# Tag Info

2

Pricing via characteristic functions arises naturally in models that involve Levy processes. Therefore I can see how Black's formula for swaptions can be generalized for Levy dynamics: As in Black's model take the annuity as numeraire, and define the relevant measure $Q$ Black assumes that under this measure the swap rate is martingale GBM, that is to say ...

1

One can write for the payoff of an swaption $$\sum_i\tau_i P_{i+1}(S_{\alpha,\beta}(T_\alpha)-K)^+$$ and therefore the pricing equation follows Joshi's explainations. To derive the above equation use that the swap rate is given by $$S_{\alpha,\beta} = \sum_i \frac{\tau_iP_{i+1}}{\sum_i\tau_iP_{i+1}}F^i,$$ where $F^i$ are the corresponding forward rates. ...

1

well just take the Bachelier formula with $r=d=0$ $S_0 = S_{\alpha,\beta}$ and then multiply by the annuity. The annuity will be $$\sum \limits_i \tau_i P_{i+1}.$$ where $P_{i+1}$ is the df for $t_{i+1}.$

1

Given that you have swap rates and Cap prices (ATM, I assume), you can back out the IVs for the time periods using by bootstrapping. Strictly speaking, you would need Caplet prices for the given strikes. In such a case, You would look at the shortest dated cap and (assume) it is made up of only one caplet. You can then use black's formula and back out ...

1

Thanks to my research leader, I found what I missed. $V_{0,1}$ is vol of swaption that matures at $T_0$ which is not 0 (as I thought), rather it is maturity of the first libor. So $V_{0,1}$ is the closest available point on market. And now this is all clear with table on page 323 in section 7.4. $V_{0,2}$ is realy vol of swaption that matures at $T_0$=1y ...

1

I'm not sure that machine learning would lead to any practical solutions here. Do you really have enough data for that kind of techniques? I would suggest a different approach: assume that the exercise is optimal, but just based on a different cost function than the expected pay-off. If you can find a function that replicates well enough the past exercise ...

1

Your question is too broad, but I there is plenty of examples of uses of machine learning to mimic human behaviour. For instance deep learning has been used 25 years ago to read checks in banks, or support vector machines 15 years ago to implement artificial vision, or bayesian networks to mimic expert diagnosis. I guess it would not be that hard to use ...

Only top voted, non community-wiki answers of a minimum length are eligible