[Quantlib-users] Simple swap and ParCoupon cashFlows

[<e.m...@ti...>]

Hello all,
we are testing the simple swap function against our pricing instruments.
We discover some strange results in the floating leg cash flows.
The simpleswap class creates the floating leg using the FloatingRateCoupo=
nVector
function.
This function uses the ParCoupon class to create the single cash flow (if=

regular)
The ParCoupon class does not use the index to compute the cash flow,
but it uses the forward rate (discount start/discount end - 1.0)/accrual.=

The problem is how the accrual is computed.
The ParCoupon class uses the termStructure dayCounter convention to compu=
te
the accrual factor.
There is any reason for that?
I would have used the index dayCounter convention, instead.
Or better: I would have computed the forward rate with the termStructure
dayCounter,
and I would have used the index DayCounter for the real accrual factor.
Am I right?

Thanks to all,
Enrico