Serving the Quantitative Finance Community

 
User avatar
pizza
Topic Author
Posts: 0
Joined: October 25th, 2002, 2:46 pm

Vega vs Gamma

September 27th, 2008, 9:32 am

The analytical expressions for Gamma and Vega of an European call areWhen plotted against S, they look very similar, and yet basically Vega = k * Gamma * S^2.What am I missing?Thanks for help.
 
User avatar
daveangel
Posts: 5
Joined: October 20th, 2003, 4:05 pm

Vega vs Gamma

September 27th, 2008, 10:07 am

This might not be what you are asking (as your question is fairly open ended) but dont forget that d1 is also a function of S.
knowledge comes, wisdom lingers
 
User avatar
pizza
Topic Author
Posts: 0
Joined: October 25th, 2002, 2:46 pm

Vega vs Gamma

September 27th, 2008, 10:36 am

Sure, but it's the same in both formulae, no?I suspect there's something special about n(d(S))) that somehow neutralises the S^2 both "in the large" and "in the small".Basically my question is, how can the plot of a f(x) look the same as x^2 * f(x)? (Aside from a scaling factor?)
 
User avatar
Paul
Posts: 7055
Joined: July 20th, 2001, 3:28 pm

Vega vs Gamma

September 27th, 2008, 10:07 pm

All the action in gamma is centred on where d_1 is zero (call this S*), outside that its value falls of rapidly, getting exponetially small. So the difference between vega and gamma is to a first approximation just S*^2.P
 
User avatar
AVt
Posts: 90
Joined: December 29th, 2001, 8:23 pm

Vega vs Gamma

September 28th, 2008, 6:39 pm

or: vega = vol*t * S^2 * gamma
 
User avatar
MCarreira
Posts: 64
Joined: January 1st, 1970, 12:00 am

Vega vs Gamma

September 28th, 2008, 7:51 pm

QuoteOriginally posted by: PaulAll the action in gamma is centred on where d_1 is zero (call this S*), outside that its value falls of rapidly, getting exponetially small. So the difference between vega and gamma is to a first approximation just S*^2.PIt's as Paul said ... the exponential decay on n(d1) is much more important than the linear change in S.The attached charts shows how the ratio between the similar charts (forward greeks, t=1,K=1: n(d1), vol*gamma, vega) is equal to S or 1/S ... so although they look similar the ratio is as expected.
Attachments
VegaGamma.zip
(60.53 KiB) Downloaded 80 times
 
User avatar
pizza
Topic Author
Posts: 0
Joined: October 25th, 2002, 2:46 pm

Vega vs Gamma

September 28th, 2008, 10:02 pm

Interesting, thanks everybody.MCarreira, it's strange you would get a linear ratio, I had expected S^2 or 1/S^2...Cheers
 
User avatar
MCarreira
Posts: 64
Joined: January 1st, 1970, 12:00 am

Vega vs Gamma

September 28th, 2008, 10:19 pm

QuoteOriginally posted by: pizzaInteresting, thanks everybody.MCarreira, it's strange you would get a linear ratio, I had expected S^2 or 1/S^2...CheersThe ratio I showed was between n(d1) and both gamma and vega; in this way you can see that the dominant shape is n(d1). The relationships are then S and 1/S.