26
Aug
2007
The Problem With Volatility
Posted by Tom as Neural Nets, Options
Is that it’s too damn volatile! I’ve been working on the next gen version of a directional volatility model and its been both frustrating and rewarding at the same time. As long as the model is re-optimized on a weekly (or daily) basis, the predictions tend to have a 60-70% accuracy.
The problem is that you can’t forward forecast it 1 year into the future because volatility is, well, “unpredictable!” 
Gasp, I said unpredictable (Christian please feel free to jump in)! Yes, its true volatility is unpredictable far out in the future BUT in smaller time frames the “stickiness law” works nicely. Noble prize winner, Robert Engle, proved (link coming) that volatility tends to cluster and that times of low volatility seem to be followed by more periods of low volatility. The same “pattern” exists for times of high volatility and I call this the stickiness law.
So predicting the direction of volatility is feasible, based on my findings, as long as you continually update your model. Since we just saw a period of high volatility recently, we’re likely to see additional periods of high volatility soon. Guess that makes sense since September and October is coming!
4 Responses
Christian Gross
August 26th, 2007 at 1:23 pm
1Funny, you mention that there are clusters… It is an interesting finding no?
Now about unpredictable, I would be very hesitant to adapt your model every week because that would imply to me that you are optimizing to the current trend. Yes I know there are those people that say you can predict in the short term, but I find those types of predictions dangerous.
If it were me I would not try to predict volatility, but figure out what the worst case scenarios could be and indicate the probability of the remainder. I find the Sherlock Holmes theory very interesting, and it is sort of what Taleb pointed out in his books.
Here is an interesting prediction, what will happen after the interest rate decision in September? The question is will it go up or down?
What I typically do is think in simple principles and then go through all of the permutations and combinations of those simple principles.
Could you predict out into the future? Good question… Have a read of “New Kind of Science”
Tom
August 27th, 2007 at 4:58 am
2CG, I’m not trying to predict volatility but rather its direction. Interestingly enough, if the model is optimized on a weekly basis, the predictions for next week tend to have a greater than 50% accuracy. Yes it is catching on the current trend but that’s what I want!
Festival of Stocks #52 - Fat Pitch Financials
September 4th, 2007 at 10:18 pm
3[...] back into the swing of things with the following Festival of Stocks submissions: The Problem With Volatility by Neural Market Trends looks at the clustering of market volatility and thinks that one can [...]
S&P500 Next Gen Volatility Model Results | Neural Market Trends
September 5th, 2007 at 4:02 pm
4[...] doubt that I’d be able to do better than 70% in predicting volatility, because of its “volatile nature,” but as a future option trader I’m willing to live with [...]
RSS feed for comments on this post · TrackBack URI
Leave a reply
previous post: Back To NJ
next post: Are Quants Dead?
to top of page...