• $(\Omega,\mathcal A,\operatorname P)$ be a probability space
  • $\left(Y^{(n)}_k\right)_{k\in\mathbb N_0}$ be a time-homogeneous Markov chain on $(\Omega,\mathcal A,\operatorname P)$
  • $D([0,1]):=\left\{f:[0,1]\to\mathbb R\mid f\text{ is càdlàg and left-continuous at }1\right\}$
  • $X^{(n)}$ be a $D([0,1])$-valued random variable on $(\Omega,\mathcal A,\operatorname P)$ with $$X^{(n)}_t=Y^{(n)}_{\lfloor nt\rfloor}\;\;\;\text{for all }t\in[0,1)$$
  • $(N_t)_{t\ge0}$ be a Poisson process on $(\Omega,\mathcal A,\operatorname P)$ with intensity $1$ independent of $Y^{(n)}$ for all $n\in\mathbb N$ and $$Z^{(n)}_t:=\begin{cases}Y^{(n)}_{N_{nt}}&\text{for }t\in[0,1)\\ Z^{(n)}_{1-}\end{cases}$$

I've read that first $(n-1)\wedge N_{n-}$ steps of $X^{(n)}$ and $Z^{(n)}$ coincide, but appear at different times ($\frac kn$ vs the $k$th jump time of $\left(N_{nt}\right)_{t\ge0}$). What's exactly meant and how can we prove it rigorously?

I've got a vague intuition what's meant: Assuming that $N$ is right-continuous (in practice, we can always find a right-continuous modification), $N$ is almost surely nondecreasing and makes almost surely jumps of size $1$. So, $t\mapsto N_{nt}$ somehow behaves like $t\mapsto\lfloor nt\rfloor$, but the time-scale is stretched. How can we formulate this rigorously?

What's confusing me most is that it's written that only the first $(n-1)\wedge N_{n-}$ steps coincide. Where does the $N_{n-}=\lim_{t\to n-}N_t$ come from? Shouldn't the first $n-1$ steps coincide?


It's easy to see that $$X^{(n)}_{[0,\:1)}=\left\{Y^{(n)}_0,\ldots,Y^{(n)}_{n-1}\right\}$$ and $$Z^{(n)}_{[0,\:1)}=\left\{Y^{(n)}_0,\ldots,Y^{(n)}_{N_{n-}}\right\}$$ for all $n\in\mathbb N$. In order to make these processes coincide, we can apply a random time change $$\lambda^{(n)}_t:=\sum_{k=0}^\infty 1_{\left[\frac kn,\:\frac{k+1}n\right)}(t)\left(\tau^{(n)}_k+(nt-k)\left(\tau^{(n)}_{k+1}-\tau^{(n)}_k\right)\right)\;\;\;\text{for }t\ge0$$ with $\tau_0:=0$, $$\tau_k:=\inf\left\{t>\tau_{k-1}:\Delta N_t=1\right\}\;\;\;\text{for }k\in\mathbb N$$ and $$\tau^{(n)}_k:=\frac{\tau_k}n\;\;\;\text{for }k\in\mathbb N$$ for $n\in\mathbb N$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.