Faster Forward Sensitivities: Reduced stochastic hedge ratios from pathwise algorithmic differentiation

Christian P Fries

arxiv: 2605.23979 · v1 · pith:R6PRFGJ2new · submitted 2026-05-13 · 💱 q-fin.RM · q-fin.CP· q-fin.PR

Faster Forward Sensitivities: Reduced stochastic hedge ratios from pathwise algorithmic differentiation

Christian P Fries This is my paper

Pith reviewed 2026-06-30 20:58 UTC · model grok-4.3

classification 💱 q-fin.RM q-fin.CPq-fin.PR

keywords stochastic hedge ratiospathwise sensitivitiesMonte Carlo simulationalgorithmic differentiationreduced basisGalerkin projectionmargin valuation adjustmentliquidity forecasting

0 comments

The pith

Reduced stochastic hedge ratios of the form sum xi X retain the full sensitivity tensor through path averages.

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

Monte-Carlo engines produce pathwise sensitivities of derivatives to model primitives, which must then be converted to hedge ratios against market instruments via a pathwise linear map. Solving that map separately on each path is costly and high-dimensional. The paper replaces the per-path solution with a low-dimensional expansion whose coefficients are fixed either by minimizing the summed squared residual across all paths or by enforcing projected moment conditions on chosen test functions. The sensitivity tensor itself is never expanded in the basis; it enters the equations only through its empirical averages. The two coefficient choices agree in special cases but diverge once hedge-instrument sensitivities become path-dependent.

Core claim

Reduced stochastic hedge ratios can be written as a linear combination of r basis functions with r much smaller than the number of Monte-Carlo paths; the coefficients are determined either by full empirical residual minimization or by projected moment equations, while the hedge-instrument sensitivity tensor is retained exactly through its path averages rather than replaced by any further expansion.

What carries the argument

The reduced expansion φ_j^r = sum_{q=1}^r ξ_j^q X_q together with retention of the full sensitivity tensor via empirical averages, solved either by least-squares residual or by Galerkin/Petrov-Galerkin moment conditions.

If this is right

Computational cost of converting primitive sensitivities to market hedge ratios drops because only an r-dimensional system is solved instead of one per path.
Numerical stability improves by lowering the dimension of each linear solve and by allowing explicit regularization on the reduced coefficients.
The same reduced ratios can be used for sensitivity-based margin valuation adjustment and for replication-consistent liquidity forecasting.
Tensor and matrix implementations of both the residual-minimization and the moment-projection criteria become feasible, including Petrov-Galerkin variants that use test functions other than the basis itself.

Where Pith is reading between the lines

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

Choosing different test functions in the moment equations may allow tailoring the hedge ratios to secondary objectives such as variance reduction or liquidity-cost minimization that the paper does not explore.
The retained-tensor approach could be combined with adaptive choice of the basis functions X_q drawn from the empirical distribution of the sensitivities themselves.
Out-of-sample hedging performance under model recalibration could be tested by holding the reduced coefficients fixed while resimulating paths under perturbed parameters.

Load-bearing premise

Empirical averages computed from the simulated paths preserve the essential information in the hedge-instrument sensitivity tensor without unacceptable bias or instability once the basis size is chosen far smaller than the path count.

What would settle it

A direct numerical comparison in which the hedge ratios or the resulting valuation-adjustment numbers obtained from the reduced basis differ materially from those obtained by solving the full per-path linear system on the same set of paths.

read the original abstract

Monte-Carlo valuation engines can generate pathwise sensitivities of a derivative value with respect to a high-dimensional vector of model primitives. Hedge ratios with respect to market instruments are then linked to these primitive sensitivities by a pathwise linear relation. Solving this relation independently on every simulated path may be expensive, unstable, and unnecessarily high-dimensional. This paper studies reduced stochastic hedge ratios of the form $\phi_j^r=\sum_{q=1}^r\xi_j^qX_q$, where the number of solution basis functions is much smaller than the number of Monte-Carlo paths. The hedge-instrument sensitivity tensor is not replaced by its own basis expansion; it is retained through empirical averages over the simulated paths. The basis ansatz alone does not determine the coefficients, so two coefficient criteria are distinguished. The first minimizes the full empirical pathwise residual $\sum_\ell\|A_\ell\phi_\ell^r-b_\ell\|_2^2$. The second is a projected moment equation requiring $\langle A\phi^r-b,Y_s\rangle_N=0$ for selected test functions. The special case $Y_s=X_s$ is the usual Galerkin choice; different test functions give a Petrov--Galerkin formulation. The criteria coincide in special cases but differ when the hedge-instrument sensitivities are path-dependent. The paper gives the tensor and matrix forms of both reductions, discusses regularization and conditioning, and records implementation considerations. The constructions are motivated by sensitivity-based margin valuation adjustment and replication-consistent liquidity forecasting, where pathwise primitive sensitivities have to be converted into hedge ratios with respect to market instruments.

