Brownian Motion and Stochastic Calculus
Brownian motion is the canonical continuous-time random process, and the Ito calculus built on it provides the rules for differentiating and integrating along its jagged, nowhere-differentiable paths, the language of modern stochastic modeling.
Definition
Brownian motion is a continuous-path process with independent stationary Gaussian increments, and stochastic calculus is the theory of integration and differentiation with respect to it and related continuous martingales, centered on the Ito integral and the Ito formula.
Scope
The area covers the construction and path properties of Brownian motion, its martingale and Markov characterizations, the Ito stochastic integral against Brownian motion and continuous martingales, the Ito formula as the chain rule of stochastic calculus, stochastic differential equations and their existence and uniqueness theory, and connections to partial differential equations through the Feynman-Kac formula.
Sub-topics
Core questions
- How is Brownian motion constructed, and what are its striking path properties?
- How can one integrate against a process whose paths have unbounded variation?
- What replaces the ordinary chain rule when the integrator is Brownian motion?
- How are stochastic differential equations defined and solved?
Key theories
- Ito integral and Ito formula
- The Ito integral defines integration against Brownian motion using its quadratic variation, and the Ito formula is the resulting chain rule, which carries an extra second-order term reflecting that quadratic variation accumulates linearly in time.
- Stochastic differential equations and Feynman-Kac
- Stochastic differential equations driven by Brownian motion have unique strong solutions under Lipschitz and growth conditions, and the Feynman-Kac formula represents solutions of associated parabolic partial differential equations as expectations over these diffusions.
Clinical relevance
Stochastic calculus is the mathematical foundation of continuous-time finance, where the Black-Scholes model prices options through an Ito process, and it pervades physics, where it describes diffusion and noise, engineering, where it underlies filtering and stochastic control, and biology, where it models population and neural dynamics under randomness.
History
Brownian motion was observed by Robert Brown, modeled physically by Einstein and Smoluchowski, and constructed rigorously by Norbert Wiener in 1923. Kiyosi Ito created the stochastic integral and the Ito formula in the 1940s, founding stochastic calculus, which later became indispensable to mathematical finance.
Key figures
- Norbert Wiener
- Kiyosi Ito
- Paul Levy
- Mark Kac
Related topics
Seminal works
- karatzas1991
- revuz1999
Frequently asked questions
- Why can't ordinary calculus be used with Brownian motion?
- Brownian paths are continuous but nowhere differentiable and have infinite variation, so the usual Riemann-Stieltjes integral and chain rule do not apply; the Ito calculus replaces them with constructions based on the finite quadratic variation of the paths.
- What is the extra term in the Ito formula?
- Because the squared increments of Brownian motion accumulate at a definite rate rather than vanishing, the stochastic chain rule includes a second-derivative term proportional to the elapsed time, which has no analogue in ordinary calculus.