Options - Better To Buy or Sell?

[sinner]

Took me only about 20 minutes of google to find some really good info on:

#1 Find the best ARMA model for a subset of time-series.
#2 Apply the ARMA model to optimal ARMA model with a GARCH(1,1)
#3 Use the resulting to predict/forecast next in time-series.

in R. For free. Total code less very small, just consists of a few API calls. Does being a compsci absolve me of not understanding all the underlying stats (I get the gist?)? :

Maybe not too useful for returns (then again ) but apparently, very very useful for predicting/forecasting (monthly) volatility.

What if my time-series is mazzas jazzed up realised vol equation?

The overriding point here however is that market estimations of volatility aren't very accurate and I believe an edge can be obtained by being a better forecaster of volatility

Suddenly my head is spinning...

 

[mazzatelli]

Gee maz, that's quite an equation for 9:30AM on a Sunday!

Yeah, my life ain't so interesting these days!

Took me only about 20 minutes of google to find some really good info on:

#1 Find the best ARMA model for a subset of time-series.
#2 Apply the ARMA model to optimal ARMA model with a GARCH(1,1)
#3 Use the resulting to predict/forecast next in time-series.

in R. For free. Total code less very small, just consists of a few API calls. Does being a compsci absolve me of not understanding all the underlying stats (I get the gist?)? :

Maybe not too useful for returns (then again ) but apparently, very very useful for predicting/forecasting (monthly) volatility.

What if my time-series is mazzas jazzed up realised vol equation?

Ha! Yes it can be looked at in GARCH format, but slightly different. My emphasis was more on smoothing large events when calculating realized vol, operating on the assumption that these events don't have a large bearing on future vol. I'm not sure what Wayne used to calc hist vol above, I'm assuming its just the normal stdev calc (if so it would contain more noise than what I've suggested).