Let $H\in$ $]0,1[$. A fractional Brownian motion $\left(B_H(t)\right)_{t\geq 0}$ can be represented as $$X(t)={1\over C(H)}\int_\mathbb{R}\left((t-s)_+^{H-{1\over2}}-(-s)_+^{H-{1\over2}}\right)dB(s)$$ where $B(t)$ is the standard Brownian motion and $$C(H)=\left(\int_0^{\infty}\left((1+s)^{H-{1\over2}}-(-s)^{H-{1\over2}}\right)^2ds+{1\over 2H}\right)^{1\over2}$$

I would like to show that $\mathbb{E}[X(t)]=0$. My professor said that the zero mean can be explained by the symmetry of $X(t)$. What did he mean? How can he tell that $X(t)$ is symmetric? Can I state anything by knowing that $\mathbb{E}[X(t)^2]=t^{2H}$?


  • $\begingroup$ Is there nobody who can help? $\endgroup$ – Ichigo Jan 14 '13 at 16:18

If $(B_s)_{s\geqslant0}$ is a standard Brownian motion, so is $(\bar B_s)_{s\geqslant0}$, where $\bar B_s=-B_s$ for every $s$. Hence, if $$ X_t=\int\varphi(s,t)\mathrm dB_s,\qquad \bar X_t=\int\varphi(s,t)\mathrm d\bar B_s, $$ for some suitable function $\varphi$, then the distributions of $X_t$ and $\bar X_t$ coincide. Since $\bar X_t=-X_t$, the proof is complete.


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.