Re: [Quantlib-users] IMM rolls

[Tom A. <tw...@ur...>]

(following up mostly for the sake of anyone reading the archives!)

I'm starting with an effective date and a tenor ("computer, what is the 
rate for the two year out of IMM, IMM rolls?"), so i don't have a 
termination date to begin with. I am using MakeVanillaSwap, which requires 
a tenor and index, but makes everything else optional and tries to work it 
out. If you use MakeVanillaSwap with a two-year tenor, an IMM effective 
date, and a ThirdWednesday date generation rule, you get a standard 
termination date, rather than an IMM one. Hence, i need to work out the 
right termination date, and explicitly give it to MakeVanillaSwap.

tom

On Tue, 2 Jun 2020, Mike DelMedico wrote:

> Glad it worked.  When I build the schedules, the dates seem to flow fine,
> so not sure why your end date wouldn't line up with the proper final IMM
> date.  See below:
>
> floatSchedule = ql.Schedule(effectiveDate,
>                            maturityDate,
>                            ql.Period(3, ql.Months),
>                            calendarUSD_FedFund,
>                            ql.ModifiedFollowing,
>                            ql.ModifiedFollowing,
>                            ql.DateGeneration.ThirdWednesday,
>                            False)
>
> for i in floatSchedule:
>    print(i)
>
> June 17th, 2020
> September 16th, 2020
> December 16th, 2020
> March 17th, 2021
> June 16th, 2021
> September 15th, 2021
> December 15th, 2021
> March 16th, 2022
> June 15th, 2022
> September 21st, 2022
> December 21st, 2022
> March 15th, 2023
> June 21st, 2023
>
>
> On Tue, Jun 2, 2020 at 2:50 PM Tom Anderson <tw...@ur...> wrote:
>
>> That mostly does it, thanks!
>>
>> For some reason, i had assumed that the accrual periods for each coupon
>> were based directly on the frequency or index tenor, rather than running
>> to the next date. Too much time spent looking at futures bundles probably.
>>
>> It seems like i also need to tweak the end date. ThirdWednesday
>> specifically doesn't do that.
>>
>> tom
>>
>> On Fri, 29 May 2020, Mike DelMedico wrote:
>>
>>> I’m pretty sure you just flip the generator from ‘Backward” to
>>> “ThirdWednesday” and you should be good to go.
>>>
>>> These are very common use now given everyone’s desire to minimize line
>>> items. Much easier to compress a book of swaps that have all the cashflow
>>> dates lined up instead of a book with dozens of different reset dates.
>>>
>>> The other popular flavor is IMM effective but with standard rolls.
>>>
>>> -Mike
>>>
>>>
>>> On Fri, May 29, 2020 at 12:45 Tom Anderson <tw...@ur...> wrote:
>>>
>>>> Hello,
>>>>
>>>> There are some dreadful interest rate swaps which use "IMM rolls" - the
>>>> cashflows are on quarterly IMM dates, and the cashflow reflects an
>> accrual
>>>> over the period from that IMM date to the next one.
>>>>
>>>> (I think. I can't find a precise definition of how these swaps work.
>> They
>>>> just exist.)
>>>>
>>>> Does QuantLib support such swaps?
>>>>
>>>> I can generate a custom Schedule where cashflows are on IMM dates, but i
>>>> don't know of a way to tweak the accrual periods for each cashflow.
>>>>
>>>> Thanks,
>>>> tom
>>>>
>>>> --
>>>> 12:28 <@ashak> why wouldn't you run it as root? Yuo don't get permission
>>>> denied's on things then ;)
>>>>
>>>>
>>>> _______________________________________________
>>>> QuantLib-users mailing list
>>>> Qua...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>>>>
>>>
>>
>> --
>> There are only two hard things in computer science: cache invalidation,
>> naming things, and off-by-one errors. -- anonymous
>

-- 
Wikipedia topics: lists of trains, Mortal Kombat characters, one-time
villains from Mario games, road intersections, boring suburban schools,
garage bands, cats, webcomics, Digimon, Bionicle characters, webforums,
characters from English soap operas, and Mortal Kombat characters that
don't exist -- Uncyclopedia