Simple Code

[L.Isella]

Dear All,
I am also interested in simple codes for QuantLib (I have just a smattering of C++ and option pricing).
If I understand correctly, the C++ code in the email prices plain vanilla European call options.
It uses the standard Black and Scholes theory for a dividend-paying stock.
No stochastic volatility, flat term structure and so on.
I am testing the results against another library (http://finance.bi.no/~bernt/gcc_prog/).
I am not totally sure whether the dividends in the example are paid continuosly or not, but if I set the yield to zero, then I am in the case of a non-dividend paying stock.
But the output of the code in the email (I simply reset the value of the yield and print out the relevant quantities) is a bit puzzling to me:

the option's maturity is 1
the underlying's value is 7
the dividend yield is 0
the risk-free rate is 0.05
the volatility is 0.1
the strike is 8
Black-Scholes value: 0

whereas the other library provides a small value (about 0.08) but definitely non-zero.
Am I misunderstanding?
Furthermore, in the case of stock which pays dividends continuously, again I get a non-zero price using the mentioned library.
Many thanks

Lorenzo

> Good afternoon,
> I just want to price a simple option with quantlib
> I know the maturity, the strike, s0, Sigma and r and I want to use Black and
> scholes.
> It must be very easy to do that with Quantlib but the example
> europeanOption.cpp causes a lot of problems to me.
> Could someone give me a simple example ??
> Thank you very much
>
> NB: in EuropeanOption, what means
> return std::exp(-r_*maturity_)
> *PlainVanillaPayoff(type_, strike_)(s0_*std::exp(x))
> *std::exp(-(x - nuT)*(x -nuT)/(2*sigma_*sigma_*maturity_))
> /std::sqrt(2.0*M_PI*sigma_*sigma_*maturity_);
> in the operator().
> Why a PI ? why multiply the payoff with a exp ?
>
> _________________________________________________________________

> Gilles,
>
> In case it helps, here is the EuropeanOption.cpp example stripped to
> its most basics:
>
> #include <ql/quantlib.hpp>
> #include <iostream>
> using namespace QuantLib;
>
> int main()
> {
> Option::Type type(Option::Call);
> Real underlying = 7.00, strike = 8.00;
> Spread dividendYield = 0.05;
> Rate riskFreeRate = 0.05;
> Volatility volatility = 0.10;
> Date todaysDate(15, May, 1998);
> Date settlementDate(17, May, 1998), exerciseDate(17, May, 1999);
> Settings::instance().evaluationDate() = todaysDate;
> DayCounter dayCounter = Actual365Fixed();
> Time maturity = dayCounter.yearFraction(settlementDate,exerciseDate);
>
> boost::shared_ptr<Exercise> exercise(new EuropeanExercise(exerciseDate));
> Handle<Quote> underlyingH(boost::shared_ptr<Quote>(new
> SimpleQuote(underlying)));
>
> Handle<YieldTermStructure>
> flatTermStructure(boost::shared_ptr<YieldTermStructure>(
> new FlatForward(settlementDate, riskFreeRate, dayCounter)));
> Handle<YieldTermStructure>
> flatDividendTS(boost::shared_ptr<YieldTermStructure>(
> new FlatForward(settlementDate, dividendYield, dayCounter)));
> Handle<BlackVolTermStructure>
> flatVolTS(boost::shared_ptr<BlackVolTermStructure>(
> new BlackConstantVol(settlementDate, volatility, dayCounter)));
>
> boost::shared_ptr<StrikedTypePayoff> payoff(
> new PlainVanillaPayoff(type, strike));
>
> boost::shared_ptr<BlackScholesProcess> stochasticProcess(
> new BlackScholesProcess(underlyingH, flatDividendTS,
> flatTermStructure, flatVolTS));
>
> EuropeanOption option(stochasticProcess, payoff, exercise);
>
> option.setPricingEngine(boost::shared_ptr<PricingEngine>(
> new AnalyticEuropeanEngine()));
>
> std::cout << "Black-Scholes value: " << option.NPV() << std::endl;
>
> return 0;
> }
>
>
> > Thank you very much
> >
> > NB: in EuropeanOption, what means
> > return std::exp(-r_*maturity_)
> > *PlainVanillaPayoff(type_, strike_)(s0_*std::exp(x))
> > *std::exp(-(x - nuT)*(x -nuT)/(2*sigma_*sigma_*maturity_))
> > /std::sqrt(2.0*M_PI*sigma_*sigma_*maturity_);
> > in the operator().
> > Why a PI ? why multiply the payoff with a exp ?
>
>
>
>
>
> _______________________________________________
> Quantlib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users

