Stochastic differential equations · Part 1 of 10
Probability, the measure-theoretic way
Stochastic calculus is about random functions of continuous time, and that is where elementary probability runs out of road. “The probability that a Brownian path stays below 1 for all ” involves uncountably many random variables at once, and making sense of it requires the measure-theoretic framework. This post collects the parts of that framework the rest of the collection relies on.
Probability spaces
A probability space is a triple . The sample space lists all possible outcomes, is a σ-algebra of subsets of called events, and assigns each event a probability, with and countable additivity: for disjoint events ,
Why not assign a probability to every subset? On with the uniform distribution, there are sets to which no consistent length can be assigned, so probabilities are defined only on a σ-algebra: a collection of sets that contains and is closed under complements and countable unions. That restriction is more than a technicality. It also carries meaning, as the next section explains.
σ-algebras as information
A useful way to think about a σ-algebra is as a state of knowledge: contains exactly the events whose occurrence you can decide with the information it represents. For a coin tossed three times, the σ-algebra generated by the first toss contains only four events: the empty event, the whole space, “the first toss is heads” and “the first toss is tails”. A finer σ-algebra means more information.
Information that grows over time is described by a filtration: an increasing family of σ-algebras, in discrete time or for in continuous time. A process is adapted if its value at each time is known at time . Adaptedness is exactly the condition that will make the Itô integral work: integrands may use the past, but never the future.
Random variables and their laws
A random variable is a function that is measurable, meaning for every Borel set , so that all of its probabilities are defined. Its law is the distribution that it induces on the real line, and , the σ-algebra generated by , is the information obtained by observing it.
Two examples show how much a simple probability space can hold.
- Uniform and Gaussian variables. Take with length as probability. Then is uniform, and , where is the standard normal distribution function, is a standard Gaussian. In the same way, any distribution on the real line can be produced from a single uniform variable.
- Infinitely many coin tosses from one number. Write in binary. The digits are independent fair coin flips, so the unit interval already contains an infinite sequence of coin tosses. The partial sums then form a simple random walk, the discrete ancestor of Brownian motion.
Expectation and convergence
The expectation is a Lebesgue integral. Its great advantage over the Riemann integral is how well it behaves under limits: the monotone and dominated convergence theorems give simple conditions under which .
Random variables can converge in several inequivalent senses:
- almost surely, if for every outside an event of probability zero;
- in probability, if for every ;
- in , if ;
- in distribution, if for every bounded continuous function .
Almost sure convergence and convergence each imply convergence in probability, which in turn implies convergence in distribution, and none of the reverse implications holds in general. The random walk illustrates two of them: almost surely, by the strong law of large numbers, while converges only in distribution, to , by the central limit theorem. In stochastic calculus, convergence does most of the work, because the Itô integral is constructed as an limit.
Conditional expectation
Given a σ-algebra , the conditional expectation is the best guess of that uses only the information in . Formally, it is the -measurable random variable that satisfies
For square-integrable , it is the orthogonal projection of onto the subspace of -measurable random variables, so it minimizes the mean squared error among all such . Three rules cover almost every calculation:
- the tower property: when ;
- taking out what is known: when is -measurable;
- independence: when is independent of .
A first glimpse of martingales
Apply these rules to the random walk with . The next step is independent of the past and has mean zero and variance one, so
Processes with this “the best prediction of the future is the present” property are martingales. Both identities have exact analogues for Brownian motion, where and are martingales, and the second one is the first sign of the quadratic variation that makes stochastic calculus different from ordinary calculus. The next post constructs Brownian motion itself.