Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let W be a brownian motion and p>0.

For which p does $S_t=W_t+t^p$ admit an equivalent martingale measure?

I recently saw at my lectures that

NFLVR cond: There does not exist a sequence $\{H_n\}_{n \geq 1}$ of predictable processes, integrable wrt S, such that there exists $t_0 , b , \epsilon >0$ $$ \int_0^{t_0} H_n(s) S_s > -1/n$$ and $$ P(\int_0^{t_0} H_n(s) S_s >b)> \epsilon$$

Is equivalent to having such a measure and I think that is pretty much my only tool, so I guess it has to be that. On the other hand I can't get any ideas on how to use the condition.

share|cite|improve this question
cameron-martin deals with $W_t + f(t)$ and their condition is $\int^t_0 | f^{\prime} |^2 < \infty$. – mike May 14 '13 at 15:48
Hi mike, thanks for the comment. I'm sure Cameron-Martin is not in our tool kit, but it's nice to know the "solution". – Henrik May 14 '13 at 16:29
o.k. let's put it this way: I think NFLVR is too much and a brutally direct effort to construct the likeihood ratios show they want to be $e^{\int f^{\prime}(t) dW_t - \frac 12 \int (f^{\prime})^2 (t) dt}$. Also, you can rule out small p by law of iterated logarithm – mike May 15 '13 at 11:44
You are right, I have actually solved it and will post my answer soon. It is much simpler than NFLVR. – Henrik May 15 '13 at 15:35
up vote 0 down vote accepted

Let me just give a short answer since probably nobody cares:

It can bee seen as a corollary to (at least the proof i have seen) of Girsanov that if we are looking for a equivalent martingale measure for a continuous semimartingale $S_t = M_t + A_t$ where $M_t$ is the local martingale part and $A_t$ is the bounded variation part, then $A_t$ must satisfy that there exists predictable $h_t$ such that $A_t=\int _0^t h_s d<M_s>$ and $\int_0^t h_s^2 d<M_s> < \infty $. ($<M>$ is quadratic variation). In our case this rules out $p\leq 1/2$ since $A_t=\int_0^t p\cdot s^{p-1} ds $ which is only square integrable if $p > 1/2$.

Assuming from now on $p>1/2$. We will see an Martingale measure does indeed exist. This is simply and application of Girsanov - e.g. by applying Novikovs condition which is here trivial - on $h_s = - p s^{p-1}$. This directly yields that $W_t - \int h_s d<W>_s=W_t + t^p$ is a brownian motion under $\mathbb{Q}$ (in particular a local martingale) and $\mathbb{P}$ and $\mathbb{Q}$ are equivalent.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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