[Luigi Ballabio]

On 1/24/06, L.Isella <[hidden email]> wrote:

> If I understand correctly, the C++ code in the email prices plain vanilla European call options.
> It uses the standard Black and Scholes theory for a dividend-paying stock.
> No stochastic volatility, flat term structure and so on.
> I am testing the results against another library (http://finance.bi.no/~bernt/gcc_prog/).
> I am not totally sure whether the dividends in the example are paid continuosly or not, but if I set the yield to zero, then I am in the case of a non-dividend paying stock.
> But the output of the code in the email (I simply reset the value of the yield and print out the relevant quantities) is a bit puzzling to me:
>
> Black-Scholes value: 0
>
> whereas the other library provides a small value (about 0.08) but definitely non-zero.
> Am I misunderstanding?

No, but you might be doing something wrong. If I compile and run the
code in the email after setting the dividendYield variable to 0.0, I
get for the option a Black-Scholes price of 0.082.  You might want to
check your program or post it here to see what's the problem.

Later,
    Luigi

[L.Isella]

On Tue, 2006-01-24 at 15:10 +0100, Luigi Ballabio wrote:
On 1/24/06, L.Isella <[hidden email]> wrote:

> > If I understand correctly, the C++ code in the email prices plain vanilla European call options.
> > It uses the standard Black and Scholes theory for a dividend-paying stock.
> > No stochastic volatility, flat term structure and so on.
> > I am testing the results against another library (http://finance.bi.no/~bernt/gcc_prog/).
> > I am not totally sure whether the dividends in the example are paid continuosly or not, but if I set the yield to zero, then I am in the case of a non-dividend paying stock.
> > But the output of the code in the email (I simply reset the value of the yield and print out the relevant quantities) is a bit puzzling to me:
> >
> > Black-Scholes value: 0
> >
> > whereas the other library provides a small value (about 0.08) but definitely non-zero.
> > Am I misunderstanding?
>
> No, but you might be doing something wrong. If I compile and run the
> code in the email after setting the dividendYield variable to 0.0, I
> get for the option a Black-Scholes price of 0.082.  You might want to
> check your program or post it here to see what's the problem.
>
> Later,
>     Luigi
>

Sorry for the previous mail I sent by mistake.
The code I am using is the following (taken from one of the previous emails):

#include <ql/quantlib.hpp>
        #include <iostream>
        using namespace QuantLib;
       
        int main()
        {
          Option::Type type(Option::Call);
          Real underlying = 7.00, strike = 8.00;
          Spread dividendYield = 0.0; //only part of the code I modified
          Rate riskFreeRate = 0.05;
          Volatility volatility = 0.10;
          Date todaysDate(15, May, 1998);
          Date settlementDate(17, May, 1998), exerciseDate(17, May, 1999);
          Settings::instance().evaluationDate() = todaysDate;
          DayCounter dayCounter = Actual365Fixed();
          Time maturity = dayCounter.yearFraction(settlementDate,exerciseDate);
         
          boost::shared_ptr<Exercise> exercise(new EuropeanExercise(exerciseDate));
          Handle<Quote> underlyingH(boost::shared_ptr<Quote>(new
                 SimpleQuote(underlying)));
         
          Handle<YieldTermStructure>
              flatTermStructure(boost::shared_ptr<YieldTermStructure>(
                 new FlatForward(settlementDate, riskFreeRate, dayCounter)));
          Handle<YieldTermStructure>
              flatDividendTS(boost::shared_ptr<YieldTermStructure>(
                 new FlatForward(settlementDate, dividendYield, dayCounter)));
          Handle<BlackVolTermStructure>
              flatVolTS(boost::shared_ptr<BlackVolTermStructure>(
                 new BlackConstantVol(settlementDate, volatility, dayCounter)));
       
          boost::shared_ptr<StrikedTypePayoff> payoff(
                 new PlainVanillaPayoff(type, strike));
         
          boost::shared_ptr<BlackScholesProcess> stochasticProcess(
                 new BlackScholesProcess(underlyingH, flatDividendTS,
                                         flatTermStructure, flatVolTS));

