A scalar diffusion $X=(X_t)_{0\leq t \leq T}$ is the unique strong solution of the SDE $$\text{d}X_t=\text{d}W_t$$ where $W=(W_t)_{0\leq t \leq T}$ is a scalar Brownian motion.

Assume that the state space of $X$ is $R$. Let $Y=(Y_t)_{t\ge 0}$ be given by $\text{d}Y_t=\text{d}t$.

Define $p(x,y;t)\text{d}t:=P(Y_{\tau}\in \text{d}t|X_0=x,Y_0=y), \tau:=\text{inf}\{t\ge 0:X_t=m\}$

Find $p(x,y;t)$.

I'm only vaguely familiar with stochastic processes, should I use Kolmogorov forward equation or backward equation to solve it?

  • $\begingroup$ Essentially you need to compute the density of the distribution of $\tau$. Ever heard of the reflection principle (for Brownian motion)? $\endgroup$
    – saz
    Commented Mar 14, 2020 at 7:09
  • $\begingroup$ @saz Yes. I know the density of the first hitting time of Brownian motion. Is it $p(x,y;t)$? $\endgroup$
    – KKKcat
    Commented Mar 14, 2020 at 20:07

1 Answer 1


Indeed as mentioned in the comments, this is simply the density for Brownian motion

$$p(x,y;t)\text{d}t:=P(Y_{\tau}\in \text{d}t|X_0=x,Y_0=y)=P[y+\tau\in dt| X_{0}=x].$$

This has been computed here Density of first hitting time of Brownian motion from its CDF.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .