Hi,I'm searching a paper about pricing swaptions on amortizing swaps. I know, that there different ways to handle this problem (option on basket of bullet swaps, one swap with the same par rate as the amortizing swap...).My Problem is, that I can't find a paper discribing on way of pricing these options in detail (I would prefere the first one). Can anyone help me?

you can equate the NPV of a non-amortizing forward swap and the remaining part (after option date) of the amortizing swap and then price the corresponding standard swaption / put this standard option into an IR model for calibration, if you have multiple call rights.Unfortunately I am not aware of a paper, but it is quite straightforward.

QuoteOriginally posted by: pcaspersyou can equate the NPV of a non-amortizing forward swap and the remaining part (after option date) of the amortizing swap and then price the corresponding standard swaption / put this standard option into an IR model for calibration, if you have multiple call rights.Unfortunately I am not aware of a paper, but it is quite straightforward.Amortizing swaption and accreting swaptions are linear combinations of one normal European and a bunch of mid-curve type swaptions. They are not traded and there is no definitive model for it. I recommend using globally calibrated BGM model for it.

I believe Pat Hagan has written about this topic and his paper might be posted in the "file share" section. The search function on this website should enable you to track it down.

QuoteOriginally posted by: DavidJNI believe Pat Hagan has written about this topic and his paper might be posted in the "file share" section. The search function on this website should enable you to track it down.standard recomendation from me: book. Good chunk of Chapter 19 spent on amortizing and accreting swaptions

Thanks for your suggestions. I've found a paper and I think it will help.

QuoteI recommend using globally calibrated BGM model for it.a question. I had only deterministic amortization in mind. Suppose the most simple case of a single callable swap. Would you prefer to use a globally calibrated BGM to pricing the standard swaption with same PV in black scholes with market volatility?

QuoteOriginally posted by: waterQuoteOriginally posted by: pcaspersyou can equate the NPV of a non-amortizing forward swap and the remaining part (after option date) of the amortizing swap and then price the corresponding standard swaption / put this standard option into an IR model for calibration, if you have multiple call rights.Unfortunately I am not aware of a paper, but it is quite straightforward.Amortizing swaption and accreting swaptions are linear combinations of one normal European and a bunch of mid-curve type swaptions. They are not traded and there is no definitive model for it. I recommend using globally calibrated BGM model for it.I use LMM calibrated to coterminal swaptions with effective strikes (strikes of vanilla swaps with PV and PV01 equal to the the coterminal swap from the accreting swap).

QuoteOriginally posted by: katastrofaQuoteOriginally posted by: waterQuoteOriginally posted by: pcaspersyou can equate the NPV of a non-amortizing forward swap and the remaining part (after option date) of the amortizing swap and then price the corresponding standard swaption / put this standard option into an IR model for calibration, if you have multiple call rights.Unfortunately I am not aware of a paper, but it is quite straightforward.Amortizing swaption and accreting swaptions are linear combinations of one normal European and a bunch of mid-curve type swaptions. They are not traded and there is no definitive model for it. I recommend using globally calibrated BGM model for it.I use LMM calibrated to coterminal swaptions with effective strikes (strikes of vanilla swaps with PV and PV01 equal to the the coterminal swap from the accreting swap).I think the local calibration should work. I would use shifted lognormal to have approximate skew represented at the same time.