          EuropeanOption option(stochasticProcess, payoff, exercise);

          option.setPricingEngine(boost::shared_ptr<PricingEngine>(
                 new AnalyticEuropeanEngine()));

          std::cout << "Black-Scholes value: " << option.NPV() << std::endl;

          return 0;
        }

I then save it in a file (simple-option.cc) and compile & run it:

lorenzo@mypc:~/Temp$ g++ -o simple simple-option.cc -lQuantLib
lorenzo@mypc:~/Temp$ ./simple
Black-Scholes value: 0

However, If I use the same parameters inside the AmericanOption example, then I get the "right" value.
It is again something trivial, 100% sure, but I copied the code from the mail!

Lorenzo

[Luigi Ballabio]

On 1/24/06, L.Isella <[hidden email]> wrote:
> The code I am using is the following (taken from one of the previous emails):
>
> ...
>
> I then save it in a file (simple-option.cc) and compile & run it:
>
> lorenzo@mypc:~/Temp$ g++ -o simple simple-option.cc -lQuantLib
> lorenzo@mypc:~/Temp$ ./simple
> Black-Scholes value: 0

Hmm. Copying the code from your email and performing the same steps gives me:
lballabio:~$ ./simple
Black-Scholes value: 0.0823425

However, I'm using the current CVS version. What QuantLib version are
you using? I'll try to compile it and reproduce the problem, but it
will take some time.

Later,
    Luigi

[L.Isella]

Hmm. Copying the code from your email and performing the same steps gives me:
lballabio:~$ ./simple
Black-Scholes value: 0.0823425

However, I'm using the current CVS version. What QuantLib version are
you using? I'll try to compile it and reproduce the problem, but it
will take some time.

Later,
   Luigi

Thanks for your efforts.
It is weird indeed...
I am on Ubuntu Linux and, as I told Dirk, I am using
the QuantLib version

 libquantlib0-dev version 0.3.9-6ubuntu3.

Sometimes I have had troubles with other tested libraries for some "details" (e.g. some function/variable whose name is uppercase in the library and called lowercase in the code), but this does not seem the case.
However, I would be curious to find out if anyone else on Ubuntu/Debian experiences the same trouble with that code.
Cheers

Lorenzo

[Fred Nastos]

Lorenzo,

Can you add the following line, near the top of the code, say right before
the first "boost::shared_ptr" definition:

        std::cout << QL_VERSION << std::endl

The output should correspond to the same version you mentioned, but it's
a simple check to make sure everything is consistent.

I am also running quantlib on linux, and have no problems.  You
could try compiling the library yourself.  It's quite straightforward.

Fred

On Tuesday 24 January 2006 10:46, L.Isella wrote:

> Hmm. Copying the code from your email and performing the same steps gives
> me: lballabio:~$ ./simple
> Black-Scholes value: 0.0823425
>
> However, I'm using the current CVS version. What QuantLib version are
> you using? I'll try to compile it and reproduce the problem, but it
> will take some time.
>
> Later,
>    Luigi
>
> Thanks for your efforts.
> It is weird indeed...
> I am on Ubuntu Linux and, as I told Dirk, I am using
> the QuantLib version
>
>  libquantlib0-dev version 0.3.9-6ubuntu3.
>
> Sometimes I have had troubles with other tested libraries for some
> "details" (e.g. some function/variable whose name is uppercase in the
> library and called lowercase in the code), but this does not seem the case.
> However, I would be curious to find out if anyone else on Ubuntu/Debian
> experiences the same trouble with that code. Cheers
>
> Lorenzo
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log
> files for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> <a href="http://sel.as-us.falkag.net/sel?cmd=lnk&kid3432&bid#0486&dat1642">http://sel.as-us.falkag.net/sel?cmd=lnk&kid3432&bid#0486&dat1642
> _______________________________________________
> Quantlib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users

[Fred Nastos]

(Ooops, I forget the the semicolon at the end of that line)

Anyway, with version 0.3.11 I also get 0.0823.

