Mo ma's used in the conventional way and as a stand alone product will fail in most instances. The numbers support that.
However used as trend analysis tools they certainly have value. My argument has always been that an experienced TA will know what a MA would do without the need to plot it. That said I use MA's always but as an instrument to find averages in time. I also know other techs that have uses for them outside the norm that are effective.
As for patterns, absolutely agree. Patterns are all about psychology, they are a tangible, measurable event.
I'm still only just into Kate's stuff but the early argument was that during a full moon we become more depressed etc and from that they found a correlation between full moons and low events. Makes sense in theory however I have found many instances of new moon highs which goes against the grain. One point that was raised addresses the level of sunshine and its effect, which I found interesting. But then weather isn't universal across the planet at any given time, yet markets are. They generally all have some connection. I would like to see data that connects astro cycles with weather events such as rainfall and then apply that to markets. The problem there though is markets are global, weather events are not.
The chart below is interesting. A few quick stats on the three years data that was tested.
- within four bars of each time point a pivot occurred that preceeded a rally greater than: 0 points - 8% 50 points - 33% 100 points - 43% 150 points - 8% 400 points - 8%
Strike percentage of 92%.
These results are impressive by any standard. I mentioned earlier the horizontal line theory; that a random plot can often have significance. This exercise is the same theory applied to time. I wanted a random number, so I approached my overgrown teenage son with smelly socks on my lounge while he watched spicks and specs to give me a random number between 50 and 120. "huh whaa?" was his first response (this kid's top 2% academically by the way - god help us) after the initial shock that someone had interrupted him, he mumbled the number 73. The chart below represents the performance of a smelly teenager about to lose his laptop for a week for having his feet on my lounge can offer.
I have to admit I've been here before, we've done this test numerous times on multiple data sets and on both X and Y axis and often found profitable results from random numbers with no importance. Feel free to test the theory yourselves. The reason I need the 'why' behind an action is that I need to know the activity of participants will support my action for tangible reasons rather than some smelly teenager with a random number.
(20min delay)