Forums

FpML Discussion

Viewing 2 posts - 1 through 2 (of 2 total)
  • Author
    Posts
  • #1817
    cfccpflumm
    Spectator

    We are working through the FpML schema files for compounding for IR swaps, and I’ve not been able to find a few fields we believe are needed. 1) Compounding start date – We see trades where the compounding may start at a later date than the first calculation period/payment date. I’ve not yet found this in FpML 2) Business day convention for compounding – we see the business day convention for other fields, but does FpML allow for a business day convention specific to compounding 3) Roll day for compounding. I’m thinking there would be a separate compounding roll convention than the roll convention for calculation periods. Any thoughts or input is welcome. Thanks! ———— What we have found: a. compoundingMethod and CompoundingMethodEnum b. compoundingRate and compoundingSpread c. firstCompoundingPeriodEndDate in fpml-ird-4-6.xsd.html_type_CalculationPeriodDates.html

    #1823
    mgratacos
    Keymaster

    Compounding in FpML is not expressed using a flag or something similar. It’s just expressed as a calculation frequency higher than the payment frequency (e.g., calculation quarterly (3 M) and payment anual (1 Y)). The compoundingMethod is only used when there is compounding and a spread so the treatment of the spread in the calculation needs to be clarified. Trying to answer your questions: If calculation period frequency is higher than payment frequency then there is compounding so then: 1) It would be defined by the calculation period. 2) It would be defined by the business day convention of the calculation period. 3) Again, it would be defined by the roll day of the calculation period. Why would you have different data fields to represent the calculation of a period different from the compounding fields to express such calculation? The calculation data fields express the parameters of how compounding, if present, should be calculated. Best Regards, Marc

Viewing 2 posts - 1 through 2 (of 2 total)
  • You must be logged in to reply to this topic.