Re: [Quantlib-users] Eris SOFR Swap Futures

[Mike D. <mik...@gm...>]

Sure no problem. Happy to help. I’ve never seen a special rate helper for
swap futures in QL but that’s probably a better question for
Luigi/Peter/Tom.  My default assumption would be that you will have to
convert back to a fair rate and use the traditional swapRateHelpers as you
suggested. Also don’t forget that those swap futures are forward imm
starting not spot.


On Mon, Oct 30, 2023 at 13:30 ze73 man <ze7...@ya...> wrote:

> Yes, understood, thanks a lot, Mike. I am trying to evaluate what is
> possible using only futures data since that is a lot cheaper to license
> than swap data. In my context, the basis you mentioned might be acceptable
> if it is well understood and then I could also look for ways to account for
> it as you wrote. I could not find an example of directly using swap futures
> in quantlib but wondered if anyone had done this. Otherwise, I'd first back
> swap rates from the futures and then use the swap helper.
>
> On Monday, October 30, 2023 at 09:59:16 AM EDT, Mike DelMedico <
> mik...@gm...> wrote:
>
>
> Keep in mind this is a CME curve you are going to build not LCH (where the
> majority of risk is transferred). So depending on your desired need of
> accuracy for whatever you are trying to value off the resulting curve, you
> need to account for that basis as well as the limited liquidity in one
> venue vs the other.  Take a look at the OI and daily volumes of one vs the
> other?
>
>
> On Mon, Oct 30, 2023 at 08:30 ze73 man via QuantLib-users <
> qua...@li...> wrote:
>
> Thanks so much for shring your thoughts, Tom and Peter!
>
> I mean to directly build the curve from the Eris futures market prices
> (available from IB for example) and from the instrument spec on the Eris
> web site. The data in their FTP server is indeed great for validation but
> we don't have a license for it. My understanding was also that we would
> need a new rate helper for OIS Swap Futures and I was wondering if anyone
> would kindly share any experience modeling swap futures in quantlib, please?
>
>
>
> On Monday, October 30, 2023 at 04:02:59 AM EDT, Peter Caspers <
> pca...@gm...> wrote:
>
>
> Well, I meant to use their PV01. But their ftp data has everything you
> need anyhow
>
> http://files.erisfutures.com/ftp/Eris_20230703_EOD_ParCouponCurve_SOFR.csv
>
> Even daily discount factors which is arguably the safest way to
> replicate their curve precisely
>
> http://files.erisfutures.com/ftp/Eris_20230703_EOD_DiscountFactors_SOFR.csv
>
> To directly build the curve from the futures prices and the PAI we
> would need a new rate helper.
>
> On Sun, 29 Oct 2023 at 19:38, Tom Anderson <tw...@ur...> wrote:
> >
> > On Sun, 29 Oct 2023, Peter Caspers wrote:
> >
> > > Hi Ze, one approach could be to convert the futures price to the
> > > equivalent par swap rate as described in the primer. Provided that you
> > > have access to the necessary additional data?
> >
> > The necessary details (accrued PAI) are openly available here:
> >
> > https://www.erisfutures.com/sofrdata
> >
> > But a problem will be that the price of an Eris future corresponds to the
> > net value of the underlying swap on the expiry date of the future, not
> > today, because it settles daily like all futures. To turn that into a
> swap
> > rate you can use for building a curve, you would need to discount that to
> > today. And to do that, you need a curve!
> >
> > Oh, or at least, you need a rate from now to the front IMM date. And you
> > can get that directly from the SOFR future which is currently in accrual,
> > so maybe this isn't so bad.
> >
> > Personally, i am a bit skeptical that there is enough liquidity in the
> > Eris markets to give good pricing information, but it's worth a shot.
> >
> > tom
> >
> > --
> > Vegetables, rice and peas.
> _______________________________________________
> QuantLib-users mailing list
> Qua...@li...
> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>
>