How to forecast implied volatility?

[Don Bright]

"Predicting" one minute ahead is simply looking at current prices to determine IV, right? What is, "is" what might be in the future, no one can know.

That's why we call it a "market" - right, LOL.

All the best,

Don

 

[annaland]

Quote from Rodney King:

Quote from annaland:

If you're interested you can read Day and Lewis (1992) "Stock Market Volatility And The Information Content Of Stock Index Options"

... If you're interested in trading based on 20 year old data, from when the market was structurally much different and the players were orders of magnitude less sophisticated and automated...

Yes, you're right that study is relatively outdated and markets have changed greatly, from tick sizes and transparency to regulations. However, my post was to show that there are studies out there that ATTEMPT at IV by using GARCH and EGARCH specifications.

 

[lauriston]

Not sure you can forecast it, but you can take advantage of extreme implied volatilities by fading them or the underlying issue. That is exactly how I trade, with put call ratios added in but not necessary condition. I am going to do a post this evening per Brett's suggestion (TraderFeed http://tinyurl.com/yqc7z4), just a brief description of how I trade. i am not good at long articles, but I'll give it a shot. Check my blog later maybe tomorrow if you are interested (do not accuse me of spamming!)

my blog: http://lauristonletter.blogspot.com/

 

[Rodney King]

Quote from Grant:

Isn’t the problem with GARCH, as used by LTCM,

... Blaming GARCH for long-tail convexity mismanagement is like blaming the laws of physics for Wernher von Braun's rockets...

 

[Walther]

Quote from Don Bright:

Guessing/forecasting, both pretty hard to do, no matter what. My point is that when selling premium based on 30Vol B-S modeling is not too good when IV rises to 45 in the near term.

And, yes, good arb fund managers are a bit more scientific about their approaches - but I even wonder about them at times (as in Amaranth, etc.) - "guessing" with such pure directional bias, not the best thing to do.

All the best,

Don

I basically agree with you this time, I just wanted to make clear that right king forecasting tools is now important part of tool chest of the best.
Maybe even in yours. Or your brother( more likely )

 

[DerivativesG]

Panzerman,

I am writing my own codes for my trading strategies in C++ and that's an excellent page on coding. Thanks for sharing.

DG.

Quote from panzerman:

Of course predicting tops and bottoms can't be done consistently by anyone, and therefore the need for "complicated" strategies.

BTW: Here is some C++ code that uses the Method of Bisection to calculate IV using Black-Scholes:

http://finance-old.bi.no/~bernt/gcc_prog/recipes/recipes/node7.html#SECTION00724000000000000000