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 gives a clean low-dimensional projection for pathwise hedge ratios that keeps the full sensitivity tensor via averages and separates residual-minimization from projected-moment criteria, which only match in special cases.

read the letter

The main thing here is a targeted reduction for turning high-dimensional pathwise sensitivities into stochastic hedge ratios. Instead of solving the linear system on every Monte Carlo path, you project the hedge ratios onto a small basis of size r much less than the path count, while the sensitivity tensor itself stays intact through empirical averages. Two ways to fix the coefficients are laid out: one minimizes the summed pathwise residual, the other solves projected moment equations, and they differ when the hedge-instrument sensitivities depend on the path.

The paper supplies the explicit tensor and matrix forms for both approaches, notes the Galerkin special case, discusses regularization and conditioning, and flags implementation points. The motivation from margin valuation adjustment and replication-consistent forecasting is straightforward. The constructions rest on standard linear algebra applied to the pathwise relation, with no circularity or hidden fitting.

What is actually new is the side-by-side treatment of the two criteria and the observation that they diverge under path dependence. That distinction is useful for anyone implementing this in practice.

The soft spot is the lack of numerical evidence in the abstract. Without reported speedups, error behavior, or conditioning results for realistic path-dependent cases, it is hard to judge how much the reduction buys or how sensitive the choice of basis and test functions really is. If the full paper has those checks, the concern shrinks; otherwise the work stays mostly formal.

This is for quants working on Monte Carlo risk engines and sensitivity-based adjustments. The math is grounded and the distinction is honest, so the paper deserves a serious referee even if it will likely need added numerics to strengthen the case.

Referee Report

2 major / 1 minor

Summary. The manuscript presents reduced stochastic hedge ratios of the form φ_j^r = sum_{q=1}^r ξ_j^q X_q with r much smaller than the Monte-Carlo path count. The hedge-instrument sensitivity tensor is retained via empirical averages rather than expanded; coefficients are obtained either by minimizing the full empirical residual sum_ℓ ||A_ℓ φ_ℓ^r - b_ℓ||_2^2 or by solving the projected moment equations <A φ^r - b, Y_s>_N = 0. The two criteria coincide for the Galerkin choice Y_s = X_s but differ otherwise under path-dependent sensitivities. Tensor and matrix forms, regularization, conditioning, and implementation notes are supplied. The constructions are motivated by sensitivity-based MVA and replication-consistent liquidity forecasting.

Significance. If the empirical averages remain stable, the framework offers a principled way to lower the dimension of the per-path or global solve when converting primitive sensitivities to market-instrument hedge ratios. The explicit separation of residual-minimization from projected-moment criteria, together with the observation that they diverge under path dependence, supplies a useful clarification that builds directly on standard linear-algebra and projection ideas. Retention of the full sensitivity tensor through averages, rather than its own basis expansion, is a clear technical strength.

major comments (2)

[Abstract] Abstract: the central claim that the hedge-instrument sensitivity tensor is retained through empirical averages without introducing unacceptable bias or instability when the basis dimension r is chosen much smaller than the path count is load-bearing for the method's reliability, yet no analysis of the statistical properties, variance, or conditioning of the resulting reduced systems is supplied beyond the general regularization discussion.
[Abstract] Abstract: while the manuscript distinguishes the residual-minimization and projected-moment criteria and notes that they differ when sensitivities are path-dependent, it supplies no concrete example (analytic or numerical) illustrating the divergence, leaving the practical significance of the distinction unverified.

minor comments (1)

The abstract is information-dense; a short additional sentence clarifying the special cases in which the two criteria coincide would improve immediate readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and positive assessment of the technical contributions. We address each major comment below and outline revisions that will strengthen the manuscript.

read point-by-point responses

Referee: [Abstract] Abstract: the central claim that the hedge-instrument sensitivity tensor is retained through empirical averages without introducing unacceptable bias or instability when the basis dimension r is chosen much smaller than the path count is load-bearing for the method's reliability, yet no analysis of the statistical properties, variance, or conditioning of the resulting reduced systems is supplied beyond the general regularization discussion.

