[Todd Morgan]

Hi Eric,

Fantastic project. I'll be quite interested to follow along with it. If I can help in any way please let me know. I could contribute data as needed.

Interesting stuff from the other data set you referenced. I've only been back at this MTP RT for 5 weeks now after a 21 month layoff from using MTP RT. So at this point I've had only 51 trades. But the results are excellent.

Yet, I too seem to be having better/more consistent results with 3min trades than with 5mins trades(same as it was when I used MTP RT before). I've even considered eliminating the 5min setups all together. I may keep the 5min charts on my screen to just to be able to look at the 5 min ATR stop once a 3min is triggered. Time will tell.

I've also brought the ETF equivalents of the E-mini's on to my screen for an added bit of diversity. They are the SPY, QQQQ, DIA, IWM. You’d be amazed at the differing triggers and results from the emini’s and their ETF equivalents. One would logically think that they would exactly the same since they’re just differing tradables of the same underlying. But the triggers and the trade result are quite different at times. It’s quite interesting.

And I've also included the most liquid sector ETF's of OIH, SMH, RTH. I'm getting very promising results from these sector ETF's on a longer term basis...60mins+.

I'm bringing these ETF's on board because I have a super low "cost of capital" and "market maker margin" requirements so it's virtually no difference for me to trade the e-mini's or the EFT Security/Share equivalents. Alas, you may or may not be interested in that data as well.

Best Regards,
Todd

Originally Posted by ericd2281

Todd et. all,

I am starting to look at and analyze/test the effects of common objective MTP filters on the overall expectancy of the results over a long period of time. I expect the experiment to last a year or so with usable results coming as early as 4-6 months.

I agree with you in that each event (trade) is independent so one could not predict the outcome, but a trader can look at past results to get a good idea of what would most likely work out better in the future. For example, I looked at the average expectancy (R-Multiple) per trade on results from 7.26.2004 to 7.22.2004 attached in another post on the board here and found the following...

3 Min. ES, NQ, YM E-Mini's Average R-Multiple/Trade: 0.494 (327 Trades)
3 Min. ES, NQ, YM E-Mini's Cumulative R-Multiples: 161.75 (99% of Total)
3 Min. ES, NQ, YM E-Mini Total Trades: 327 Trades (70% of Total Trades)

5 Min. ES, NQ, YM E-Mini's Average R-Multiple/Trade: 0.009 (138 Trades)
5 Min. ES, NQ, YM E-Mini's Cumulative R-Multiples: 1.25 (1% of Total)
5 Min. ES, NQ, YM E-Mini Total Trades: 138 Trades (30% of Total Trades)

Because of this large discrepancy I am currently only trading the 3 Minute Signals. Granted there will surely be some 5 minute trades that would end up being profitable that I did not look at based on the past history but I am ok with that since I know that over the long run I should get somewhere around 99% of the expectancy with 30% less data entry, scrambling to get orders entered, commissions, slippage etc. Assuming a $20k account and only 1 contract traded on the 138 5 minute signals (I fully understand this is incorrect position sizing and totally wrong, but it is an intentionally low assumption to make a point) and $5/side commission, this would reduce your return for the year by $1,380 or roughly 6.9 percentage points. Assuming a $20k account and an average of ~7 contracts (taken from Matt's current RT results here- http://www.mtptrader.com/showthread.php?p=5055#post5055 ) traded on the 138 5 minute signals and $5/side commission, this would reduce your return for the year by $10,253 or 51 percentage points. Certainly a trader could choose to take all of the signals, shown above but I do not fully understand why they would take them if they understand the above results.

It will surely be interesting to see how things work out in the analysis I will be conducting to see if any 'filters' work in the same manner as the data shows above.

Also, any questions/comments are appreciated especially since it is difficult for me to explain this in a web bulletin board setting.

Regards,
Eric