Interesting! Now my own opinion would be that if "they" were only talking about the 50 and 200 day moving averages, then “they” can't possibly be very serious about technical analysis. I've explained how I arrived at the 84 day MA several times on other boards here at iHub but, I don't mind going into it again because I feel it's important. I never understood the reasoning behind the 50 and 200 day moving averages and it has never been explained to my satisfaction exactly how anyone arrived that those numbers to begin with. That is to say; how do we know those number are valid in the first place? My own research on the subject concluded in the assumption that someone (perhaps Joseph Granville or Richard Fabian) arbitrarily started using those averages and perhaps taught it to a few others who taught it to still others. In the MARKET TECHNICIANS ASSOCIATION JOURNAL (Issue 11 - May 1981) on page 53 we read: "There has been a tendency among market technicians to use the same time frames in a variety of models, i.e. , to use a 200-day moving average or a 13-week momentum for every stock in the portfolio. This consistency ignores the fact that different markets have different cyclical characteristics." I even disagree with that philosophy to some extent. I believe that you need to draw from science and physics in particular to identify the primary cycles as well as laws that affect all of life and investing in particular and then see if we can arrive at one set of indices that seems to work for the most investments. I do not believe that the 50 or 200 day moving averages that are used so widely today were specifically designed and tested for validity. I use the 50 and 200 day moving averages on my charts because the rest of the world seems to pay attention to them and makes decisions based upon them and I want to know what the rest of the world (in terms of the investment community) is thinking. I do not however, feel that those averages have any particular validity other than that of an accepted standard. So, several years ago while doing my own research into why anyone in their right mind would use 14 days (or 2 and 4/5 weeks) as the foundation for calculating the Relative Strength Index (RSI) or Stochastics when that number did not seem coincide mathematically with any of the other indices, I began thinking about moving averages as well. I remembered reading somewhere that the number 14 was used for the RSI because 14 days is one half the lunar cycle. But, that logic seemed ludicrous to me because the price charts were not based upon calendar days. Price charts are based upon trading days. So, I decided to figure out how many trading days there are in half of the lunar cycle. That was a problem in itself because the number of trading days is not the same every year especially when you take leap year into consideration. So, I started with the actual lunar cycle which has more than 12 full moons in a year (by-the-way) and then took the average number of trading days over a 4 year period to account for leap year and also averaged the lunar cycle over a four year period so that I was working with the same time frame and I determined that to match the Stochastics and RSI to the lunar cycle in terms of trading days, that I had to use 20.42 for a full lunar cycle and 10.21 for a half lunar cycle. Then I set up a model consisting of all the stocks I had traded in the previous 4-years and I ran back testing with 14 base Stochastics vs 20.42 base Stochastics and 14 base RSI vs 10.21 bass RSI and I was amazed at how much more accurately the timing was with actual tops and bottom vs actual peaks in the indices tested. That did not solve the moving average problem however, so I invented the 84.33 day moving average which approximates 1/3 of a year in terms of trading days (everything happens in threes theory) and I back tested that. I was totally amazed at how much more accurately the 84.33 day exponential moving average was when compared to the 50 day simple moving average for identifying support and resistance levels. For 10 and 20 day moving averages which are very close to being correct already, I now use 10.21 and 20.42 exponential moving averages instead. I encourage you not to take my word for it but, to take these numbers and do your own back testing and see what kind of results you get. In my opinion, it will make a subtle yet, extremely more profitable difference in your trading results over a four year period.
Market Data 
Markets 
AI