Authors: We agree that the manuscript would benefit from a more explicit treatment of the statistical properties of the reduced estimators. The current text emphasizes the algebraic constructions and regularization for conditioning but does not supply bias/variance analysis or Monte-Carlo stability results for r ≪ N. In the revision we will add a dedicated subsection (or short appendix) containing (i) a brief asymptotic argument on consistency of the empirical averages as N → ∞ for fixed r and (ii) numerical illustrations of estimator variance and conditioning across a range of r/N ratios. revision: yes
Referee: [Abstract] Abstract: while the manuscript distinguishes the residual-minimization and projected-moment criteria and notes that they differ when sensitivities are path-dependent, it supplies no concrete example (analytic or numerical) illustrating the divergence, leaving the practical significance of the distinction unverified.

Authors: The referee is correct that the distinction is stated theoretically but not illustrated. We will insert a short analytic example (a low-dimensional linear case with explicitly path-dependent A_ℓ) together with a small numerical table showing the coefficient vectors obtained under each criterion. This will make the practical relevance of the divergence concrete without lengthening the paper substantially. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper derives reduced hedge ratios via standard linear-algebra projections (residual minimization and Petrov-Galerkin moment equations) applied to an explicit pathwise linear relation between primitive sensitivities and hedge ratios. The basis expansion is introduced only for the solution coefficients; the sensitivity tensor is retained through empirical averages, and the two criteria are shown to coincide or diverge according to whether test functions match the basis. These steps are self-contained applications of least-squares and projection methods with explicit tensor/matrix forms; no equation reduces a claimed result to a fitted parameter by construction, no uniqueness theorem is imported from self-citation, and no ansatz is smuggled via prior work. The derivation therefore remains independent of its inputs.

Axiom & Free-Parameter Ledger

3 free parameters · 1 axioms · 0 invented entities

Review performed on abstract only; ledger entries are therefore limited to elements explicitly invoked in the abstract.

free parameters (3)

basis dimension r
Chosen much smaller than number of paths; specific selection rule not stated in abstract.
basis functions X_q
Not specified how the X_q are chosen or constructed.
test functions Y_s
Selected for the projected-moment criterion; choice left open in abstract.

axioms (1)

domain assumption Existence of a pathwise linear relation linking primitive sensitivities to hedge-instrument sensitivities
Invoked in the opening paragraph to motivate the reduction problem.

pith-pipeline@v0.9.1-grok · 5820 in / 1383 out tokens · 26820 ms · 2026-06-30T20:58:01.987383+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

29 extracted references · 13 canonical work pages · 2 internal anchors

[1]

Regression sensitivities for initial margin calculations

ALBANESE, CLAUDIO; CAENAZZO, SIMONE; FRANKEL, OLIVER. Regression sensitivities for initial margin calculations. SSRN Electronic Journal, 2016. Avail- able at SSRN: https://ssrn.com/abstract=2763488; DOI: https: //doi.org/10.2139/ssrn.2763488

work page doi:10.2139/ssrn.2763488 2016
[2]

Algorithmic differentiation for callable exotics

ANTONOV, ALEXANDRE. Algorithmic differentiation for callable exotics. SSRN Electronic Journal, 2017. Available at SSRN: https://ssrn.com/ abstract=2839362

2017
[3]

Conquering the Greeks in Monte Carlo: efficient calculation of the market sensitivities and hedge-ratios of finan- cial assets by direct numerical simulation

