QuantLib › quantlib-users

price bermudan swaption with 2 curves ?

Classic

List

Threaded

8 messages Options

alex belyakov

price bermudan swaption with 2 curves ?

how to price bermudan swaption with different discounting and forwarding curves in quantlib?

------------------------------------------------------------------------------
LIMITED TIME SALE - Full Year of Microsoft Training For Just $49.99!
1,500+ hours of tutorials including VisualStudio 2012, Windows 8, SharePoint
2013, SQL 2012, MVC 4, more. BEST VALUE: New Multi-Library Power Pack includes
Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9/22/13.
http://pubads.g.doubleclick.net/gampad/clk?id=64545871&iu=/4140/ostg.clktrk
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

Luigi Ballabio

Re: price bermudan swaption with 2 curves ?

As far as I know, it's not possible at this time. The available
engines for Bermudan swaptions all work on trees, which don't separate
forwarding and discounting.

Luigi

On Sat, Sep 21, 2013 at 4:35 PM, Alexander Belyakov <[hidden email]> wrote:

> how to price bermudan swaption with different discounting and forwarding
> curves in quantlib?
>
>
> ------------------------------------------------------------------------------
> LIMITED TIME SALE - Full Year of Microsoft Training For Just $49.99!
> 1,500+ hours of tutorials including VisualStudio 2012, Windows 8, SharePoint
> 2013, SQL 2012, MVC 4, more. BEST VALUE: New Multi-Library Power Pack
> includes
> Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9/22/13.
> http://pubads.g.doubleclick.net/gampad/clk?id=64545871&iu=/4140/ostg.clktrk
> _______________________________________________
> QuantLib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>

--
<https://implementingquantlib.blogspot.com>
<https://twitter.com/lballabio>

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60133471&iu=/4140/ostg.clktrk
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

Klaus Spanderen-2

Re: price bermudan swaption with 2 curves ?

the FdHullWhiteSwaptionEngine and FdG2SwaptionEngine support separate forward and discount curves. Assumption is that both curves share the same OU (or G2++) dynamics. In the absense of reference results test cases are still missing for a multi curve set-up.

regards

Klaus

On Wednesday, September 25, 2013 12:24:37 PM Luigi Ballabio wrote:

> As far as I know, it's not possible at this time. The available

> engines for Bermudan swaptions all work on trees, which don't separate

> forwarding and discounting.

> Luigi

> On Sat, Sep 21, 2013 at 4:35 PM, Alexander Belyakov <[hidden email]> wrote:

> > how to price bermudan swaption with different discounting and forwarding

> > curves in quantlib?

> >

> > --------------------------------------------------------------------------

> > ---- LIMITED TIME SALE - Full Year of Microsoft Training For Just $49.99!

> > 1,500+ hours of tutorials including VisualStudio 2012, Windows 8,

> > SharePoint 2013, SQL 2012, MVC 4, more. BEST VALUE: New Multi-Library

> > Power Pack includes

> > Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9/22/13.

> > http://pubads.g.doubleclick.net/gampad/clk?id=64545871&iu=/4140/ostg.clktr

> > k

> > _______________________________________________

> > QuantLib-users mailing list

> > [hidden email]

> > https://lists.sourceforge.net/lists/listinfo/quantlib-users

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60133471&iu=/4140/ostg.clktrk
_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

Luigi Ballabio

Re: price bermudan swaption with 2 curves ?

Ok, I stand corrected. Twice.

Luigi

On Sep 25, 2013 10:52 PM, "Klaus Spanderen" <[hidden email]> wrote:

Hi

the FdHullWhiteSwaptionEngine and FdG2SwaptionEngine support separate forward and discount curves. Assumption is that both curves share the same OU (or G2++) dynamics. In the absense of reference results test cases are still missing for a multi curve set-up.

regards

Klaus

On Wednesday, September 25, 2013 12:24:37 PM Luigi Ballabio wrote:

> As far as I know, it's not possible at this time. The available

> engines for Bermudan swaptions all work on trees, which don't separate

> forwarding and discounting.

>

>

> Luigi

>

> On Sat, Sep 21, 2013 at 4:35 PM, Alexander Belyakov <[hidden email]> wrote:

> > how to price bermudan swaption with different discounting and forwarding

> > curves in quantlib?

