A quick one folks. I have to fly to Hong Kong first thing tomorrow so and I still have a million-and- one things to do.

Tonight, I’ll answer the questions raised about ‘what period ATR to identify mean ranges’. The key principle is to ensure I am comparing similar populations - similar in structure volatility. Figure 1 will best explain what I do.

03-14-2008-fig-1.jpg

FIGURE 1 13-w AUDUSD

Let’s assume I want to assess the current ATR of the AUDUSD and I am trading the 18-day timeframe (monthly trend), I first place the 250-day and 60-day Hart swings on a daily chart. The two approximate the 12-month and 13-week Barros Swings. The moment we place the swings on the chart, we see that there are two clear divisions of volatility. The period commencing 07/25/07 marks the beginning of some large and choppy 13-w swings. Prior to that (from the 250-day low), the swings were gentler and corrections shallower.

Next, I place a Market-Analyst Probability Box on the entire period from 03/09/06 to 02/29/28. The blue Prob Box shows we had an ATR of 81.2 with a standard deviation of 46. The Red Prob Box shows the earlier period had an ATR of 63.4 and a standard deviation of 24.4. Finally, I place a green Prob Box to confirm the current stats: the green box shows the ATR has increased to about 120 with a standard deviation of about 60.

I want to fine down the ATR farther. Normally, I like to see at least 30 bars in the swing; in this case, we only have 28 for the last swing, but it will do as an illustration. Figure 2 shows that the ATR has dropped marginally in the last swing structure. This issues a warning that the market will move into congestion. So I’ll be watching the Primary Sell Zone of 949o to 8412.

03-14-2008-figure-2.jpg

FIGURE 2 13-w AUDUSD

Note that the above is not a full analysis of the AUDUSD; but it does serve to show how I measure the ATR and to what uses I put it.