《《ch4Ito》.pdfVIP

Stochastic Analysis in Quantitative Finance S. G. Kou National University of Singapore and Columbia University This version Spring 2014 Chapter 4 ItÙ Formula for Smooth and Non-Smooth Functions 4.1 Di§erences between Calculus and Stochastic Calculus In calculus the Riemann integral is deÖned as the limit of sum. However to compute the Riemann integral without going through the complicated process of taking limits, one can use the chain rule (also called the fundamental theorem of calculus), which says that f(x ) f(x ) = Z T f (x )dt;0 T 0 t 0 0 0 where f is the Örst derivative of f. The derivative f is deÖned as 0 f(y) f(x) df(x) f (x) = lim = : y!x y x dx However, in stochastic calculus, although the stochastic integral is deÖned via the limit of sum of simple functions, for a stochastic process X the derivative t 0 dXt X and are not deÖned. The term dX or f(t)dX are only interpreted by t dt t t their corresponding integration forms, i.e. Rt dX and Rt f(s)dX : 0 s 0 s The second di§erence is that in caculus we have the chain rule (also called the fundamental theorem in calculus) is a result which links a Riemann integral (deÖned as the limit of sums) to the Örst derivative (deÖned as the limit of the ratio of di§erences): f(x) f(0) = Z0x f (s)ds:0