Hi there,I am bootstrapping a Euro swap, the curve will be used for rich/cheap analysis and so by definition will not match the observed curve exactly.I was hoping to use Euribor rates at the short end, however there is a pronounced kink in the curve between the 12m Euibor rate and the 2y swap rate as can be seen from the jpeg attached. It's been there since mid 2009, I assume this is because Euibor's are unsecured and the swaps are now priced using OIS based methods?However this raises some questions:Given that I am interested in the relative value of swaps, should I continue to include Euibors in the bootstrap process?It seems likely that if I continue to include Euribors I will pick up relative pricing effects due to the change in the relationship between Euirbor rates and swap rates rather than the relationship between swap rates themselves.In light of this problem would there be any problem in just starting the bootstrap from the 1y swap rate?Given that the payments are annual this can be solved mathematically (I just assume the spot discount rate=1, ignoring the 2 day settlement). However are there any reasons based on financial theory (or anything else) why I shouldn't go down this route? This way I would only be using only swaps in the bootstrapping process, so the resulting curves would show only the relative value of the swaps, with no other effects which is appealing in my situation.If the above is deemed to be impractical what set of bootstrapping instruments should I use?I need to be able to build the curves every day for the last 10 years, so using futures may be difficult if I can't get the historic vols for the convexity adjustment.FRA's are also potentially problematic as its hard to know which ones were most liquid on an historical basis. That said if I have to use one or other of those, its good to know and I will find a way.Lastly I am using swaps with the fixed leg quoted on a 30/360 euro bond basis as I thought this was the most common (and hence liquid) quotation method, however a paper I am reading seems to think Act/365 is more common can someone please confirm which is correct. Baz

Hi Baz,A brief comment regarding your last point (unfortunately I can't provide too much insight to your earlier comments), but it seems that (for EURIBOR and LIBOR): 1 year maturity it's Annual 30/360 vs 3 month Euribor, 2-10 year maturities: Annual 30/360 vs 6 month Euribor; This is based on the ISDA FIX documenation. I hope that helps to clarify.Perhaps you could help me: I'm looking for sources to determine the most appropriate inputs to bootstrap the different curves (zero/swap and basis etc.) for various currencies? Of course there isn't necessarily one correct answer, but something to work with would be great. I have already looked into this for some currencies.Thanks,Searle

Thanks Searle,When you say sources are you looking for, papers/books?I assume you want references for 2 curve pricing?Baz

Hi Baz,I am looking for sources to determine which are the best specific instruments to use for each currency (for zero, basis curves etc.). I.e. which MM instruments to use, which futures/FRA's and up to what tenor, and then the respective bonds,swaps etc. As I said, there are no absolutes here. I have been doing this for some of the currencies already, but unless you trade in those markets (or know traders who do) then it can be hard to determine the appropriate inputs (and all their conventions).Thanks for your help Searle

Hi Searle,Sorry but I'm an outsider looking in too.I know a couple of traders in fixed income and they seem pretty cagey about giving out details.Its an interesting question and I'd love to know the answer. But from what I have heard I'm not sure there is one definitive set of instruments not in the new two curve world anyway.Baz