Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Complex integral via complex contour

  1. Jul 12, 2009 #1
    Hi, I am using the Fourier transform to price a European put option. I have obtained the following integral:

    [tex]q(x,\tau)=\frac{1}{2\pi}\int_{-\infty}^{\infty}\frac{e^{-[\omega^2-i\omega(\gamma-3)+2(\gamma-1)]\tau}}{(1-i\omega)(2-iw)}d\omega[/tex]

    which I need to solve. I have solved this through the use of the convolution theorem, finding the inversion for

    [tex]G(x)=e^{-[\omega^2-i\omega(\gamma-3)+2(\gamma-1)]\tau}}[/tex] and

    [tex]\frac{1}{(1-i\omega)(2-iw)}[/tex]

    To obtain the result

    [tex]q(x,\tau)=e^{-\gamma\tau+x} N(-d_1)-e^{2x} N(-d_2)[/tex] with
    [tex]d_{1,2}=\frac{(\gamma\mp1)\tau+x}{\sqrt{2\tau}}[/tex]

    Where N(*) corresponds to the cumulative normal distribution.

    I want to be able to obtain this result not using the convolution theorem but rather constructing a complex contour. I am having difficulty. I attempted to construct a rectangular contour connecting the real axis between -R and R, the line Im(w)=-3/2 between -R and R and the sides of the rectangle closing off the two lines. I then took R->infinity. However when I calculate 2*pi*i Residue(w=-i) I obtain a complex number which doesnt make sense... I am not sure what I am doing wrong.

    Any suggestions would be very welcome. It is also interesting to note that the Residues (not 2*pi*i*residues) at the two poles (w=-i, w=-2i) correspond to the coefficients in the solution of N(-d1) and N(-d_2).

    Thanks in advance!
    Alex
     
  2. jcsd
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Can you offer guidance or do you also need help?
Draft saved Draft deleted



Similar Discussions: Complex integral via complex contour
Loading...