I was researching various way to compute weighted averages, and came across something seemingly unrelated, called the Simpson's paradox: Simpson's paradox, or the Yule–Simpson effect, is a phenomenon in probability and statistics, in which a trend appears in different groups of data but disappears or reverses when these groups are combined. Wikipedia has a very good article on it: https://en.wikipedia.org/wiki/Simpson's_paradox Here is how it's illustrated: As it turned out, there is a relationship between weighted averages (such as exponential moving averages) and the counter-intuitive effects of the Simpson's paradox. Given the preponderance of the "what is a trend" threads in ET, and the dominance of moving averages in technical analysis, I thought you guys might be interested in this.
I've heard about this in regards to college admission, but how could it apply to trading? Can you give an example?
on this picture data looks like came different distributions. application to trading - regime switching models
In that graph, imagine that the red and blue lines are short-term exponential moving averages of the stock price, indicating an uptrend, and the black line is a long-term exponential moving average of the price, indicating a downtrend. Would you buy, sell, or hold? That's what comes to my mind when I look at this chart.
This contains a good applied illustration using portfolio weightings, reported performance, and the potential to mislead observers, depending on how data is aggregated.
The way I would apply that to trading is simple: 1. do not compare directly (and decide which is better) trading results from different instruments, timeframes, methods etc. 2. from the above: focus on particular instrument, timeframe and optimize your method 3. any in general: there is no perfect way to trade... but that's just me. Your conclusions could be different.
Interesting post. The aspect that is readily applicable to trading is the section under Vector Interpretation and the link for Bayesian Networks. Using Price and Volume as the axis, one has to include the cases of the three other quadrants. The above quadrant is P+, V+. With these two variables, there are three other permutations. Each require their own quadrant in the above Vector Interpretation. Applying the logic of P+ when V- in the same quadrant as P+, V+ demonstrates the suggested Simpson's Paradox. Applying Parallelograms within the context of directed acyclic graph can provide insight from a visual geometric perspective. The links to these concepts are within the quoted above text starting at Simpson's Paradox. If one pic is worth a thousand words, this would be one: http://www.allofthisisforyou.com/gallery-2/homer2009/
This is what I would do: Buy on the first blue dot, sell on the 4th blue dot. Then, buy again on the first red dot and sell again on the 4th red dot.