Re: [Quantlib-users] SOFR curve bootstrapping

[philippe h. <pha...@ma...>]

Very convincing argument!!

Regards

Philippe Hatstadt
+1-203-252-0408


> On Oct 16, 2024, at 7:20 AM, Tom Anderson <tw...@ur...> wrote:
> 
> On Tue, 15 Oct 2024, philippe hatstadt wrote:
> 
>> I suppose I could bootstrap a linear or stepwise convexity adjustment of futures and calibrate it so that I reprice OIS swaps for which there are market quotes?
> 
> Indeed.
> 
>> Or I might as well just build my curve with the latter and price forwards from it.
> 
> Indeed!
> 
> You may feel there are advantages to going the indirect route and using futures plus a calibrated convexity adjustment. For example, there is solid market data for futures at three-month intervals out for the first few years. How confident are you that OIS quotes at corresponding points represent real, tradeable prices, rather than figments of dealers' imaginations? In calibrating the convexity, you could stick to more reliable benchmark points for the OIS side, while the final curve gains greater granularity from the futures. You may or may not find this argument convincing!
> 
> tom
> 
>> Regards
>> 
>> Philippe Hatstadt
>> +1-203-252-0408
>> 
>> 
>>>> On Oct 15, 2024, at 9:36 AM, Tom Anderson <tw...@ur...> wrote:
>>> 
>>> On Mon, 14 Oct 2024, philippe hatstadt via QuantLib-users wrote:
>>> 
>>>> Does QuantLib have the ability to build a SOFR curve by bootstrapping SOfR futures in the front. and swaps in the back?
>>> 
>>> Yes. Just use the appropriate combination of rate helpers:
>>> 
>>> https://www.quantlib.org/reference/class_quant_lib_1_1_overnight_index_future_rate_helper.html
>>> 
>>> https://www.quantlib.org/reference/class_quant_lib_1_1_o_i_s_rate_helper.html
>>> 
>>>> What about convexity adjustment for futures versus forwards?
>>> 
>>> There is a convexityAdjustment parameter to the OvernightIndexFutureRateHelper constructor.
>>> 
>>>> Is there a closed form aproximate on available?
>>> 
>>> HullWhite::convexityBias can give you a convexity adjustment given a volatility and some other parameters, if that's what you're after:
>>> 
>>> https://rkapl123.github.io/QLAnnotatedSource/db/df5/class_quant_lib_1_1_hull_white.html#a61adab892d30472df50fa149f29478e7
>>> 
>>> Although i'm not sure it's the correct calculation for accruing futures like SOFR - this dates from the LIBOR era.
>>> 
>>> Also, in my experience, this sort of thing doesn't work. Market prices for futures plus market quotes for volatility do not result in forwards compatible with market prices for swaps. There are other forces at work which push swaps and futures prices around independently.
>>> 
>>>> Lastly what about options on SOFR futures?
>>> 
>>> QuantLib has extensive support for various sorts of work with options but i've never used any of it.
>>> 
>>> tom
>>> 
>>> --
>>> Interesting, but possibly aimed at madmen. -- Charlie Brooker, on
>>> Torchwood
>> 
> 
> --
> The book reminds me of reading unformatted code, something from a
> decompiler. -- dade_, on 'Infinite Jest'