Building a volatility-momentum system

By Thomas Stridsman
The article “New approaches to volatility” (Active Trader, September 2010) looked at various ways to calculate volatility, as well as how to combine volatility with momentum to form a trading strategy framework.

In most cases volatility can be explained by how much each period’s price deviates from a “baseline” price. More often than not the baseline price is the moving average of prices, although it doesn’t have to be — and sometimes shouldn’t be. (See “Baseline primer” on p. 27 for more information about different baseline prices.)

The commonly used standard deviation is a common volatility measure, as is the lesser-known mean absolute deviation, both of which were discussed in last month’s article, along with momentum and average range. “Volatility calculations” on p. 30 compares some of these volatility tools.

The average range differs from the other volatility measures in that it measures only the price movement within bars, not the movement between them. As a result, although like the other volatility measures it will fluctuate from one bar to the next, it does not scale up or down based on the number of periods in the calculation. (In other words, the average range calculated over a shorter look-back period will likely be similar to that calculated over a longer look-back period, whereas short-term and long-term standard deviation calculations will likely be much different.) This makes it possible to use the range as a “normalizing” factor for both price and volatility.

Designing a deviation and momentum strategy
This month builds on last month’s article by constructing a system consisting of standard deviation and momentum. In so doing we will think of the number of periods used to calculate a volatility distance, conceptually understood as the price distance needed to span the length of one market “cycle,” from peak to peak or trough to trough, and set the momentum length at half that number. (Note: Whether there really exists a market cycle of any length, isn’t that important; it’s simply a way to conceptualize the system’s inner logic. “New approaches to volatility” explains the cycle concept and how it relates to the look-back periods used for the volatility and momentum calculations.)

The volatility distance, which is how far price needs to travel from one extreme of the cycle to the other to generate a long or short entry, will be the standard deviation times a user-specified multiplier. The momentum portion of the system will be implied, in that the volatility distance will be the momentum needed for the price to travel the distance from one extreme to the other over half the cycle length.

The price to enter long with a stop order will be the low price half a cycle length ago plus the volatility distance. (Conceptually, the low price half a cycle ago can be thought of as somewhere near the low extreme area of the cycle.) The rule is reversed for short trades.

With the entry price levels representing extreme price boundaries within a cycle (peaks and troughs), the system enters trades when these levels are surpassed, and remains in a trade as long as the market continues to set new extreme prices. You can think of this as if the cycle within which we are operating is riding an uptrend or downtrend of an even larger cycle, moving from trough to peak or peak to trough.

For the complete article, see the October 2010 issue of Active Trader magazine. Click here to subscribe.



|

|