AVELLANEDA, MARCO; GAMBA, ROBERTA. Conquering the Greeks in Monte Carlo: efficient calculation of the market sensitivities and hedge-ratios of finan- cial assets by direct numerical simulation. InMathematical Finance – Bachelier Congress 2000, Springer Finance, pp. 93–109. Springer, 2002. DOI: https: //doi.org/10.1007/978-3-662-12429-1_6

work page doi:10.1007/978-3-662-12429-1_6 2000
[4]

Sensitivities for Bermudan options by regression methods

BELOMESTNY, DENIS; MILSTEIN, GRIGORIN; SCHOENMAKERS, JOHNG M. Sensitivities for Bermudan options by regression methods. WIAS preprint, 2007

2007
[5]

Estimating security price derivatives using simulation.Management Science, 42(2):269–285, 1996

BROADIE, MARK; GLASSERMAN, PAUL. Estimating security price derivatives using simulation.Management Science, 42(2):269–285, 1996

1996
[6]

Algorithmic differentiation: adjoint Greeks made easy

CAPRIOTTI, LUCA; GILES, MICHAELB. Algorithmic differentiation: adjoint Greeks made easy. SSRN Electronic Journal, 2011. Available at SSRN: https: //ssrn.com/abstract=1801522

2011
[7]

AAD and least squares Monte Carlo: fast Bermudan-style options and XV A Greeks

CAPRIOTTI, LUCA; JIANG, YUPENG; MACRINA, ANDREA. AAD and least squares Monte Carlo: fast Bermudan-style options and XV A Greeks. SSRN Elec- tronic Journal, 2016. Available at SSRN: https://ssrn.com/abstract= 2842631; DOI:https://doi.org/10.2139/ssrn.2842631

work page doi:10.2139/ssrn.2842631 2016
[8]

CV A sensitivities, hedging and risk

CRÉPEY, STÉPHANE; LI, BOTAO; NGUYEN, HOANG; SAADEDDINE, BOUAZZA. CV A sensitivities, hedging and risk. arXiv:2407.18583, 2024. Available at arXiv: https://arxiv.org/abs/2407.18583

work page arXiv 2024
[9]

finmath-lib: Mathematical Finance Library

FINMATH.NET. finmath-lib: Mathematical Finance Library. Java library and source code repository, 2026. Available at https://www.finmath. net/finmath-lib/; source code: https://github.com/finmath/ finmath-lib. Accessed May 13, 2026

2026
[10]

Automatic Backward Differentiation for American Monte-Carlo Algorithms (Conditional Expectation)

FRIES, CHRISTIANP. Automatic backward differentiation for American Monte- Carlo algorithms, conditional expectation.Risk, April 2018. Working paper ver- sion: SSRN Electronic Journal, 2017, available at SSRN:https://ssrn.com/ abstract=3000822; arXiv:https://arxiv.org/abs/1707.04942. ©2026 Christian Fries20V ersion 0.4.2 (20260512) Faster Forward Sensitivi...

work page internal anchor Pith review Pith/arXiv arXiv 2018
[11]

Back to the future: comparing forward and backward differentiation for forward sensitivities in Monte-Carlo simulations

FRIES, CHRISTIANP. Back to the future: comparing forward and backward differentiation for forward sensitivities in Monte-Carlo simulations. SSRN Elec- tronic Journal, 2018. Available at SSRN: https://ssrn.com/abstract= 3106068

2018
[12]

FRIES, CHRISTIANP. Fast stochastic forward sensitivities in Monte Carlo sim- ulations using stochastic automatic differentiation (with applications to initial margin valuation adjustments).Journal of Computational Finance, 22(4):103– 125, 2019. DOI: https://doi.org/10.21314/JCF.2018.359. Work- ing paper version: SSRN Electronic Journal, 2017, available at...

work page doi:10.21314/jcf.2018.359 2019
[13]

John Wiley & Sons, 2007

FRIES, CHRISTIANP.Mathematical Finance: Theory, Modeling, Implementation. John Wiley & Sons, 2007

2007
[14]

Replication-Consistent Liquidity Forecasting for Deriva- tives – Forward Funding Sensitivities and a Liquidity Valuation Adjustment for Settlement Lags

