[Paper Review] Integration by Parts Formula, Derivative Formula, and Transportation Inequalities for SDEs Driven by Fractional Brownian Motion

Xiliang Fan|arXiv (Cornell University)|Feb 24, 2013

Stochastic processes and financial applications6 citations

TL;DR

This paper establishes integration by parts formulas, derivative formulas, and transportation inequalities for solutions to stochastic Volterra equations driven by fractional Brownian motion. Using Malliavin calculus and pathwise analysis, it derives Bismut-type derivative formulas and Talagrand-type transportation cost inequalities, enabling gradient estimates, strong Feller properties, and regularity results for the law of the solution under both uniform and L² metrics.

ABSTRACT

This paper is devoted to study a class of stochastic Volterra equations associated with fractional Brownian motion. We first prove the Driver type integration by parts formula and the shift Harnack type inequalities. As a direct application, we provide an alternative method to describe the regularities of the law of the solution. Secondly, by using the Malliavin calculus, the Bismut type derivative formula is established, which is then applied to the study of the gradient estimate and the strong Feller property. Finally, we establish the Talagrand type transportation cost inequalities for the law of the solution on the path space with respect to both the uniform metric and the $L^2$-metric.

Motivation & Objective

• To develop integration by parts formulas for solutions of stochastic Volterra equations with fractional Brownian motion.
• To establish Bismut-type derivative formulas via Malliavin calculus for gradient estimates and the strong Feller property.
• To derive Talagrand-type transportation cost inequalities on path space with respect to uniform and L² metrics.
• To provide an alternative method for analyzing the regularity of the law of the solution using these functional inequalities.

Proposed method

• Application of Driver-type integration by parts formula to characterize the law of the solution.
• Use of Malliavin calculus to derive a Bismut-type derivative formula for the transition semigroup.
• Establishment of shift Harnack inequalities as a tool for gradient estimates.
• Derivation of Talagrand-type transportation cost inequalities on path space under both uniform and L² metrics.
• Combination of pathwise techniques and Malliavin calculus to analyze regularity and ergodicity properties.
• Use of the solution's Malliavin differentiability to obtain gradient estimates and strong Feller behavior.

Experimental results

Research questions

• RQ1 Can an integration by parts formula be established for solutions of SDEs driven by fractional Brownian motion?
• RQ2 How can Malliavin calculus be used to derive a Bismut-type derivative formula for such processes?
• RQ3 What transportation cost inequalities hold for the law of the solution on path space under different metrics?
• RQ4 Can these functional inequalities be used to deduce gradient estimates and the strong Feller property?
• RQ5 What is the regularity of the law of the solution, and how can it be characterized via these tools?

Key findings

• An integration by parts formula of Driver type is established for the solution of the stochastic Volterra equation driven by fractional Brownian motion.
• A Bismut-type derivative formula is derived using Malliavin calculus, enabling gradient estimates and the strong Feller property.
• Shift Harnack inequalities are obtained, providing a new method to study the regularity of the law of the solution.
• Talagrand-type transportation cost inequalities are proven for the law of the solution on path space with respect to both the uniform metric and the L²-metric.
• The derived functional inequalities lead to quantitative estimates on the transition density and the smoothness of the law.
• The results demonstrate the applicability of Malliavin calculus and pathwise methods in analyzing SDEs with long-range dependence.