> >

> >

> > --------------------------------------------------------------------------

> > ---- LIMITED TIME SALE - Full Year of Microsoft Training For Just $49.99!

> > 1,500+ hours of tutorials including VisualStudio 2012, Windows 8,

> > SharePoint 2013, SQL 2012, MVC 4, more. BEST VALUE: New Multi-Library

> > Power Pack includes

> > Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9/22/13.

> > http://pubads.g.doubleclick.net/gampad/clk?id=64545871&iu=/4140/ostg.clktr

> > k

> > _______________________________________________

> > QuantLib-users mailing list

> > [hidden email]

> > https://lists.sourceforge.net/lists/listinfo/quantlib-users

Peter Caspers-4

Re: price bermudan swaption with 2 curves ?

a quick test of the two engines shows similar results in both
single and multicurve settings:

Hull White 1F, model volatility 1%, reversion speed 1%
bermudan swaption 3% vs Euribor6m, 10y maturity, yearly exercises

6m forwarding curve = discounting curve = flat 3%

GSR Integral Engine 0.054418
TreeSwaptionEngine 0.054443
FD Engine 0.054424

6m forwarding curve flat 3%, discounting curve flat 1%

GSR Integral Engine 0.061361
FD Engine 0.061363

as a side note: the FD engine seems to consider all flows with

payment date > exercise date

to be part of the exercise into right. From my experience a more
suitable criterion to identify the optional flows is

reset date >= exercise date

what do you think ?

kind regards
Peter

Luigi Ballabio <[hidden email]> writes:

> Ok, I stand corrected. Twice.
>
> Luigi
>
> On Sep 25, 2013 10:52 PM, "Klaus Spanderen" <[hidden email]>
> wrote:
>
>
> Hi
>
>
>
> the FdHullWhiteSwaptionEngine and FdG2SwaptionEngine support
> separate forward and discount curves. Assumption is that both
> curves share the same OU (or G2++) dynamics. In the absense of
> reference results test cases are still missing for a multi curve
> set-up.
>
>
>
> regards
>
> Klaus
>
>
>
> On Wednesday, September 25, 2013 12:24:37 PM Luigi Ballabio
> wrote:
>
> > As far as I know, it's not possible at this time. The available
>
> > engines for Bermudan swaptions all work on trees, which don't
> separate
>
> > forwarding and discounting.
>
> >
>
> >
>
> > Luigi
>
> >
>
> > On Sat, Sep 21, 2013 at 4:35 PM, Alexander Belyakov <
> [hidden email]> wrote:
>
> > > how to price bermudan swaption with different discounting and
> forwarding
>
> > > curves in quantlib?
>
> > >
>
> > >
>
> > >
> --------------------------------------------------------------------------
>
> > > ---- LIMITED TIME SALE - Full Year of Microsoft Training For
> Just $49.99!
>
> > > 1,500+ hours of tutorials including VisualStudio 2012,
> Windows 8,
>
> > > SharePoint 2013, SQL 2012, MVC 4, more. BEST VALUE: New
> Multi-Library
>
> > > Power Pack includes
>
> > > Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9
> /22/13.
>
> > > http://pubads.g.doubleclick.net/gampad/clk?id=64545871&iu=/
> 4140/ostg.clktr
>
> > > k
>
> > > _______________________________________________
>
> > > QuantLib-users mailing list
>
> > > [hidden email]
>
> > > https://lists.sourceforge.net/lists/listinfo/quantlib-users
>
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60133471&iu=/4140/ostg.clktrk
> _______________________________________________
> QuantLib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users

Ferdinando M. Ametrano-2

Re: price bermudan swaption with 2 curves ?

to me 0.054 and 0.061 seem sensible results in the 2 different discounting scenario. I don't know if the model is correct, but that's the kind of differences I would expect

also payment date > exercise date makes sense: in a 7NC2 you can exercise in two year into a 5 year swap, etc., and the standard 5 year swap does not have any payment at its start.

please correct me if I'm wrong...