FRIES, CHRISTIANP. Replication-Consistent Liquidity Forecasting for Deriva- tives – Forward Funding Sensitivities and a Liquidity Valuation Adjustment for Settlement Lags. SSRN Electronic Journal, 2025. Available at SSRN: https://ssrn.com/abstract=5514678; DOI: https://doi.org/ 10.2139/ssrn.5514678

work page doi:10.2139/ssrn.5514678 2025
[15]

Stochastic algorithmic differentiation of (expectations of) discontinuous functions (indicator functions)

FRIES, CHRISTIANP. Stochastic algorithmic differentiation of (expectations of) discontinuous functions (indicator functions). SSRN Electronic Journal,
[16]

Available at SSRN: https://ssrn.com/abstract=3282667; DOI: https://doi.org/10.2139/ssrn.3282667; arXiv: https:// arxiv.org/abs/1811.05741

work page doi:10.2139/ssrn.3282667
[17]

Stochastic automatic differentiation: automatic differen- tiation for Monte-Carlo simulations.Quantitative Finance, 19(6):1043–1059,

FRIES, CHRISTIANP. Stochastic automatic differentiation: automatic differen- tiation for Monte-Carlo simulations.Quantitative Finance, 19(6):1043–1059,
[18]

Working paper version: SSRN Electronic Journal, 2017, available at SSRN: https://ssrn.com/abstract=2995695

DOI: https://doi.org/10.1080/14697688.2018.1556398. Working paper version: SSRN Electronic Journal, 2017, available at SSRN: https://ssrn.com/abstract=2995695

work page doi:10.1080/14697688.2018.1556398 2018
[19]

Melting sensitivities: exact and approximate margin valuation adjustments

FRIES, CHRISTIANP; KOHL-LANDGRAF, PETER; VIEHMANN, MARIO. Melting sensitivities: exact and approximate margin valuation adjustments. SSRN Elec- tronic Journal, 2018. Available at SSRN: https://ssrn.com/abstract= 3095619; DOI:https://doi.org/10.2139/ssrn.3095619

work page doi:10.2139/ssrn.3095619 2018
[20]

Smoking adjoints: fast Monte Carlo Greeks.Risk, 19(1):88–92, 2006

GILES, MICHAELB; GLASSERMAN, PAUL. Smoking adjoints: fast Monte Carlo Greeks.Risk, 19(1):88–92, 2006

2006
[21]

Springer, 2004

GLASSERMAN, PAUL.Monte Carlo Methods in Financial Engineering. Springer, 2004. ©2026 Christian Fries21V ersion 0.4.2 (20260512) Faster Forward Sensitivities Fries, Christian P

2004
[22]

SIAM, 2nd edition, 2008

GRIEWANK, ANDREAS; WALTHER, ANDREA.Evaluating Derivatives: Princi- ples and Techniques of Algorithmic Differentiation. SIAM, 2nd edition, 2008

2008
[23]

Adjoints and Automatic (Algorithmic) Differentiation in Computational Finance

HOMESCU, CRISTIAN. Adjoints and automatic algorithmic differentiation in computational finance. arXiv:1107.1831, 2011. Available at arXiv: https: //arxiv.org/abs/1107.1831

work page internal anchor Pith review Pith/arXiv arXiv 2011
[24]

Rolling adjoints: fast Greeks along Monte Carlo scenarios for early-exercise options

JAIN, SHASHI; LEITAO, ALVARO; OOSTERLEE, CORNELISW. Rolling adjoints: fast Greeks along Monte Carlo scenarios for early-exercise options. SSRN Elec- tronic Journal, 2017. Available at SSRN: https://ssrn.com/abstract= 3093846

2017
[25]

Model calibration, risk measurement, and the hedging of derivatives

LI, ANLONG. Model calibration, risk measurement, and the hedging of derivatives. SSRN Electronic Journal, 1999. Available at SSRN: https: //ssrn.com/abstract=899081; DOI: https://doi.org/10.2139/ ssrn.899081

1999
[26]

Valuing American options by simulation: a simple least-squares approach.Review of Financial Studies, 14(1):113–147, 2001

LONGSTAFF, FRANCISA; SCHWARTZ, EDUARDOS. Valuing American options by simulation: a simple least-squares approach.Review of Financial Studies, 14(1):113–147, 2001

2001
[27]

