John Madarasz: Thanks for getting back to me. Back in the eighties I took a look at various moving averages extensively using Technifilter Plus (Dos Based in those days). I also used Dow Jones Market Analyzer Plus as well. Dow Jones was my data provider. I long ago tossed the studies, but as I recall I looked at daily bar moving averages crosses optimized from 1 to 200. Of course, I now have an additional 17 years of data since I ran those studies as I believe I ran them in early 1987. The buy/sell or sell short/cover was triggered by a close above or below the shorter term moving average. The trades were executed at the open of the next day.
As I recall I found no combination worthy of use as a stand alone strategy. It took my old 286 hours and hours to run the back tests as it ran through the data using thousands of combinations.
My recent test of the moving average strategies mentioned in the post I first responded to using TradeStation, bare out my original findings even with the additional 17 years of data. However, I took a look at the data TradeStation is providing and unfortunately find it less than desirable for index strategy back testing. For the NAZ Composite, TradeStation's data from 1984 back only has the Close price inserted into the Open, High, Low, Close fields. For the Dow Jones Industrials, while TradeStation has the High, Low and Close going back to the twenties (most likely due to the original method of calculating those numbers, the theoretical numbers method), from 1994 back TradeStation has the Close for each day inserted in the Open field. So I have no Open numbers to work with going back to 1970.
Bottom line, my calculations using TradeStation are flawed. However, due to my previous testing albeit 17 years ago, I have strong doubts regarding the results Markettell is reporting.
Regards,
LG |