On Fri, Sep 27, 2013 at 7:51 AM, Peter Caspers <[hidden email]> wrote:

a quick test of the two engines shows similar results in both
single and multicurve settings:

Hull White 1F, model volatility 1%, reversion speed 1%
bermudan swaption 3% vs Euribor6m, 10y maturity, yearly exercises

6m forwarding curve = discounting curve = flat 3%

GSR Integral Engine 0.054418
TreeSwaptionEngine 0.054443
FD Engine 0.054424

6m forwarding curve flat 3%, discounting curve flat 1%

GSR Integral Engine 0.061361
FD Engine 0.061363

as a side note: the FD engine seems to consider all flows with

payment date > exercise date

to be part of the exercise into right. From my experience a more
suitable criterion to identify the optional flows is

reset date >= exercise date

what do you think ?

kind regards
Peter

Luigi Ballabio <[hidden email]> writes:

> Ok, I stand corrected. Twice.
>
> Luigi
>
> On Sep 25, 2013 10:52 PM, "Klaus Spanderen" <[hidden email]>
> wrote:
>
>
> Hi
>
>
>
> the FdHullWhiteSwaptionEngine and FdG2SwaptionEngine support
> separate forward and discount curves. Assumption is that both
> curves share the same OU (or G2++) dynamics. In the absense of
> reference results test cases are still missing for a multi curve
> set-up.
>
>
>
> regards
>
> Klaus
>
>
>
> On Wednesday, September 25, 2013 12:24:37 PM Luigi Ballabio
> wrote:
>
> > As far as I know, it's not possible at this time. The available
>
> > engines for Bermudan swaptions all work on trees, which don't
> separate
>
> > forwarding and discounting.
>
> >
>
> >
>
> > Luigi
>
> >
>
> > On Sat, Sep 21, 2013 at 4:35 PM, Alexander Belyakov <
> [hidden email]> wrote:
>
> > > how to price bermudan swaption with different discounting and
> forwarding
>
> > > curves in quantlib?
>
> > >
>
> > >
>
> > >
> --------------------------------------------------------------------------
>
> > > ---- LIMITED TIME SALE - Full Year of Microsoft Training For
> Just $49.99!
>
> > > 1,500+ hours of tutorials including VisualStudio 2012,
> Windows 8,
>
> > > SharePoint 2013, SQL 2012, MVC 4, more. BEST VALUE: New
> Multi-Library
>
> > > Power Pack includes
>
> > > Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9
> /22/13.
>
> > > http://pubads.g.doubleclick.net/gampad/clk?id=64545871&iu=/
> 4140/ostg.clktr
>
> > > k
>
> > > _______________________________________________
>
> > > QuantLib-users mailing list
>
> > > [hidden email]
>
> > > https://lists.sourceforge.net/lists/listinfo/quantlib-users
>
>
>

> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60133471&iu=/4140/ostg.clktrk

> _______________________________________________
> QuantLib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60133471&iu=/4140/ostg.clktrk

_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

Peter Caspers-4

Re: price bermudan swaption with 2 curves ?

thank you, yes I also think the pricing looks fine, in particular because the code used in each case could not be more indpendent.

what I meant is, in your example 7nc2 (traded today in EUR vs. 6m) the underlying may be written

calcstart calcend=payment

01/10/13 01/10/14
01/10/14 01/10/15
01/10/15 03/10/16
03/10/16 02/10/17
02/10/17 01/10/18
01/10/18 01/10/19
01/10/19 01/10/20

and a similar 6m schedule on the float side, and the exercise schedule (assuming 2OD notification period)

notification effective

29/09/15 01/10/15
29/09/16 03/10/16
28/09/17 02/10/17
27/09/18 01/10/18
27/09/19 01/10/19

where I want to use the notification dates as the exercise dates in quantlib terminology. Then exercising on 29/09/15 would include

the payments on 01/10/15 (and similar for the subsequent notification dates), which is one too much. You could remove the first two

periods from the underlying, but then the problem still occurs for the other exercise dates.

Using the effective dates as exercise dates would work (this is what I did below, i.e. I assumed a null notice period in the end), but is in my

opinion not precisely what you want for pricing.

On 27 September 2013 09:57, Ferdinando M. Ametrano <[hidden email]> wrote:

to me 0.054 and 0.061 seem sensible results in the 2 different discounting scenario. I don't know if the model is correct, but that's the kind of differences I would expect

also payment date > exercise date makes sense: in a 7NC2 you can exercise in two year into a 5 year swap, etc., and the standard 5 year swap does not have any payment at its start.

please correct me if I'm wrong...

