Therein lies the issue with regard to understanding and profiting from the markets. There are many books, speakers, traders, programs, firms, funds, and markets in which people operate many of which may or may not be acting in the participants best interest. Therefore it is incumbant upon an individual before risking capital to acknowledge this fact and act accordingly.
Just to be clear, only the first paragraph of the preceding post's quote attributed to me was my own text. Everything following that first paragraph is Grob's response. Just a note to avoid any confusion.
Thanks for taking the trouble to do the search and to do the equity curve. I couldn't find the post. Is there a way you can post a way to get to it? You may want to relabel the equity curve so it is consistent with the thing you put up as what you backtested. TIA.
What was the slippage and/or commission that was setup on this system? What type of orders? (Limit, Stop, Market) ... I really don't care about theories or who's right/wrong... From my observation, this test seems like it's not tested right to draw the conclusion you have made.
I think this is just the first one of a family of four equity curves chicken is doing. It has one small slip up, maybe. He may have done it with the wrong size bars (5 min as shown) or he misslabelled it as 5 min bars instead of 30 min bars. When he gets to the second progression, one using pairs of slopes, the backtest gets more complicated. The third progression is done with 30 minute bars on two charts where the two charts are offset by 15 minutes and pairs of slopes are involved from progression two are involved. That is a very long and involved backtest. The fourth progression, the retrace progression, is done using an iteration of the third one. If he posts the source of the post in 2000, then people can go there and see what was posted for the four progressions instead of just the first one which was designed to show a base line beginning point for iterative refinement and how doing refinements really improves the profit margins. So far, we only have an equity curve based on the wrong bar durations and perhaps the wrong starting bar. We will probably not get any further than we are now. None of this is drill oriented, it is just a learning series of four ways to mechanically trade where each one succeeds the former in level of money making. Using 15 bar days is a common beginning approach to take the heat off getting a routine going. Attached is a third progression covering a day... it is not readable for sure but those doing it were looking at phantom bars out 15 to 30 minutes ahead of the market and seeing how the market turned at different points that day. It just conveys that real forming bars do tell you what is going on and from this you can improve your holds and profit taking by stages of improvement. On this day, the end of the day close value projected ahead(crosshatchy red bars) and the real bar (solid orange) were coincident. All of this stuff gets a person to see the market in vector quantities instead of scalars. I am not holding my breath for chicken to post the html to take you to the series of posts on the four progressions nor do I think he will correct the back test to 30 minutes nor will he do the remaining three backtests to demonstrate how iterative refinement of trading bar signals work. This was in another forum unlike ET. Have a laugh on me.
The post is found by doing a google search on groups. Limit the search to author "Jack Hershey". At one time your posts were quite prolific on the usenet groups misc.invest.futures and misc.invest.technical. There is no need to relabel anything as the test of your drill shows the results of your drill as you described it precisely.
thanks. Your illustration is showing 5 min bars. I assume yopu used 30 min bars as you state here. All orders were market at the time of the BO on the bar, no bar data for the trades is useful except to choose the value of the BO. Are you going to do the other three progressions?? Check the chart I posted so you know how to do the pairs of slopes. Could you post the reference as a red thing with a red line under it? People may want to go to where you abstrated the part you did and see what was going on to use the baseline you backtested.
The test was performed on 30 minute bars precisely as described for progression # 1. A thread describing a traders attempt to follow the remaining three progressions is found in here: http://www.elitetrader.com/vb/showthread.php?s=&threadid=56093 It doesn't show an improved result as it was practiced by the thread originator. That leaves two explanations: 1) The practitioner was in error 2) the theory is flawed. I tend to favor explanation #2.