Stochastic differential equations · Part 3 of 10
Building the Itô integral
Many models of random dynamics have the form “rate of change equals drift plus noise”. Written as an integral equation, such a model reads
and everything hinges on the last term. This post defines it.
Why the obvious definition fails
For a smooth integrator , the integral is a limit of Riemann–Stieltjes sums , and it does not matter where in the point is chosen. Brownian motion has infinite variation, and suddenly the choice matters. Compare left and right endpoints in sums for :
The two kinds of sums converge to limits that differ by . Itô’s choice is the left endpoint. It is the natural one for modeling, because the integrand at each step is fixed before the next increment of noise arrives, the way a trading strategy has to be chosen before prices move.
Simple adapted integrands
Fix a filtration to which is adapted, with independent of for . A simple adapted process is a step function
where each is a bounded random variable that is -measurable, that is, known at the start of its step. For such processes there is only one reasonable definition:
The Itô isometry
The key identity is that this map preserves the norm:
To see why, write and expand the square. For , the increment is independent of , which is known at time , so the cross term is zero. For a diagonal term, is independent of , so . Summing gives exactly . Adaptedness is what makes every step of this argument work.
Extending by approximation
Now let be any adapted process with (and a mild measurability condition). Such processes can be approximated by simple adapted processes , in the sense that . By the isometry, the integrals form a Cauchy sequence in , so they converge, and the limit does not depend on the approximating sequence. That limit is the Itô integral of .
Viewed as a process in its upper limit, inherits three properties from the simple case:
- it has a version with continuous paths;
- it is a martingale with mean zero;
- its variance is .
If only satisfies almost surely, a localization argument (stopping before the integral gets too large) still defines , but then it may only be a local martingale. The difference matters, and it comes back in the post on martingales.
Quadratic variation and covariation
For a continuous process , the quadratic variation is the limit in probability of as the mesh of a partition of goes to zero. The previous post showed that . For an Itô process
the drift contributes nothing, because it has finite variation, and
The covariation of two processes is defined by polarization, . In differential shorthand, all of this is captured by a multiplication table:
and for independent Brownian motions and . These rules are exactly what Itô’s formula needs.
The example that explains everything
Return to . Using left-endpoint sums and the algebraic identity ,
So
not as ordinary calculus would suggest. The correction is exactly what keeps the right side a martingale with mean zero. Evaluating at midpoints instead gives the Stratonovich integral, which obeys the ordinary chain rule but is not a martingale. The next post turns this example into a general rule.