Assuming the dynamics of the exchange rate between two currencies at time $t$ is given by:

$$ dX_t=\Delta r X_t dt+ σ X_t dW_t$$

Is the FX Reverse process $\frac{1}{X_t}$ a brownian motion?

How can Ito's Lemma be applied to prove that?

  • $\begingroup$ Plese write down your model for $FX$ - then on can look at $1/FX$. $\endgroup$ – Ric Nov 20 '14 at 14:18
  • $\begingroup$ 1/Fx is supposed to be brownian process. How to prove it. $\endgroup$ – user13524 Nov 20 '14 at 16:45
  • $\begingroup$ Its a GBM Process not a brownian. See below. $\endgroup$ – Drew Nov 20 '14 at 19:30
  • $\begingroup$ I tried to keep you notation but frankly the way you expressed your question and added details in comments were not very community-friendly. Please pay some attention to formatting and clarity next time. $\endgroup$ – SRKX Nov 21 '14 at 1:00

Well, if you assume Fx is a Brownian Motion $W_t$ then $\frac{1}{X_t} = -\frac{1}{X^2_t} \bullet X_t + \frac{1}{X^3_t} \bullet \langle X\rangle_t = -\frac{1}{X^2_t} \bullet X_t + \frac{1}{X^3_t} \bullet \sigma^2 X^2_t t$.

So $d\bigl(\frac{1}{X_t}\bigr) = -\frac{1}{X_t} [(\Delta r + \sigma^2 ) dt + \sigma dW_t ]$

Setting $\frac{1}{X_t} = M_t$, we see it is not a Brownian Motion, but a Geometric Brownian Motion. I think thats the proof you were asked.

This is Yor's notation where $H \bullet X = \int H dX_s$

| improve this answer | |
  • $\begingroup$ Nice solution! I also think that it is important that it is GBM, which is a positive process. With BM we could run into serious problems at zero. $\endgroup$ – Ric Nov 21 '14 at 7:24
  • $\begingroup$ I changed it to $\Delta r$ on seeing the change made above. $\endgroup$ – Drew Nov 21 '14 at 15:13
  • $\begingroup$ Your quadratic appears missing the fraction $\frac{1}{2}$. $\endgroup$ – Gordon Apr 1 '15 at 15:54

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.