|According to Martin Pring in his work "Technical Analysis Explained," a diffusion indicator is an oscillator made of a basket of securities that comprise an index and often measures the part of the market that is heading up. "An example might be the percentage of the 30 stocks comprising the Dow Jones Industrial Average that are above their 30-day moving averages," Pring adds.|
|The statistical diffusion indicator I have developed that is seen in the bottom window of Figure 1 is an oscillator that moves between zero percent (0) and 100% (1.0) and is constructed of the individual Dow Jones Industrial Average (DJIA) securities. The indicator measures the percentage of the Dow stocks whose 50-day linear regression slope indicator is above zero and whose 50-day R-squared indicator is above its critical level. Thus, this statistical diffusion indicator is made up of a basket of stocks that individually are in strong intermediate-term uptrends. The current reading of 0.766 indicates that 77% (rounded up) of the stocks within the DJIA are still in strong uptrends. However, does the combining of these individually strong trending stocks mean that the statistical diffusion indicator is signaling that the overall DJIA is also in a strong uptrend? Logically we might think so, but to answer this question properly we need to apply statistical analysis to the diffusion index.|
|FIGURE 1: DJIA DIFFUSION INDEX, DAILY. The bottom window shows the DJIA diffusion index and a developing head & shoulders pattern. The top window shows the linear regression slope indicator, the second lower window shows that R-squared Indicator, and the third lower window the relative standard error indicator.|
|Graphic provided by: MetaStock.|
|From mid-August 2010 forward it looks like the diffusion index is forming a head & shoulders pattern, so I decided to use a time frame that included the length of this pattern and chose to use 130 days. I then plotted the linear regression trendline with a 130-day time period. I also set the linear regression slope indicator, the R-squared indicator, and the relative standard error indicator to the same time period.|
|The result of this analysis from mid-November 2010 shows that the linear regression slope indicator indicates that the linear regression trendline has been decelerating. Since I know that price deceleration normally occurs just before a trend reversal, I know this indicator is warning me of a possible reversal in trend. I also note that the R-squared indicator is moving downward, informing me that the strength of this trend is weakening. However, the relative standard error indicator is below 0.2, indicating low volatility. |
Two things we should know about volatility when it comes to statistical analysis. One, low volatility suggests that the current trend will continue; and two, volatility normally increases to a high level just before a change in trend occurs. Thus, volatility is the indicator to watch. As long as it remains low, I expect the diffusion index to continue to move upward. However, once volatility starts to increase at a fast pace, I know that the diffusion index is close to a reversal.
|In conclusion, the statistical analysis of the DJIA diffusion index suggests that the DJIA is nearing a possible reversal in trend. However, until volatility starts to increase in the diffusion index, the DJIA will most likely continue higher.|
Click here for more information about our publications!