On Fri, Sep 27, 2013 at 7:51 AM, Peter Caspers <[hidden email]> wrote:

a quick test of the two engines shows similar results in both
single and multicurve settings:

Hull White 1F, model volatility 1%, reversion speed 1%
bermudan swaption 3% vs Euribor6m, 10y maturity, yearly exercises

6m forwarding curve = discounting curve = flat 3%

GSR Integral Engine 0.054418
TreeSwaptionEngine 0.054443
FD Engine 0.054424

6m forwarding curve flat 3%, discounting curve flat 1%

GSR Integral Engine 0.061361
FD Engine 0.061363

as a side note: the FD engine seems to consider all flows with

payment date > exercise date

to be part of the exercise into right. From my experience a more
suitable criterion to identify the optional flows is

reset date >= exercise date

what do you think ?

kind regards
Peter

Luigi Ballabio <[hidden email]> writes:

> Ok, I stand corrected. Twice.
>
> Luigi
>
> On Sep 25, 2013 10:52 PM, "Klaus Spanderen" <[hidden email]>
> wrote:
>
>
> Hi
>
>
>
> the FdHullWhiteSwaptionEngine and FdG2SwaptionEngine support
> separate forward and discount curves. Assumption is that both
> curves share the same OU (or G2++) dynamics. In the absense of
> reference results test cases are still missing for a multi curve
> set-up.
>
>
>
> regards
>
> Klaus
>
>
>
> On Wednesday, September 25, 2013 12:24:37 PM Luigi Ballabio
> wrote:
>
> > As far as I know, it's not possible at this time. The available
>
> > engines for Bermudan swaptions all work on trees, which don't
> separate
>
> > forwarding and discounting.
>
> >
>
> >
>
> > Luigi
>
> >
>
> > On Sat, Sep 21, 2013 at 4:35 PM, Alexander Belyakov <
> [hidden email]> wrote:
>
> > > how to price bermudan swaption with different discounting and
> forwarding
>
> > > curves in quantlib?
>
> > >
>
> > >
>
> > >
> --------------------------------------------------------------------------
>
> > > ---- LIMITED TIME SALE - Full Year of Microsoft Training For
> Just $49.99!
>
> > > 1,500+ hours of tutorials including VisualStudio 2012,
> Windows 8,
>
> > > SharePoint 2013, SQL 2012, MVC 4, more. BEST VALUE: New
> Multi-Library
>
> > > Power Pack includes
>
> > > Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9
> /22/13.
>
> > > http://pubads.g.doubleclick.net/gampad/clk?id=64545871&iu=/
> 4140/ostg.clktr
>
> > > k
>
> > > _______________________________________________
>
> > > QuantLib-users mailing list
>
> > > [hidden email]
>
> > > https://lists.sourceforge.net/lists/listinfo/quantlib-users
>
>
>

> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60133471&iu=/4140/ostg.clktrk

> _______________________________________________
> QuantLib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60133471&iu=/4140/ostg.clktrk

_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users

Ferdinando M. Ametrano-2

Re: price bermudan swaption with 2 curves ?

yes Peter, exercising on 29/09/15 should not include the payment(s) on 01/10/15, so using the effective dates as exercise dates is relevant.

as from measuring the time to exercise from today to notification date instead of (spot) value date to effective date: your mileage might vary, but to me it seems an ill posed problem. As a matter of fact the volatility you will use to calculate the variance has been distilled from swaption prices, so as long as you are coherent in measuring times (from value date) you should be ok using effective date.

On Fri, Sep 27, 2013 at 11:07 AM, Peter Caspers <[hidden email]> wrote:

thank you, yes I also think the pricing looks fine, in particular because the code used in each case could not be more indpendent.

what I meant is, in your example 7nc2 (traded today in EUR vs. 6m) the underlying may be written

calcstart calcend=payment

01/10/13 01/10/14
01/10/14 01/10/15
01/10/15 03/10/16
03/10/16 02/10/17
02/10/17 01/10/18
01/10/18 01/10/19
01/10/19 01/10/20

and a similar 6m schedule on the float side, and the exercise schedule (assuming 2OD notification period)

notification effective

29/09/15 01/10/15
29/09/16 03/10/16
28/09/17 02/10/17
27/09/18 01/10/18
27/09/19 01/10/19

where I want to use the notification dates as the exercise dates in quantlib terminology. Then exercising on 29/09/15 would include

the payments on 01/10/15 (and similar for the subsequent notification dates), which is one too much. You could remove the first two

periods from the underlying, but then the problem still occurs for the other exercise dates.

