There has been a ton of talk in the markets about the lack of volatility lately. Bill over at NO DooDahs! put up an interesting study yesterday, that got me thinking that I should revisit some of my own volatility research from a while back to provide a different perspective.

Average True Range (ATR) is an indicator that technical analyst use to measure period to period volatility in the form of the range over which markets trade – basically meaning the high to the low (plus any opening gap if there was one). The advantage of ATR over measures that are based on closing prices is that period to period changes (closes) do not necessarily give the full picture. After all, I know I’ve seen plenty of times when markets have run up a lot and/or down a lot, but ended the day essentially flat.

When I wrote an introduction to ATR article for Trade2Win about a year ago, I introduced the idea of using a normalized version of the indicator to really make the best use of it. We would expect price ranges to vary with the current market price. By that I mean a 1% move on a market at 100 is 1, while a similar move on a market at 1000 is 10. Thus, we would expect ATR to be higher in general terms for a higher priced market. If, however, we normalize the ATR reading by taking it as a percentage of current price, we can evaluate volatility over time without having a bias imparted by changing prices.

Here’s an example:

What you see above is the monthly S&P 500 chart (from the futures using continuous contract prices). It goes all the way back to the mid-1980s. If you look very closely, you can make out the Crash of 1987. Below prices I have plotted normalized ATR and the standard ATR. Notice how flat and smooth the lower plot of standard ATR is. There is a tiny little blip in 1987, then flat, then a steady rise through the late 1990s and in to the early part of this decade. ATR rose, as one would expect, with prices. It then dropped when prices fell, though interestingly it has not come back as the index has moved back up toward the highs.

Now compare that to normalized ATR. The Crash is much more obviously presented because normalize ATR captures the relatively violence of that time. Notice also that normalized ATR didn’t peak out until later because it also caught the relatively volatility that occured during the stock market sell-off.

Now let’s zoom in an take a closer look with the weekly chart.

Here again, we can really see how volatile things were during the bear market of the early 2000s. After that, the volatility dried up completely. It has stayed low during the whole of the recovering going back to 2003/2004.

And we can even see the pattern in the daily chart.

Again, the spikes in ATR have all come thanks to stock market declines. Whenever the market has been trending higher, things have been pretty narrowly traded.

This is the sort of thing you will see in most markets. The declines are much sharper and more violent than the rises. It’s panic selling, really. Have you ever heard of panic buying? Traders don’t normally act that way.

What we see on the monthly chart above is one of the reasons I’ve been a stock market bull for quite a while now. The normalized ATR is down around where it was in the middle 1990s before the market really took off. Until the readings start getting much higher than that, I’m not going to have any real worries about the long-term direction of prices.

Struggling with support & resistance and knowing what the key market levels are? Check out the Price Distribution Analysis methods I use.

Bookmark to:

Volatility Pointing at More Stock Market Upside
Modeling the Stock Current Market on 1987
Looking at Volatility Across Markets
Revisiting Average True Range (ATR)
Watching for a Market Explosion or Implosion