Re: [Quantlib-users] 1st iteration: failed at 1st alive instrument

[Luigi B. <lui...@gm...>]

Is that a probability curve you're bootstrapping?  If so, it seems that
even with probability of default = 1 (the second bound
in  f[2.22045e-16,1]) the pricer can't get the sign of the NPV to change
(the two values in [1.165963e+00,3.797374e-01] are the NPVs at the
bounds).  I don't see a workaround here...

Luigi


On Thu, Apr 27, 2023 at 12:06 PM Ben Watson <ben...@ma...>
wrote:

> I know that this topic has come up before. In this case, there is nothing
> specifically wrong with the code – the same bit of code works for 350+ CDS
> curves.
>
>
>
> This issues is that the CASINO GUICHARDPERRACHON looks like it is about to
> default. The 6m CDS spread is 11659. The root solver is failing.
>
>
>
> My question is… is there anything that could be done to control the root
> solver in this case? i.e.  adjust the tolerance or some other work around?
>
>
>
> Date      Ticker    ShortName
>                              RedCode              Tier
> Ccy        DocClause          Spread6m          Spread1y
> Spread2y            Spread3y            Spread4y
> Spread5y              Spread7y            Spread10y
>
> 2023-04-24        GROUPE             CASINO GUICHARDPERRACHON
> FG554O              SNRFOR              EUR       MM14
> 1.16596253              1.22492036        1.06136462
> 0.98585902        0.93722387              0.90058129
> 0.84189616        0.78298615
>
>
>
> 1st iteration: failed at 1st alive instrument, pillar December 20th, 2023,
> maturity December 20th, 2023, reference date April 26th, 2023: root not
> bracketed: f[2.22045e-16,1] -> [1.165963e+00,3.797374e-01]
>
>
>