Hi i know this question has been answered before as well.Assuming 10 year swap rate is 3.5% and 9 year swap rate is 3%, is my daily roll down = (10y Swap - 9y Swap)*not*1/360, assuming not = 10m then => daily rolldown pnl = (50bps)*10m*(1/360) = 138 EUR...?So in a 100m EUR 10y swap if I were paying I'd lose a total of 1380EUR a day due to roll down - this seems a bit cheap to me.......Further I have used the 9y swap rather than the 1y forward 9year swap as 1year from now i'd have a 9y swap assuming that the market doesn't move the 9y swap today should be equal to the 9y swap 1 year from now..Does this make sense or am I being thick about something here...?

This brings me to another question - the payer in a swap is going to lose carry, further he is also short convexity - where is his advantage in a swap?Thanks,Doc

Looks correct... Who says there has to be an "advantage"?

Well the payer seems to be at an inherent disadvantage in the fact that:He is short convexity and he also always seems to be lose money if the market stays still...i am sure you know but this seems to be a bit of a case like a short gamma short theta case in options..(he loses if the market moves and if the market sits still!)So the swap at outset isn't really fair because of the market is static and/or rallies the payer would lose money..Maybe this doesn't make much sense but let me know your thoughts.. (maybe im missing something small here..)

Well, don't you think that the mkt price, i.e. the rate in this case, incorporates whatever asymmetries might exist?P. S. I thought I should mention that Antti Ilmanen's "Understanding the Yield Curve" series of papers talks about this and I can't recommend it highly enough, since it talks about all these things in some detail.

The convexity is priced into the rate already and is related to volatility. This is why the yield curve may even start going down again at some point in the long end...meaning the payer starts paying less for what you call "disadvantage"..... so it is not really a disadvantage at all, as the mkt will correct for it. 100 year swap will probably have a much lower rate than a 30yr swap due to convexity.

ahhhh ok makes sense i have the series of papers haven't gotten around to actually reading them though - thanks for the advice anyway Gmike and Martinghoul!