Title:Stochastic Calculus via Stopping Derivatives

Abstract:This article provides a derivative-based development of core aspects of stochastic calculus. We define a notion of stopping derivative, which is a form of right derivative with respect to stopping times. Using this, we define the drift and variance rate of a process as stopping derivatives for conditional expectation and conditional variance respectively. We prove a version of Itô's formula, expressing how the quantities of drift and variance rate transform under the application of a twice continuously differentiable function. Another result establishes a characterisation of continuous local martingales as continuous processes that have zero drift. We also give a fundamental theorem of calculus for stopping derivatives, which relates the stopping-derivative-defined drift and variance rate to the parameters used in the definition of a process via a stochastic integral.