Hi, I am trying to build a zero coupon curve for french bonds. I use a mix of BTF, BTAN and OAT to build my curve. I get somme funny results especially around the 2Yr BTAN, I was wondering if some of you guys could review my excel spreadsheet and eventually comment to let me know if I am right or wrong and how I can implement itthanks a lot in advanceP

It is common to run into problems building curves where data changes from one type of financial security to another (egs. futures switching into par swaps, or in your case money market instruments into par swaps).I lack time the to give you any specifics just now, but you will find it much easier to solve for discount factors first, then use a simple transformation to convert the discount factors into zero coupon rates. The formulas in your column j are extremely clumsy. You can solve for the discount factors much more simply and elegantly if you investigate and implement a concept called the cumulative discount factor (sometimes called the value of a basis point). Another thing you will need to compute is the time in years between swap coupons, so that you can weight the afore-mentioned cumulative discount factor correctly. Any decent basic article on curve building can be adapted to your particular market. Try for example the paper by Uri Ron posted on the Bank of Canada website.

thanks for your answer, I agree with you a cumulative DF approach is more elegant, although, I find it difficult with this approach to tackle the maturities of the bonds. In my S/S I take the maturities of the bonds and they are different wether it's BTAN BTF or OAT (not the same settlement date). In a cumulative approach I must assume the settlement date is the same for all of them and the fraction of year is 1, 2, 3 etc... (in france coupons are paid annually)thoughts?thanks

don't forget that a zc bond curve is different from zc swap curve. Don't forget the accrual rate; and you have to solve the problem of the lack of bond for each year.