dX = dW, X₀ = 0, t = 4
p is N(0, 4); p(0, 4)
0.1995.
If , the density of obeys the Fokker–Planck equation .
This page is easier if you are comfortable with:
The first term transports probability with the drift, as in a continuity equation; the second spreads it out by diffusion. For and the equation is the heat equation (the heat equation), solved by with peak density ; for this is . The solution is the normal density of Brownian motion, see the normal density and the heat kernel.
With constant drift and , is normal with mean and variance : at , mean 1.5 and variance 12. For the Ornstein–Uhlenbeck process , , the stationary density (a standard normal) makes the right-hand side zero: the flux vanishes when .
The equation is derived from Itô's lemma by applying it to a test function and integrating by parts; the partial differential equation methods are those of separation of variables. Applications include price models, population dynamics and statistical physics; the dual formulation for expectations is the Feynman–Kac formula.
The Fokker–Planck equation describes how the probability density of a randomly moving particle spreads and drifts with time.
The density spreads out and its peak falls as time passes.
These examples are worked by hand, step by step, so you can compare your own working.
For , and start at 0, find the peak density at .
For , find the mean and variance at .
For and , verify that makes the flux zero at .
A few situations come up again and again. Knowing them saves time and helps you spot answers that cannot be right.
p is N(0, 4); p(0, 4)
0.1995.
mean
6.
stationary variance σ²/(2θ)
0.5.
stationary density at 0
0.5642.
For dX = μ(x)dt + σ(x)dW, ∂p/∂t = −∂(μp)/∂x + ½∂²(σ²p)/∂x². With μ = 0 and σ = 1 it is the heat equation.
X_t ~ N(0, t) with t = 4: p(0).
Mean of X₃.
Stationary variance σ²/(2θ).
Normal with variance 0.5, at 0.
only diffusion ✗
include −(μp)ₓ ✓
Both terms.
σ ✗
½σ² ✓
Variance.
any ✗
∫p dx = 1 always ✓
Continuity equation.
same ✗
p is the density of X_t ✓
Forward equation.
Try each one on paper, then open the answer. Use the solver above to see the full working.
0.3989
0.133
2.5
20
1
0.1995
Unlimited fresh questions with instant marking. Choose a level, work it out on paper, then type your answer.
Solve: …
∂p/∂t = −(μp)ₓ + ½(σ²p)ₓₓ.
The density of a diffusion over time.
The heat equation, see heat equation.
Set ∂p/∂t = 0 and solve.
Brownian motion.
geometric Brownian motion.
More in Advanced statistics and stochastic processes and the formula sheets.