Stochastic modelling in mathematical and computational finance

点击翻译按钮获取中文摘要

Stochastic modelling in mathematical and computational finance is the focus of the proposed research program.  Mathematical and computational finance considers the uncertain future behaviour of financial or economic variables and systems; presents a theory for the valuation and risk management of derivative securities; and provides a quantitative framework for examining various investment, managerial, and regulatory decisions. Various probabilistic models, statistical estimation methods, and computational algorithms which are motivated by financial applications shall be considered.  Addressing applied problems in a realistic framework also drives new theoretical research and is the impetus for novel theoretical advances in probability and statistics.  The research integrates various aspects of mathematical and computational finance starting with the development of new stochastic models for fundamental financial and economic quantities such as interest rates and asset volatility. We shall develop new pricing and risk management theories methodologies that allow market participants to value and hedge financial derivatives exposed to credit risk.  We also plan to study the stochastic equations which characterize these new valuation and risk management methods, deriving explicit solutions where possible but  focusing on realistic modelling which requires the creation of new efficient computational algorithms for solving these equations.***The first objective of the proposed research program is the study of forward -backward stochastic differential equations (FBSDEs) and applications in mathematical finance. An FBSDE is a coupled system of stochastic equations with components that evolve forward in time from a specified initial condition and components that evolve backward in time from a random terminal condition.  We shall use FBSDEs to characterize a new pricing methodology for credit risk derivatives such as defaultable bonds and extend this method.  The second objective is the development of numerical methods for the solution of FBSDEs since the class of FBSDEs with explicit solutions is limited.  We shall further develop a new numerical method we created, based on the fast Fourier transform, to higher dimensions.  The third objective involves the study of problems in mathematical finance that can be characterized as producing or depending on large amounts of high- dimensional data. Our goal is to extend to financial contexts certain modelling and statistical techniques for high- dimensional data that effectively reduce the dimension to the point that an accurate lower dimensional model can be implemented.  Infinite dimensional models of forward interest rate processes shall be the first example considered so that, by reducing the dimension,  we can create new parsimonious financial models that preserve the key features of the theoretical model and the observed data.**