On Tuesday 24 January 2006 15:18, Fred Nastos wrote:

> Lorenzo,
>
> Can you add the following line, near the top of the code, say right before
> the first "boost::shared_ptr" definition:
>
> std::cout << QL_VERSION << std::endl
>
> The output should correspond to the same version you mentioned, but it's
> a simple check to make sure everything is consistent.
>
> I am also running quantlib on linux, and have no problems.  You
> could try compiling the library yourself.  It's quite straightforward.
>
> Fred
>
> On Tuesday 24 January 2006 10:46, L.Isella wrote:
> > Hmm. Copying the code from your email and performing the same steps gives
> > me: lballabio:~$ ./simple
> > Black-Scholes value: 0.0823425
> >
> > However, I'm using the current CVS version. What QuantLib version are
> > you using? I'll try to compile it and reproduce the problem, but it
> > will take some time.
> >
> > Later,
> >    Luigi
> >
> > Thanks for your efforts.
> > It is weird indeed...
> > I am on Ubuntu Linux and, as I told Dirk, I am using
> > the QuantLib version
> >
> >  libquantlib0-dev version 0.3.9-6ubuntu3.
> >
> > Sometimes I have had troubles with other tested libraries for some
> > "details" (e.g. some function/variable whose name is uppercase in the
> > library and called lowercase in the code), but this does not seem the
> > case. However, I would be curious to find out if anyone else on
> > Ubuntu/Debian experiences the same trouble with that code. Cheers
> >
> > Lorenzo
> >
> > -------------------------------------------------------
> > This SF.net email is sponsored by: Splunk Inc. Do you grep through log
> > files for problems?  Stop!  Download the new AJAX search engine that
> > makes searching your log files as easy as surfing the  web.  DOWNLOAD
> > SPLUNK!
> > <a href="http://sel.as-us.falkag.net/sel?cmd=lnk&kid3432&bid#0486&dat1642">http://sel.as-us.falkag.net/sel?cmd=lnk&kid3432&bid#0486&dat1642
> > _______________________________________________
> > Quantlib-users mailing list
> > [hidden email]
> > https://lists.sourceforge.net/lists/listinfo/quantlib-users
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log
> files for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
> _______________________________________________
> Quantlib-users mailing list
> [hidden email]
> https://lists.sourceforge.net/lists/listinfo/quantlib-users

--
_______________________________________________________
Fred Nastos
Ph.D. Candidate

Department of Physics Tel: [phone]
University of Toronto Fax: [phone]
60 St. George Street
Toronto, ON  M5S 1A7

Web: www.physics.utoronto.ca/~nastos
_______________________________________________________

[Dirk Eddelbuettel]

It looks like it is an issue with Ubuntu. or the older QL 0.3.9 library.
I also get Lorenzo's invalid answer of 0 when I use the supplied QL 0.3.9
packages on Ubuntu.

Interestingly enough, I do get the correct answer using RQuantLib rather than
the little C++ source code:

edd@joe:~$ echo 'library(RQuantLib); EuropeanOption("call", 7, 8, 0, 0.05, 1, 0.1)' | R --slave
Concise summary of valuation for EuropeanOption
      value       delta       gamma        vega       theta         rho
 0.08234248  0.21613477  0.41869080  2.05158494 -0.17410929  1.43060089
     divRho
-1.51294336

This uses the standard Ubuntu packages, rebuilt for the C++ transition:

edd@joe:~$ dpkg -l r-base-core libquantlib-0.3.9c2 r-cran-rquantlib
Desired=Unknown/Install/Remove/Purge/Hold
| Status=Not/Installed/Config-files/Unpacked/Failed-config/Half-installed
|/ Err?=(none)/Hold/Reinst-required/X=both-problems (Status,Err: uppercase=bad)
||/ Name           Version        Description
+++-==============-==============-============================================
ii  libquantlib-0. 0.3.9-6ubuntu3 Quantitative Finance Library -- development
ii  r-base-core    2.1.1-1        GNU R core of statistical computing language
ii  r-cran-rquantl 0.1.12-3build1 GNU R package interfacing the QuantLib finan


I don't have access to an older Debian system with 0.3.9 to compare; with
0.3.11 the correct answer comes up with the same C++ code posted a few days
ago.  

Dirk

--
Hell, there are no rules here - we're trying to accomplish something.
                                                  -- Thomas A. Edison

[Dirk Eddelbuettel]

It is something in the date calcs that must have gotten fixed since 0.3.9.

