Re: [Quantlib-users] SOFR index runtime Error

[Peter C. <pca...@gm...>]

I'd use a tenor matching the coupon frequency of your swaption, e.g.
Euribor3M if the SOFR payments are quarterly in the swap underlying
the swaption. You can also attach a SOFR forecasting curve to the
Euribor3M index. I guess that would in fact be a quite good
approximation of a SOFR swaption, whether it references term SOFR or
overnight SOFR.
Thank you
Peter

On Sun, 16 Oct 2022 at 04:42, Mark <gr...@gm...> wrote:
>
> Jonathan and Peter,
>
> Thank you for your response. If I used Euribo as a proxy to SOFR for swaption calculation, will Euribo1m a better choice than Euribo6m?
>
> Thanks,
> Mark
>
> On Sat, Oct 15, 2022 at 11:58 AM Peter Caspers <pca...@gm...> wrote:
>>
>> Just to add to what Jonathan said: QuantLib does not model SOFR
>> swaptions correctly at the moment, so even if the technical issue is
>> resolved you'll not get an accurate valuation.
>> Best, Peter
>>
>> On Sat, 15 Oct 2022 at 11:23, Jonathan Sweemer <sw...@gm...> wrote:
>> >
>> > Hi Mark,
>> >
>> > You can find the source of the error message here: https://github.com/lballabio/QuantLib/blob/master/ql/processes/gsrprocess.cpp#L92-L93
>> >
>> > Looks like your Sofr term structure somehow contains a forward from 0.915068 to 0.912329, which of course is not allowed, but it's a bit hard to know why without seeing the rest of your code.
>> >
>> >
>> > On Sat, Oct 15, 2022 at 11:58 AM Mark <gr...@gm...> wrote:
>> >>
>> >> Hi,
>> >>
>> >> If I used index=Euribor6M(termstructure) as the yield curve I can calculate European swaption through GSR model, but once I changed the index to index=Sofr(termstructure), I will get below error:
>> >>
>> >> atmSwaption.NPV()
>> >> Return _QuantLib.Intrument_NPV(self)
>> >> RuntimeError:  G(t,w) should be called with w (0.912329) not lesser than t(0.915068)
>> >>
>> >> Does anyone know what this error message means?
>> >>
>> >> Thanks,
>> >> Mark
>> >> _______________________________________________
>> >> QuantLib-users mailing list
>> >> Qua...@li...
>> >> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>> >
>> > _______________________________________________
>> > QuantLib-users mailing list
>> > Qua...@li...
>> > https://lists.sourceforge.net/lists/listinfo/quantlib-users