ql ois curve bootstrapping

[Grison PG Pierre (External DEXIA-US)]

Hello,

 

I’m encountering some problems when I try to bootstrap a OIS curve to get risk free discount factors:

 

1st iteration: failed at 20th alive instrument, maturity September 5th, 2017, reference date September 4th, 2014: root not bracketed: f[1e-013,0.908467] -> [1.474373e-001,9.462471e-001]

 

It seems that for a certain pillar (2y), the 1d solver cannot find a solution in the initial range. I tried to change the initial bounds to [0;1]  but it did not help that much. Have you already encounter this problem? How can it be solved?

 

Many thanks,

 

Pierre

 

 

------------------------------------------------------------------------------
Slashdot TV.  
Video for Nerds.  Stuff that matters.
http://tv.slashdot.org/
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

[Stefano Portolan]

Hi Pierre,

I don't know whether I can be of any help but at the beginning, while playing out with boostrapping ratehelpers, I got the same problem.
I got rid of it simply noting that I was putting WRONG rate values: In order to put a quoted 0.04 (%!!!) as 1.e-4 or 0.0004, I was simply giving 0.04 ... something like 4%.

I don't know if it applies to your case, but I see that with a 3 year maturity you are boostrapping quotes ranging between 0.1 and 0.9 ... is it right at first place?

Hoping being of some help :P

Best.
Stefano