Let $\left(X_n\right)_{n\geq 1}$ be a sequence of i.i.d. real random variables, with $\mathbb E(X_1)=0$, $\operatorname{var}(X_1)=1$. Let $S_n=X_1+\cdots+X_n$.

Prove that for any $A>0$, $$\displaystyle \left\{\limsup_{n\rightarrow \infty}\frac{S_n}{\sqrt{n}}>A\right\}\in\cap_{n\geq 1}\sigma(X_i,i\geq n)$$

And deduce that $$\displaystyle \mathbb P\left(\limsup_{n\rightarrow \infty}\frac{S_n}{\sqrt{n}}=+\infty\right)=1$$


We observe that for any $m \in \mathbb N$,

$$\limsup_{n\rightarrow\infty} \frac{S_n}{\sqrt n} = \limsup_{n\rightarrow\infty} (\frac{1}{\sqrt n}\sum_{i=m}^n X_i + \frac{1}{\sqrt n}\sum_{i=1}^{m-1} X_i ).$$

The second term in parentheses goes to 0 as $n\rightarrow\infty$. Thus we find

$$\limsup_{n\rightarrow\infty} \frac{S_n}{\sqrt n} = \limsup_{n\rightarrow\infty} \frac{1}{\sqrt n}\sum_{i=m}^n X_i .$$

and so

$$\{\limsup_{n\rightarrow\infty} \frac{S_n}{\sqrt n} > A\} = \{\limsup_{n\rightarrow\infty} \frac{1}{\sqrt n}\sum_{i=m}^n X_i > A\} .$$

The limsup of a sequence of random variables which is measurable with respect to a certain $\sigma$-algebra is also measurable with respect to that $\sigma$-algebra. Thus the event on the right side above is in $\bigcap_{n=m}^\infty \sigma(X_i : i\geq n)$. This proves the first statement.

For the second, we apply the Kolmogorov zero-one law to deduce that $P\{\limsup_{n\rightarrow\infty} \frac{S_n}{\sqrt n} > A\} = 0$ or $1$. So, to prove the second claim it suffices to show that this event has positive probability. By the central limit theorem, $\lim_{n\rightarrow\infty} P\{\frac{S_n}{\sqrt n} > A\} > 0$. So, with positive probability there exist arbitrarily large $n$ with $\frac{S_n}{\sqrt n} > A$. Thus $P\{\limsup_{n\rightarrow\infty} \frac{S_n}{\sqrt n} > A\} > 0$, which proves the result.


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.