Robert, a Forum-Twitter subscriber, asked me how I work out ‘normal’ range and volume.
The software I use, Market Analyst [1], calculates the mean and standard deviation of average true range of the bars bounded by two dates. I use Barros Swings to identify the dates.
My historical norm is usually determined by the structure of the Second Higher Timeframe. Since I trade the 18-day swing (monthly trend), my second higher timeframe is the 12-month swing (yearly trend). Figure 1 shows that the yearly trend has a confirmed sideways structure. Hence my calculations will begin from the March 2000 high.
The next thing I do is start a data set from the most recent First Higher Timeframe extreme - for the 18-day trader, this is the 13-week extreme. In this case. I started a set at the March 2009 low. Finally I have a set that defines the trader’s timeframe (me, 18-day) structures that have taken place since the most recent First Higher Timeframe extreme.
Figure 2 shows the various calculations.
- The Green shows the 12-month data (from : March 2000 high). The ATR is 17 to 18 with a standard deviation of 11 to 12.
- The Purple shows the 13-week data (from March 2009 low). The ATR is 16 to 17with a standard deviation of 7 to 8
- The Red shows the 18-day data (from July 7 2009 low). The ATR is 14 with a standard deviation of 6.
- The Brown shows the latest 5-day data (From Feb 5 2010 low). The ATR is 12 with a standard deviation of 5.
Normal for me is mean +1 to mean -1/2 standard deviation. Thus Normal range in the current move is 17 to 10; the Normal Range historically is 30 (18 + 12) to 11 (17 - 6).
Generally I like to see current ranges and data close to their historical data.
I do the same calculations for Normalised Volume.
Figure 2 also shows why I use Normalised Volume rather than Volume. You’ll notice the “U” shape brought about by the expiry of futures contracts. The Volume tends to peak in the week of expiry and then levels off.
Figures 3 shows the Normalised Volume. As you can see, there is no ‘U’ shape. I have added Figure 4 which shows the usual “U” shape without the ATR data.
By the way, a friend of mine sells a software program that creates Normalised Volume off csv data. You can contact him at:
“Kym Haines” <khaines@jordanit.com.au>.
The software is not ‘pretty’; in fact it still has a DOS interface. But it is effective and that’s what counts in my book.
I make the usual disclaimer: I obtain no benefit from recommending the program. I just happen to think it’s great value.
[2]
FIGURE 1 12-M S&P
[3]
FIGURE 2 S&P Mean & Std
[4]
Figure 3 S&P Normalised Volume
[5]
Figure 4 S&P Volume