<t>This all very much depends on how you define theta. I always like to define it as d/dt assuming asset prices follow their forward curves and vols move to forward vols - and that theta doesn't include the costs of financing your portfolio. In that framework a zero coupon bond doesn't have theta, n...

If S is a lognormal asset, then so is S^2, just with a different drift and volatility. So the same formula should apply after you make those adjustments.

<t>The standard answer to this question: there are two possibilities if you go straight into work and skip the PhD. One is that you learn a lot on the job and are taught/teach yourself quant skills, and 4-5y from now you're much further ahead career-wise than you'd be if you did a PhD then started w...

They mean implied vol, not realised vol. Implied vol is just another way of expressing the market option price, though, so it doesn't say much to answer your question by saying "vol changes"; it's equivalent to saying that the option price changes.

<t>The problem is that you don't know what vol to put in for FRA_2 in the swaption calculation.You can't use the caplet vol for that FRA, because that's a vol from time=now to fixing date 2. But the swaption expires on fixing date 1, so you need the vol of FRA_2 from time=now to fixing date 1. That ...

Depends on the organization of the group, but generally middle office->front office is a more difficult transition than front office quant from one market to another.

<t>Pretty certain it's not the same. e.g. in a driftless process you can easily calculate the mean and variance of the distribution at some halfway point. Under OU you can use Bayes Rule to calculate the midpoint distribution as well; the mean and variance don't agree with the driftless case.TheBrid...

<t>Not sure I fully understand your question, but it sounds like you're asking why e.g. a 1.20 call on EUR/USD knocking up and out at 1.30 would price different than a 1/1.20 put on USD/EUR knocking down at 1/1.30?It shouldn't - they're exactly the same derivative contract. Are you properly adjustin...

<t>If there's no market to trade options in, then implied vol doesn't have any meaning - you can't hedge your vega exposure, and so you're stuck with it.Two options: you find the other side to the trade with a different customer, or you ride out the risk and dynamcially delta hedge your options. Doi...

<t>This field is still pretty young - young enough that there is no standard terminology for roles. Jobs run a gamit from pure programming roles to pure quantative research roles, and so do the programs training for them, but it's hard to tell from the name exactly where something fits in that range...

<t>The SDE for a Brownian bridge is dB(t) = ( B(T) - B(t) ) / ( T - t ) dt + dB', where B' is a regular Brownian motion. So you can write your integral asI = Integrate[ Exp[-k t] dB(t), { t, 0, T } ] = Integrate[ Exp[-k t] ( B(T) - B(t) ) / ( T - t ) dt, { t, 0, T } ] + Integrate[ Exp[-k t] dB'(t), ...

FENICS just doesn't have a terribly good model for this stuff. Not bad, but pretty out of date compared to what the street is doing.

<t>I think you're just mistaken in what copulas are for. They provide correlation between marginals: both linear and nonlinear correlation. They're not just about nonlinear correlation. The point of using copulas isn't always to get nonlinear correlation effects into a model: it's to get *any* kind ...

Say z=y/x; then Ito gives youdz = dz/dy * dy + dz/dx * dx + 1/2*d^2z/dy^2 * dy^2 + 1/2*d^2z/dx^2 * dx^2 + d^2z/(dx dy) * dx dyThe last term is where the correlation comes in; dx dy = x y rho sig_x sig_y dt.