Swap Payment Dates, Day Counts, and Discounting a Delayed Cash Flow
Summary
The answer explains how a delayed swap payment should be handled when a payment date shifts because of the calendar. If the final floating accrual period is extended by two days, the floating rate and interest payment should reflect the added accrual under the leg’s applicable day-count convention. The cash flow is then discounted from its actual, later payment date.
It also warns that swap valuation requires distinct forward and discount rates in the framework described. To determine the final floating fixing, the respondent says to build a forward-rate curve from swaps of multiple maturities. The question mentions a Vasicek model and LIBOR, but the answer does not give a detailed comparison of curve-building approaches or specify conventions for every leg. It concludes that for a longer-maturity swap, a two-day adjustment would usually be smaller than the bid-ask spread for a buy-side participant.
Key ideas
- An extended floating accrual period changes the final floating payment according to the relevant day-count convention.
- Discount the cash flow from the shifted payment date.
- Separate forward-rate construction from discounting when valuing the swap.
- A curve built from swaps with multiple maturities can supply the forward rates needed for the final fixing.
- The response expects a two-day shift to be small relative to bid-ask spread for a longer swap.
Tags
Full text
# Answer by Phil H (score 0) # Pricing of Interest rate swap with start ex. 01/06/2015 to 03/06/2015 - 2 extra days? Change discount factor and fixed payments? I hope you can help me. So let say we have an interest rate swap, with the following characteristic: Start in 30/06/2015. End in 02/07/2019 It has fixed payment every year, and floating every half year. The floating leg = From Hull (2012) the floating leg is just set equal to the notional, if we just have had a payment. However, the fixed leg: Due to weekends, in ex. year 3, the fixed payment will be done 02/07/2018. Fixed leg have day count convention: 30/360. So basically, do we change the discount factor and the payment because of the extra 2 days? Instead of year 3, we will have year 3.005 to discount the cash flow. (using the convention actual/365 -> have read this was normal) and do we change the payment as well? Can we get 1.005 * Notional * Swap rate. (30/360) -> again we are two days late? One more question: I am currently using the Vasicek model to find the discount factors. But I have read, that some uses the LIBOR rate instead? Do you have any experience regarding this ?? ## Answer by Phil H (score 0) https://quant.stackexchange.com/a/26020 The floating rate used to determine the interest payment in any given float leg roll depends on the fixing appropriate to the start of the roll. So if the only change to this IRS structure is to extend the final roll by 2 days, then that final rate will just be calculated to include the extra days using the day count convention, and that payment will be discounted from the slightly further payment date. However, some things you mention suggest that you are not calculating separate forward and discount rates (I.e. using traditional textbook method), so you will first need to construct a curve of forward rates using multiple swap maturities in order to work out what the final float fixing will be. In any case, for a longer maturity swap, I would expect 2 days to make less difference than the bid-ask spread, particularly for a buy-side actor.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.