AR + option models - Anyone interested?

CFA Level II
#1

I like understanding things rather than learning by heart properties I will forget the next day. So I have done a few models in Excel to simulate an AR phenomenon and option price behaviours (with the BS equation). That was helpful for me. If you are interested I can put the file somewhere on a server and release the link later today… Tell me. Marc

#2

Here it is. Let me know if you find that interesting / helpful. http://www.hannebert.com/CFA/AR_and_BS_models.zip

#3

I have downloaded your zip yet. But what are you trying to model with AR? The asset price or the volatility of the asset price?

#4

With the AR model, I wanted to understand the “conditions” for variance stationary for AR1, AR2 and AR3 models. For example in the Secret Sauce it is said that AR1 is variance stationary with b1 < 0. Intuitively I said to myself “this cannot be” and I wanted to experiment. As matter of fact, AR1 is variance stationary only when the absolute value of b1 is less than 1.

#5

BTW, to have an AR1 or AR2 model, just put 0 in b2 / b3 or b3, respectively.

#6

Why don’t you use GARCH models for volatility? Duan Jin-chuan wrote many papers about it. Are you sure you implemented the stochastic process correctly? I have never heard anyone using AR as an option pricing model. I don’t even know what is the correct specification of the stochastic process. Normally an option pricing model with stochastic volatility process will have two random variables. To implement it properly, you need to implement advanced variance reduction technique. I highly doubt you can implement this in excel easily…

#7

ymc, There is a misunderstanding here. There are different files to understand better a) AR models and b) options. Files are in the same zip but there is NO relation between the two objectives. So I’m NOT modeling options using and AR model - never thaught you could do it before your post actually! Those models are extreemly simple: pure application of the CFA L2 program. Nothing more.

#8

ymc Wrote: ------------------------------------------------------- > Why don’t you use GARCH models for volatility? > Duan Jin-chuan wrote many papers about it. > > Are you sure you implemented the stochastic > process correctly? I have never heard anyone using > AR as an option pricing model. I don’t even know > what is the correct specification of the > stochastic process. > > Normally an option pricing model with stochastic > volatility process will have two random variables. > To implement it properly, you need to implement > advanced variance reduction technique. I highly > doubt you can implement this in excel easily… I think autoregressive doesn’t matter in option pricing in the same way that drift doesn’t matter.