I re-iterate that the A/D line is positively biased due to the greater statistical occurance of rises over declines and even more so as you move out to longer time frames and thus, is worthless as an indicator. It is devoid of any data input regarding price and volume.
You consistantly mention only 1 circumstance that favors your bullish outlook and that is a scenario where a stock rises 5 points in 1 day and falls back 1 penny over 5 days. True, this will produce a falling A/D line, yet there is not just 1 but 4 scenarios for the A/D. Two are bullish and 2 are bearish.
1) Stock rises 5 points one day, falls back 1 penny for each of 5 days after. Result: Falling A/D, yet stocks rise over time. Complete disconnect between A/D and price.
2) Stock rises 1 point in one day but falls back 0.25 point over 5 days. Result: Falling A/D, and falling price. Complete correlation between A/D and price.
3) Stock falls 1 point over 1 day but rises, 0.25 point over 5 days. Result: Rising A/D with rising price. Complete correlation between A/D and price.
4) Stock falls 1 point over 1 day but then rises 1 penny for 5 days. Result: Rising A/D but falling price. Complete disconnect between A/D and price.
Thus, there are 4 scenarios, 2 bullish and 2 bearish and for each bullish and bearish set, one is correlated between A/D and price and the other shows a complete disconnect. Taking this alone into consideration there are 2 bullish and 2 bearish situations giving each a 50% shot. Breaking them up and thorwing the disconnects into the opposing divergent catagory still makes it 50%. The A/D line is worthless.
That leaves us with the fact that over time, stocks statistically rise 62% of the time and fall back 38% of the time giving favor to the bullish side. Since the A/D only concerns whether stocks rose or declined, odds favor the bullish side. Yet, as we have seen in the above 4 scenarios, one of them will produce a rising A/D yet have falling price.
Using larger time frames does indeed smooth out the data, but since the bullish side is always favored statistically, all you do is smooth out the bearish side from the smaller time frames and the result is a bullish A/D, regardless of what the underlying price does. The further out you go, the more you smooth out which makes the weekly A/D much more of an illusion than the daily.
If you use a form of analysis that favors the positive, the result is biased and thus worthless. The weekly is even more positively biased and thus even more worthless. Furthermore, the A/D line has zero time correlation as it has has peaked and bottomed at various times in no consistant manner with the peaks and bottoms in the stock market making it even more useless in timing and in diagosing the health of the market at any specific time. |
