It’s well known that following a period of contraction, we can expect an explosive move. The question of course is, what measure of contraction do we use?
In this blog, I’ll look at some of the traditional methods as well as the one I like.
Figure 1 shows the contraction methods of which I shall be writing about in this blog.
The first indicator is the average true range. You tailor the look-back period to suit your timeframe. Since I am an 18-day (monthly trend) trader, I have used a look-back of 18 days. The red lines are linear regression bands set 2.0 standard deviations apart. Whenever the ATR reaches the bottom band, I determine that sufficient contraction has occurred to warn of a big move. Right now it is in a neutral phase.
The second indicator is Larry Connors Historical Volatility Ratio (HVR). I use a setting of 5 and 100. A drop below 0.5 suggests a big move to come. Right now it is in a neutral phase.
The third is John Bollinger’s Bandwidth. The red lines represent 2.0 standard deviations for the highs and 1.0 for the low. You’ll notice that the indicator dropped below the lower standard deviation and is now starting to rise. So far we have not had a corresponding expansion in price. Theoretically this is supposed to indicate a forthcoming range expansion. I say theoretically because I have no practical experience with this indicator.
The one I use is to compare the mean and standard deviation between two legs of a move. In the top pane of Figure 1, you see that I have drawn two probability boxes; each box shows the average true range of the bars within the box. In addition they show the standard deviation of the ATR of the bars within the box.
Before I explain how I use this tool, you need to know that because I expect to see a skewed sample, I designate normal to be mean +1 stdev to mean -1/2 stdev (and so on) rather than the mathematical mean +/- 1 stdev, mean +/- 2 stdev etc.
Notice that the second box’s ATR has dropped to mean -1.5 stdev of the first box’s ATR and stdev. This is a severe contraction. In addition, the most current three inside days are all smaller than mean -1.5 ATR of the second’s box’s ranges.
The information all leads to the conclusion I expect to see a large range (mean +2 stdev or greater). The method does not indicate direction, just magnitude.
Figure 1 Contraction Tools
Refer this blog post to a friend or colleague…