Computing deltas of callable LIBOR exotics in forward LIBOR models.Journal of Computational Finance, 7:107–144, 2004

PITERBARG, VLADIMIR. Computing deltas of callable LIBOR exotics in forward LIBOR models.Journal of Computational Finance, 7:107–144, 2004

2004
[28]

Hedged Monte-Carlo: low variance derivative pricing with objective probabilities.Physica A: Statistical Mechanics and its Applications, 289(3–4):517–525, 2001

POTTERS, MARC; BOUCHAUD, JEAN-PHILIPPE; SESTOVIC, DRAGAN. Hedged Monte-Carlo: low variance derivative pricing with objective probabilities.Physica A: Statistical Mechanics and its Applications, 289(3–4):517–525, 2001. DOI: https://doi.org/10.1016/S0378-4371(00)00554-9

work page doi:10.1016/s0378-4371(00)00554-9 2001
[29]

Pricing and hedging American-style options: a simple simulation-based approach.Journal of Computational Fi- nance, 13(4):95–125, 2010

WANG, YANG; CAFLISCH, RUSSELE. Pricing and hedging American-style options: a simple simulation-based approach.Journal of Computational Fi- nance, 13(4):95–125, 2010. DOI: https://doi.org/10.21314/JCF. 2010.220. ©2026 Christian Fries22V ersion 0.4.2 (20260512) Faster Forward Sensitivities Fries, Christian P . Notes Classification: JEL-class: C15, C63, G13,...

work page doi:10.21314/jcf 2010

[1] [1]

Regression sensitivities for initial margin calculations

ALBANESE, CLAUDIO; CAENAZZO, SIMONE; FRANKEL, OLIVER. Regression sensitivities for initial margin calculations. SSRN Electronic Journal, 2016. Avail- able at SSRN: https://ssrn.com/abstract=2763488; DOI: https: //doi.org/10.2139/ssrn.2763488

work page doi:10.2139/ssrn.2763488 2016

[2] [2]

Algorithmic differentiation for callable exotics

ANTONOV, ALEXANDRE. Algorithmic differentiation for callable exotics. SSRN Electronic Journal, 2017. Available at SSRN: https://ssrn.com/ abstract=2839362

2017

[3] [3]

Conquering the Greeks in Monte Carlo: efficient calculation of the market sensitivities and hedge-ratios of finan- cial assets by direct numerical simulation

AVELLANEDA, MARCO; GAMBA, ROBERTA. Conquering the Greeks in Monte Carlo: efficient calculation of the market sensitivities and hedge-ratios of finan- cial assets by direct numerical simulation. InMathematical Finance – Bachelier Congress 2000, Springer Finance, pp. 93–109. Springer, 2002. DOI: https: //doi.org/10.1007/978-3-662-12429-1_6

work page doi:10.1007/978-3-662-12429-1_6 2000

[4] [4]

Sensitivities for Bermudan options by regression methods

BELOMESTNY, DENIS; MILSTEIN, GRIGORIN; SCHOENMAKERS, JOHNG M. Sensitivities for Bermudan options by regression methods. WIAS preprint, 2007

2007

[5] [5]

Estimating security price derivatives using simulation.Management Science, 42(2):269–285, 1996

BROADIE, MARK; GLASSERMAN, PAUL. Estimating security price derivatives using simulation.Management Science, 42(2):269–285, 1996

1996

[6] [6]

Algorithmic differentiation: adjoint Greeks made easy

CAPRIOTTI, LUCA; GILES, MICHAELB. Algorithmic differentiation: adjoint Greeks made easy. SSRN Electronic Journal, 2011. Available at SSRN: https: //ssrn.com/abstract=1801522

2011

[7] [7]

AAD and least squares Monte Carlo: fast Bermudan-style options and XV A Greeks

CAPRIOTTI, LUCA; JIANG, YUPENG; MACRINA, ANDREA. AAD and least squares Monte Carlo: fast Bermudan-style options and XV A Greeks. SSRN Elec- tronic Journal, 2016. Available at SSRN: https://ssrn.com/abstract= 2842631; DOI:https://doi.org/10.2139/ssrn.2842631

work page doi:10.2139/ssrn.2842631 2016

[8] [8]