Using the effective dates as exercise dates would work (this is what I did below, i.e. I assumed a null notice period in the end), but is in my

opinion not precisely what you want for pricing.

On 27 September 2013 09:57, Ferdinando M. Ametrano <[hidden email]> wrote:

to me 0.054 and 0.061 seem sensible results in the 2 different discounting scenario. I don't know if the model is correct, but that's the kind of differences I would expect

also payment date > exercise date makes sense: in a 7NC2 you can exercise in two year into a 5 year swap, etc., and the standard 5 year swap does not have any payment at its start.

please correct me if I'm wrong...

On Fri, Sep 27, 2013 at 7:51 AM, Peter Caspers <[hidden email]> wrote:

a quick test of the two engines shows similar results in both
single and multicurve settings:

Hull White 1F, model volatility 1%, reversion speed 1%
bermudan swaption 3% vs Euribor6m, 10y maturity, yearly exercises

6m forwarding curve = discounting curve = flat 3%

GSR Integral Engine 0.054418
TreeSwaptionEngine 0.054443
FD Engine 0.054424

6m forwarding curve flat 3%, discounting curve flat 1%

GSR Integral Engine 0.061361
FD Engine 0.061363

as a side note: the FD engine seems to consider all flows with

payment date > exercise date

to be part of the exercise into right. From my experience a more
suitable criterion to identify the optional flows is

reset date >= exercise date

what do you think ?

kind regards
Peter

Luigi Ballabio <[hidden email]> writes:

> Ok, I stand corrected. Twice.
>
> Luigi
>
> On Sep 25, 2013 10:52 PM, "Klaus Spanderen" <[hidden email]>
> wrote:
>
>
> Hi
>
>
>
> the FdHullWhiteSwaptionEngine and FdG2SwaptionEngine support
> separate forward and discount curves. Assumption is that both
> curves share the same OU (or G2++) dynamics. In the absense of
> reference results test cases are still missing for a multi curve
> set-up.
>
>
>
> regards
>
> Klaus
>
>
>
> On Wednesday, September 25, 2013 12:24:37 PM Luigi Ballabio
> wrote:
>
> > As far as I know, it's not possible at this time. The available
>
> > engines for Bermudan swaptions all work on trees, which don't
> separate
>
> > forwarding and discounting.
>
> >
>
> >
>
> > Luigi
>
> >
>
> > On Sat, Sep 21, 2013 at 4:35 PM, Alexander Belyakov <
> [hidden email]> wrote:
>
> > > how to price bermudan swaption with different discounting and
> forwarding
>
> > > curves in quantlib?
>
> > >
>
> > >
>
> > >
> --------------------------------------------------------------------------
>
> > > ---- LIMITED TIME SALE - Full Year of Microsoft Training For
> Just $49.99!
>
> > > 1,500+ hours of tutorials including VisualStudio 2012,
> Windows 8,
>
> > > SharePoint 2013, SQL 2012, MVC 4, more. BEST VALUE: New
> Multi-Library
>
> > > Power Pack includes
>
> > > Mobile, Cloud, Java, and UX Design. Lowest price ever! Ends 9
> /22/13.
>
> > > http://pubads.g.doubleclick.net/gampad/clk?id=64545871&iu=/
> 4140/ostg.clktr
>
> > > k
>
> > > _______________________________________________
>
> > > QuantLib-users mailing list
>
> > > [hidden email]
>
> > > https://lists.sourceforge.net/lists/listinfo/quantlib-users
>
>
>

> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60133471&iu=/4140/ostg.clktrk

> _______________________________________________
> QuantLib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users

------------------------------------------------------------------------------
October Webinars: Code for Performance
Free Intel webinars can help you accelerate application performance.
Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
the latest Intel processors and coprocessors. See abstracts and register >
http://pubads.g.doubleclick.net/gampad/clk?id=60133471&iu=/4140/ostg.clktrk

_______________________________________________
QuantLib-users mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/quantlib-users