How A CTA May Use Volatility To Set Protective Stops
A non-high-tech measure of *historical volatility is given by the range of market prices over the course of a trading interval, this is usually a day or a week. The range of prices is defined as the difference between the high and the low for that given trading interval. If the range of the current day lies beyond the range of the previous day (Gap- up or down) the current days range must include the distance between the current days range and yesterday’s close. This is what is referred to as the “True Range”. The true range for a gap-down day is the difference between yesterday’s settlement price and today’s low. On the flip side, the true range for a gap-up day is the difference between today’s high and the previous day’s settlement price.
To grind this down a bit further, a tick is the smallest increment by which prices can move in a given futures or commodity market. The next step would be to translate the dollar value for 1 tick in the given market being traded, (Ex: The minimum tick value in corn futures is $12.50 or ¼ cent). To use corn data as an example, data shows that 90 percent of all observations between 2004- 2014 had a daily true range equal to or less than 26 ¼ cents. Therefore a CTA who was long corn futures, may want to set a protective sell stop 26 ¼ cents below the previous days close, as the probability of being whip-sawed out of the market are 1 in 10. Similarly, a CTA who had short sold corn would want to set their stop at least 26 ¼ cents above the previous day’s closing price. The dollar value for this stop would be $1,312.50 per contract, in corn.
Now, instead of concentrating on the true range for a day or a week, it may be more suitable and efficient for a CTA to work with the average true range over the past “N” trading sessions, wherein “N” is any number found to be most effective through back testing their trading methodology (Ex: 9 days, 20 days, 4 weeks, etc.). The theory is that the range for the past “N” periods is a more reliable and consistent indicator of volatility as compared to the true range from the immediately preceding trading session. An example would be to calculate the average true range over the past 20 trading sessions in corn futures and to use this number for placing protective stops. As an aside, this philosophy could be flipped around and be used for entry, which I’ll cover in a future article.
As one last example this average true range methodology could be slightly modified by working with a fraction or multiple of the volatility estimate. Ex: A CTA might want to set their protective stop equal to 150 percent of the average true range for the past “N” trading sessions, (The famous Turtle traders used this methodology by taking the 20-day average true range and then setting their stops equal to 200 percent or 2x this number). The theory is that the fraction or multiple enhances and increases the probability of not being taken out of a valid trade due to market “noise”.
*Historical Volatility – HV’ is the realized volatility of a financial instrument over a given time period. Generally, this measure is calculated by determining the average deviation from the average price of a financial instrument in the given time period.