Re: [Quantlib-users] Error in ImpliedVolatility method

[Ashish B. <ash...@gm...>]

Hi Luigi,

Thanks for your reply. I hope to keep my code the same in new version, we
only need to remove the keywords on it. No other changes are required
around IV code.

Regards
Ashish

On Fri, 2 Feb 2024 at 20:25, Luigi Ballabio <lui...@gm...>
wrote:

> Hello,
>     in version 1.30, the impliedVolatility method was overloaded (it's now
> also possible to pass a dividend schedule) and SWIG can't support keyword
> arguments in this case.
>
> In previous versions, the implied volatility was given by default a
> minimum value of 0.0001 because smaller values could cause problems in the
> calculations.  In more recent ones, those problems were solved and we could
> lower the minimum value to 1e-7.   My guess is that the correct value is
> 5e-5, but in version 1.27 it was below the minimum value and couldn't be
> returned.  The minimum value was probably good enough within accuracy and
> was returned instead of the correct one.
>
> Hope this helps,
>     Luigi
>
>
> On Fri, Feb 2, 2024 at 12:15 PM Ashish Bansal <ash...@gm...>
> wrote:
>
>> Hi all,
>>
>> We are using the QL v1.27 through Python which are upgrading to recent
>> version v1.32. We are calculating the implied volatility of the vanilla
>> option using the ImpliedVolatility method as following:
>> self.IV = option.impliedVolatility(targetValue=self.marketPremium,
>> process=process)
>>
>> This was working fine in v1.27 but in v1.32 it is given the following
>> error:
>> [image: image.png]
>>
>> I did look at the release change log but have not found anything around
>> it.
>>
>> Secondly, the value of same put option is showing difference in IV as
>> following when i remove the keywords and run:
>> v1.27: 0.0001
>> v1.32: 0.00005 (5.0100000000000005e-05)
>>
>> Kindly help if we need to make more changes for this.
>>
>> Regards
>> Ashish
>> _______________________________________________
>> QuantLib-users mailing list
>> Qua...@li...
>> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>>
>