Most academic papers refer to relative volatility as one of two things:
1. The volatility of a security compared to that of an appropriate index.
or, closely related
2. The volatility of a security explained by an appropriate index.

Then there are technical factors such as the Relative Volatility Index, available for review here, although that isn’t really what I’m talking about.

I read Curtis Faith’s new book, and one thing that he mentions is something I’ve found interesting for a while: market regimes. He refers to them as market states and the info is available on pages 25 and 26 if you have the book. While this area of study has received some coverage, Faith breaks it down into a simple system that I like: Volatility and Trend (he uses the terms stable/trending and quiet/volatile). This creates four market regimes:

Low Volatility, Trending	High Volatility, Trending
Low Volatility, Not Trending	High Volatility, Not Trending

The volatility component that describes the regime is what I call (and he may have called too, don’t remember exactly) relative volatility. My interpretation differs from the academic, in that I would like to compare the volatility of a security to its volatility at some point in history.My first foray into this was simply to create two series, the volatility xx periods back and a moving average xx periods back of the previously defined volatility. Standard deviation measures could be added to the moving average, I suppose, to determine relatively high and relatively low volatility - but I haven’t gone that far yet.This initial measure has shown some interesting results in individual futures contracts when combined with basic time-dependent always-in order techniques. My continuous contract transformation may be suboptimal though (linear instead of logarithmic, switch period, maybe?), because the results can be quite interesting when on high volume time periods for given individual contracts, but when testing on long-term continuous contracts the results aren’t nearly as interesting/profitable. Still, this measure may provide a very slight edge if partnered with correct money management and order entry rules.

I’ll post more as this volatility measure develops. A quick Google, JSTOR, and Blackwell Synergy search didn’t pull up too much in this vein, rather the aforementioned definitions, so I plan to look more. But, if anyone has any ideas or has come across similar work, please comment and let me know!

I have added another model to the Models page on foquant.com. This is a variation of the original Black Scholes model, however this uses the formulas from Black’s 1976 paper, interpreted in T.W. Epps’ Pricing Derivative Securities, for pricing options on futures.

Following up on my prior post, I have uploaded another model to the Models page of foquant.com.

This model alters the prior model, as it now allows the underlying to pay a dividend. In a sense making the original model redundant, this new model will also accurately price a non-dividend-paying underlying. But, the separation of names calls attention to the differing equations.