Profitable ETF Trading Strategies: 4 ways to measure volatility
Volatility is such an important part of any shorter term trader’s trading strategy that it is worth taking a few moments to consider several different ways that it can be measured and understood. Depending on the typical length of time you plan to hold your positions you may find one or the other of these methods more suitable for you.
The bottom line is that the method you choose should be responsive to significant changes in the time period you favor, and be sensitive enough to give you actionable information. You should be prepared to spend some time trying out different parameter settings until you find the best trade off between smoothness and sensitivity.
By smoothness, I mean how well the parameter settings filter out or smooth over normal noise variation in the data, yet still making it clear that an important condition in the market’s volatility has just changed.
Long term traders can simply use beta, a comparison of the asset’s volatility to that of the market to find out if, in general the asset is more or less noisy than the market. I recommend considering the correlation to the market to see how much in parallel the asset will move when the market moves.
As an example: if you intend to be an intermediate to long term trader you may be well served by considering volatility as defined by “Annualized historical volatility”. This method describes volatility as measured by the standard deviation of price for the lookback period, then annualizing it using normal statistical methods. This gives us a reasonable approximation of the kind of volatility we could see over longer periods.
Intermediate term traders to swing traders (holding from months to weeks) can get good information from simple standard deviation of price over the holding period, without a need to annualize, because you don’t intend to hold that long.
Swing traders need something a little more sensitive and reliable than standard deviation, which loses its authority to describe volatility when sample size is less than 30 periods. A swing trader would have to look at hourly price data for standard deviation to be meaningful and sometimes that data is hard to get or unreliable.
Average True Range (ATR) is probably a better volatility measure for swing traders as it is more sensitive and accurate than standard deviation in shorter time frames. By going a step further and dividing ATR by price you have ATR% which allows you to fairly compare an asset against itself through time or to compare different assets at the same time. Plotting a time series of ATR% gives a much better representation of volatility through time than straight ATR, whose line is skewed by changes in price.
Simple ATR time series actually disguise changes in volatility. If an asset goes up in price but has exactly the same relative volatility, the ATR time series will rise, which will usually be interpreted as increasing volatility when that is absolutely not the case. Choose your graphs wisely.
Depending on your time frame, you should carefully consider the effect of volatility on your strategy and choose a measure that best meets your analysis needs.