CV A sensitivities, hedging and risk

CRÉPEY, STÉPHANE; LI, BOTAO; NGUYEN, HOANG; SAADEDDINE, BOUAZZA. CV A sensitivities, hedging and risk. arXiv:2407.18583, 2024. Available at arXiv: https://arxiv.org/abs/2407.18583

work page arXiv 2024

[9] [9]

finmath-lib: Mathematical Finance Library

FINMATH.NET. finmath-lib: Mathematical Finance Library. Java library and source code repository, 2026. Available at https://www.finmath. net/finmath-lib/; source code: https://github.com/finmath/ finmath-lib. Accessed May 13, 2026

2026

[10] [10]

Automatic Backward Differentiation for American Monte-Carlo Algorithms (Conditional Expectation)

FRIES, CHRISTIANP. Automatic backward differentiation for American Monte- Carlo algorithms, conditional expectation.Risk, April 2018. Working paper ver- sion: SSRN Electronic Journal, 2017, available at SSRN:https://ssrn.com/ abstract=3000822; arXiv:https://arxiv.org/abs/1707.04942. ©2026 Christian Fries20V ersion 0.4.2 (20260512) Faster Forward Sensitivi...

work page internal anchor Pith review Pith/arXiv arXiv 2018

[11] [11]

Back to the future: comparing forward and backward differentiation for forward sensitivities in Monte-Carlo simulations

FRIES, CHRISTIANP. Back to the future: comparing forward and backward differentiation for forward sensitivities in Monte-Carlo simulations. SSRN Elec- tronic Journal, 2018. Available at SSRN: https://ssrn.com/abstract= 3106068

2018

[12] [12]

FRIES, CHRISTIANP. Fast stochastic forward sensitivities in Monte Carlo sim- ulations using stochastic automatic differentiation (with applications to initial margin valuation adjustments).Journal of Computational Finance, 22(4):103– 125, 2019. DOI: https://doi.org/10.21314/JCF.2018.359. Work- ing paper version: SSRN Electronic Journal, 2017, available at...

work page doi:10.21314/jcf.2018.359 2019

[13] [13]

John Wiley & Sons, 2007

FRIES, CHRISTIANP.Mathematical Finance: Theory, Modeling, Implementation. John Wiley & Sons, 2007

2007

[14] [14]

Replication-Consistent Liquidity Forecasting for Deriva- tives – Forward Funding Sensitivities and a Liquidity Valuation Adjustment for Settlement Lags

FRIES, CHRISTIANP. Replication-Consistent Liquidity Forecasting for Deriva- tives – Forward Funding Sensitivities and a Liquidity Valuation Adjustment for Settlement Lags. SSRN Electronic Journal, 2025. Available at SSRN: https://ssrn.com/abstract=5514678; DOI: https://doi.org/ 10.2139/ssrn.5514678

work page doi:10.2139/ssrn.5514678 2025

[15] [15]

Stochastic algorithmic differentiation of (expectations of) discontinuous functions (indicator functions)

FRIES, CHRISTIANP. Stochastic algorithmic differentiation of (expectations of) discontinuous functions (indicator functions). SSRN Electronic Journal,

[16] [16]

Available at SSRN: https://ssrn.com/abstract=3282667; DOI: https://doi.org/10.2139/ssrn.3282667; arXiv: https:// arxiv.org/abs/1811.05741

work page doi:10.2139/ssrn.3282667

[17] [17]

Stochastic automatic differentiation: automatic differen- tiation for Monte-Carlo simulations.Quantitative Finance, 19(6):1043–1059,

FRIES, CHRISTIANP. Stochastic automatic differentiation: automatic differen- tiation for Monte-Carlo simulations.Quantitative Finance, 19(6):1043–1059,

[18] [18]

Working paper version: SSRN Electronic Journal, 2017, available at SSRN: https://ssrn.com/abstract=2995695

DOI: https://doi.org/10.1080/14697688.2018.1556398. Working paper version: SSRN Electronic Journal, 2017, available at SSRN: https://ssrn.com/abstract=2995695

work page doi:10.1080/14697688.2018.1556398 2018

[19] [19]

