QuantLib - Asset Swap Cash Flow Final Period

Question

I am comparing CFs of asset swaps in QuantLib to the asset swap screen (ASW) in Bloomberg. I noticed that the final payments of both swap legs do not include an interest payment for the final period. In Bloomberg, however, they do.

For example, taking the first asset swap from the test suite swapping DE0001135275 as of 04-24-2007.

Final periods fixed rate bond CFs

2034-01-04      4.00000
2035-01-04      4.00000
2036-01-04      3.99991
2037-01-05    104.00009

Final periods swap CFs

            fixed_leg   float_leg
2035-01-04  4.00000     2.552582
2035-07-04  NaN         2.510446
2036-01-04  3.99991     2.552582
2036-07-04  NaN         2.524490
2037-01-05  100.00000   100.000000

Is it possible to change this behavior in either QuantLib or BB so that the CFs align?

Thanks for any hints.

PS I tried posting to quantlib-users mailing list but my subscription is pending for approval so I decided to post here as well. Sorry for the duplication.

Answer 1

The final payments on both legs of an AssetSwap should obviously be included wherever you choose to value it, both for the par Asset Swap and Market Asset Swap cases.

In the QuantLib test suit you refer to, I see the final interest payments do exist, and are consistent with Bloomberg's pricer.

Taking the first example for that example, for the fixed (bond) leg you have:

for cf in list(parAssetSwap.leg(0))[-5:]:
        print(cf.date(), cf.amount())

Which outputs:

January 4th, 2034 4.0000000000000036
January 4th, 2035 4.0000000000000036
January 4th, 2036 3.9999101729171427
January 5th, 2037 4.000089827082864
January 5th, 2037 100.0

And the floating leg...

for cf in list(parAssetSwap.leg(1))[-5:]:
        print(cf.date(), cf.amount())

Which gives you:

July 4th, 2035 2.510446097912533
January 4th, 2036 2.552582336038789
July 4th, 2036 2.5244895866360917
January 5th, 2037 2.5666315972451503
January 5th, 2037 100.0