&P 500/VIX smile calibration puzzle.Finance and Stochastics, 28(1):27–79, 2024

    JulienGuyon. Dispersion-constrainedmartingaleschrödingerproblemsandtheexact joint S&P 500/VIX smile calibration puzzle.Finance and Stochastics, 28(1):27–79, 2024

    2024

  7. [15]

    Does the term-structure of equity at-the-money skew really follow a power law?Risk, pages 1–6, August 2023

    Julien Guyon and Mehdi El Amrani. Does the term-structure of equity at-the-money skew really follow a power law?Risk, pages 1–6, August 2023

    2023

  8. [16]

    Volatility models in practice: Rough, Path-dependent or Markovian? arXiv preprint arXiv:2401.03345, 2024

    Eduardo Abi Jaber et al. Volatility models in practice: Rough, Path-dependent or Markovian? arXiv preprint arXiv:2401.03345, 2024

    2024 arXiv

  9. [17]

    The quintic Ornstein-Uhlenbeck volatility model that jointly calibrates SPX & VIX smiles.Risk, pages 1–6, July 2023

    Eduardo Abi Jaber, Camille Illand, et al. The quintic Ornstein-Uhlenbeck volatility model that jointly calibrates SPX & VIX smiles.Risk, pages 1–6, July 2023

    2023

  10. [18]

    Joint pricing of VIX and SPX options with stochastic volatility and jump models

    Thomas Kokholm and Martin Stisen. Joint pricing of VIX and SPX options with stochastic volatility and jump models. The Journal of Risk Finance, 16(1):27–48, 2015

    2015

  11. [19]

    Callegaro O

    G. Callegaro O. Bonesini and A. Jacquier. Functional quantization of rough volatility and applications to volatility derivatives.Quantitative Finance, 23(12):1769–1792, 2023

    2023

  12. [20]

    Forward variance dynamics: Bergomi’s model revisited

    Sidi Mohamed Ould Aly. Forward variance dynamics: Bergomi’s model revisited. Applied Mathematical Finance, 21(1):84–107, 2014

    2014

  13. [21]

    Optimal quadratic quantization for numerics: the Gaussian case.Monte Carlo Methods Appl., 9(2):135–165, 2003

    Gilles Pagès and Jacques Printems. Optimal quadratic quantization for numerics: the Gaussian case.Monte Carlo Methods Appl., 9(2):135–165, 2003

    2003

  14. [22]

    Recursive marginal quantization of the Euler scheme of a diffusion process.Applied Mathematical Finance, 22(5):463–498, 2015

    Gilles Pagès and Abass Sagna. Recursive marginal quantization of the Euler scheme of a diffusion process.Applied Mathematical Finance, 22(5):463–498, 2015

    2015

  15. [23]

    Empirical analysis of rough and classical stochastic volatility models to the SPX and VIX markets.Quantitative Finance, 22(10):1805–1838, 2022

    Sigurd Emil Rømer. Empirical analysis of rough and classical stochastic volatility models to the SPX and VIX markets.Quantitative Finance, 22(10):1805–1838, 2022

    2022

  16. [24]

    Legendre-Gauss quadrature weights and nodes.Matlab function lgwt

    Greg Von Winckel. Legendre-Gauss quadrature weights and nodes.Matlab function lgwt. URL: http://www. mathworks. com/matlabcentral/fileexchange/4540 Wessel P, Smith WHF (1998), New, improved version of the Generic Mapping Tools released, EOS Trans. AGU, 79:579, 2004. 28

    1998

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.