Itô's stochastic calculus: its surprising power for applications (Q972809)

This paper gives a comprehensive survey of Kiyosi Itô's work together with important generalizations and applications. The first part is about the early work and includes the Lévy-Itô decomposition of Lévy processes and results for transition probabilities of diffusion processes. Furthermore stochastic integration with respect to the Brownian motion is introduced and Itô's formula for these integrals is stated. In the last part of this first section the method of successive approximations is used to show existence and uniqueness of the solution of stochastic differential equations (SDEs) if its coefficients are Lipschitz continuous and satisfy a linear growth condition. The second section generalizes the theory of stochastic integration to semimartingales. This is achieved by decomposing the semimartingale into a pure-jump part and a continuous part, that is again decomposed into a locally square integrable process and a process of bounded variation. Now the stochastic integral with respect to a semimartingale can be defined and Itô's formula for these processes is given. Compensated Poisson random measures and Itô jump processes are defined and studied, including Itô's formula, a martingale representation theorem and a Girsanov theorem. In the last section the theory introduced so far is applied to mathematical finance, assuming that the market consists of a risk-free bond and a risky asset that can be modeled as a one-dimensional SDE. Important concepts introduced are the Black-Scholes partial differential equation, equivalent martingale measures and Merton's integro-differential equation for Itô jump processes.