If we forward the dates from 1998/1999 to 2006/2007, it all works:

edd@joe:~$ grep 200 euopt.cpp
  Date todaysDate(15, May, 2006);
  Date settlementDate(17, May, 2006), exerciseDate(17, May, 2007);
edd@joe:~$ g++ -o euopt euopt.cpp -lQuantLib
edd@joe:~$ ./euopt
Black-Scholes value: 0.0823425

even on the Ubuntu system with the older 0.3.9 library.

Dirk

--
Hell, there are no rules here - we're trying to accomplish something.
                                                  -- Thomas A. Edison

[zheng wang-4]

I am checking the Excel AddIns and have a few questions:
a. When settlementdate is bigger than valuateDate, DF numbers are undefined.
b. Can we set different valuatedate and settlementdate for each instrument for the curve
c. what is use of the cell G3 in swap convention section?
d. what is trigger used for?
Thanks.
-Jim

---

Yahoo! Mail
Use Photomail to share photos without annoying attachments.

[zheng wang-4]

One more:  the BPS in the SimpleSwap has numbers of over 8,000,000
which seems too big. what is the definition for it?

zheng wang <[hidden email]> wrote:

I am checking the Excel AddIns and have a few questions:
a. When settlementdate is bigger than valuateDate, DF numbers are undefined.
b. Can we set different valuatedate and settlementdate for each instrument for the curve
c. what is use of the cell G3 in swap convention section?
d. what is trigger used for?
Thanks.
-Jim

---

Yahoo! Mail
Use Photomail to share photos without annoying attachments.

---

Do you Yahoo!?
With a free 1 GB, there's more in store with Yahoo! Mail.

[eric ehlers]

Hi Jim

>  a. When settlementdate is bigger than valuateDate, DF numbers are
> undefined.

Yes if all you do is increment the settlement date it breaks the
curve, you need to make the corresponding changes to the fixing days
and the dates in the curve.

>  b. Can we set different valuatedate and settlementdate for each instrument
> for the curve

No.  The evaluation date is contained the Settings object, in the
default QuantLib build there is one global instance of Settings.
QuantLib supports the concept of a session, allowing multiple
instances of the Settings object, but this is not implemented in the
current release of QuantLibAddin (0.3.11).  Please see the links below
for prior discussion on this subject.

http://sourceforge.net/mailarchive/forum.php?thread_id=8151830&forum_id=4299
http://sourceforge.net/mailarchive/forum.php?thread_id=7875915&forum_id=4299
http://sourceforge.net/mailarchive/forum.php?thread_id=7144644&forum_id=4299

The upcoming QuantLibAddin 0.3.12 release implements a session as a
workbook allowing all objects requiring a given instance of the
evaluation date to be grouped in a single workbook.  But this wouldn't
really let you use a different evaluation date for each instrument in
the curve.

>  c. what is use of the cell G3 in swap convention section?

#/fixing days for the swap rate helpers.

>  d. what is trigger used for?

To force a dependency which isn't known to Excel's calculation tree.
For example qlSetQuote and qlGetDf both operate on the same underlying
object but Excel has no way to know that so the trigger forces qlGetDf
to be recalculated if qlSetQuote changes.

Regards,
Eric

[Luigi Ballabio]

On 1/25/06, zheng wang <[hidden email]> wrote:
> One more:  the BPS in the SimpleSwap has numbers of over 8,000,000
>  which seems too big. what is the definition for it?

It is the variation of the NPV of a leg when the paid rate changes of
a basis point. Whether it's too big or not depends on the notional of
the swap. What parameters were you using?

Luigi

[Luigi Ballabio]

On 2/2/06, zheng wang <[hidden email]> wrote:

> Luigi Ballabio <[hidden email]> wrote:
> On 1/25/06, zheng wang wrote:
> > One more: the BPS in the SimpleSwap has numbers of over 8,000,000
> > which seems too big. what is the definition for it?
>
> It is the variation of the NPV of a leg when the paid rate changes of
> a basis point. Whether it's too big or not depends on the notional of
> the swap. What parameters were you using?
>
> The sample excel YC_SwapDemo.xls, which has notional of 1,000,000
> but the BPS is 8 times bigger.

Hmm. It doesn't look right. Then again, I'm not familiar with the Excel add-in.
Eric, can you step in?

Thanks,
    Luigi