Melting sensitivities: exact and approximate margin valuation adjustments

FRIES, CHRISTIANP; KOHL-LANDGRAF, PETER; VIEHMANN, MARIO. Melting sensitivities: exact and approximate margin valuation adjustments. SSRN Elec- tronic Journal, 2018. Available at SSRN: https://ssrn.com/abstract= 3095619; DOI:https://doi.org/10.2139/ssrn.3095619

work page doi:10.2139/ssrn.3095619 2018

[20] [20]

Smoking adjoints: fast Monte Carlo Greeks.Risk, 19(1):88–92, 2006

GILES, MICHAELB; GLASSERMAN, PAUL. Smoking adjoints: fast Monte Carlo Greeks.Risk, 19(1):88–92, 2006

2006

[21] [21]

Springer, 2004

GLASSERMAN, PAUL.Monte Carlo Methods in Financial Engineering. Springer, 2004. ©2026 Christian Fries21V ersion 0.4.2 (20260512) Faster Forward Sensitivities Fries, Christian P

2004

[22] [22]

SIAM, 2nd edition, 2008

GRIEWANK, ANDREAS; WALTHER, ANDREA.Evaluating Derivatives: Princi- ples and Techniques of Algorithmic Differentiation. SIAM, 2nd edition, 2008

2008

[23] [23]

Adjoints and Automatic (Algorithmic) Differentiation in Computational Finance

HOMESCU, CRISTIAN. Adjoints and automatic algorithmic differentiation in computational finance. arXiv:1107.1831, 2011. Available at arXiv: https: //arxiv.org/abs/1107.1831

work page internal anchor Pith review Pith/arXiv arXiv 2011

[24] [24]

Rolling adjoints: fast Greeks along Monte Carlo scenarios for early-exercise options

JAIN, SHASHI; LEITAO, ALVARO; OOSTERLEE, CORNELISW. Rolling adjoints: fast Greeks along Monte Carlo scenarios for early-exercise options. SSRN Elec- tronic Journal, 2017. Available at SSRN: https://ssrn.com/abstract= 3093846

2017

[25] [25]

Model calibration, risk measurement, and the hedging of derivatives

LI, ANLONG. Model calibration, risk measurement, and the hedging of derivatives. SSRN Electronic Journal, 1999. Available at SSRN: https: //ssrn.com/abstract=899081; DOI: https://doi.org/10.2139/ ssrn.899081

1999

[26] [26]

Valuing American options by simulation: a simple least-squares approach.Review of Financial Studies, 14(1):113–147, 2001

LONGSTAFF, FRANCISA; SCHWARTZ, EDUARDOS. Valuing American options by simulation: a simple least-squares approach.Review of Financial Studies, 14(1):113–147, 2001

2001

[27] [27]

Computing deltas of callable LIBOR exotics in forward LIBOR models.Journal of Computational Finance, 7:107–144, 2004

PITERBARG, VLADIMIR. Computing deltas of callable LIBOR exotics in forward LIBOR models.Journal of Computational Finance, 7:107–144, 2004

2004

[28] [28]

Hedged Monte-Carlo: low variance derivative pricing with objective probabilities.Physica A: Statistical Mechanics and its Applications, 289(3–4):517–525, 2001

POTTERS, MARC; BOUCHAUD, JEAN-PHILIPPE; SESTOVIC, DRAGAN. Hedged Monte-Carlo: low variance derivative pricing with objective probabilities.Physica A: Statistical Mechanics and its Applications, 289(3–4):517–525, 2001. DOI: https://doi.org/10.1016/S0378-4371(00)00554-9

work page doi:10.1016/s0378-4371(00)00554-9 2001

[29] [29]

Pricing and hedging American-style options: a simple simulation-based approach.Journal of Computational Fi- nance, 13(4):95–125, 2010

WANG, YANG; CAFLISCH, RUSSELE. Pricing and hedging American-style options: a simple simulation-based approach.Journal of Computational Fi- nance, 13(4):95–125, 2010. DOI: https://doi.org/10.21314/JCF. 2010.220. ©2026 Christian Fries22V ersion 0.4.2 (20260512) Faster Forward Sensitivities Fries, Christian P . Notes Classification: JEL-class: C15, C63, G13,...

work page doi:10.21314/jcf 2010