Give your opinion about these results on back test | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Display your trades on your chart? Give a look at this 1 reply

Opinion about these Results 0 replies

share your opinions about these results? 11 replies

please give me your opinion about this chart 4 replies

Why EA test is forward test better than back test?? 1 reply

Attachments: Give your opinion about these results on back test
Exit Attachments

Give your opinion about these results on back test

  • #1
  • First Post: Jun 27, 2007 6:28pm Jun 27, 2007 6:28pm
Hi
Imagine you have a system to run in GBPUSD with 1M time frame, 24h/24h, and it produce in last month (28/05 to 28/06), 18 trades all positive J, back tested in 3 different brokers.
-Do think is possible?
-What is your classification of it?
-Do you think it can be a loser system in any conditions, different from last month?
I join for analysis, a back test report; I will appreciate constructive opinions, analysis, etc. about results, not about the system. THK
Attached File(s)
File Type: zip M E V A.zip   11 KB | 600 downloads
OldSchool
  • #2
  • Jun 27, 2007 7:13pm Jun 27, 2007 7:13pm
the quality of the results depends on the system.

However in my experience...mechanical systems that operate on any timeframe less than 15 minutes are just trading noise and the results will not be consistent month to month.

The backtesting on different brokers means nothing. If you really want to test the stability of the system, test it on different currencies AND different timeframes....however testing on different timeframes is just for hardcore, perfectionist system developers.

And don't test it over anything less than four months per currency.
  • #3
  • Jun 27, 2007 7:18pm Jun 27, 2007 7:18pm
Mtav:

The modeling quality is only 25%. I've been told that at least 90% needs to be obtained to get 'worthy' data. I've put a few EA's together on lower timeframes and had some really crazy 1000% return results. Demo them and they loose...It is hard, down right difficult, to put together a short timeframe EA. You may want to PM stockwet or aprasai. They've done a lot of EA work....
  • #4
  • Jun 27, 2007 7:22pm Jun 27, 2007 7:22pm
Quoting Palmer
Mtav:

The modeling quality is only 25%. I've been told that at least 90% needs to be obtained to get 'worthy' data. I've put a few EA's together on lower timeframes and had some really crazy 1000% return results. Demo them and they loose...It is hard, down right difficult, to put together a short timeframe EA. You may want to PM stockwet or aprasai. They've done a lot of EA work....
Yes...90% is mandatory.
  • #5
  • Jun 28, 2007 3:21am Jun 28, 2007 3:21am
Hi there
Many tanks for your clear opinion aicccia and Palmer J
It supposes I’m aware of the questions you say, thanks.
I have my system run for past month in a standard account in with 1H time frame, at same time i have open a chart in 1M time frame, witch have download every tick from real account, so the data is very accurate, at least for my broker, i guess.
I will open now one mini account with $1000, to see if it guarantees consistent results as i get for 1H.
The system is run for get long runs, and in 1M, for me we can get better entry and exit points then in 1H, this is the point I’m study last month.
Many thanks in anticipation, for opinions.
OldSchool
  • #6
  • Jun 28, 2007 4:46am Jun 28, 2007 4:46am
Quoting Palmer
Mtav:

The modeling quality is only 25%. I've been told that at least 90% needs to be obtained to get 'worthy' data. I've put a few EA's together on lower timeframes and had some really crazy 1000% return results. Demo them and they loose...It is hard, down right difficult, to put together a short timeframe EA. You may want to PM stockwet or aprasai. They've done a lot of EA work....
Modeling quality of atleast 90% is great for all other timeframes but for 1M timeframe you cannot get more than 25%. 25% in 1M timeframe is 90% for 5, 15, 30, or 1 hour timeframe.
  • #7
  • Jun 28, 2007 8:43am Jun 28, 2007 8:43am
Quoting WhiteSand
Modeling quality of atleast 90% is great for all other timeframes but for 1M timeframe you cannot get more than 25%. 25% in 1M timeframe is 90% for 5, 15, 30, or 1 hour timeframe.
Hi
same opinion here! thk
OldSchool
  • #8
  • Jun 28, 2007 9:14am Jun 28, 2007 9:14am
Hi
That’s it, modeling quality is more simple then everybody talk about, for H1 modeling quality of 90% only means it use time frame for example 30m! as for 1M there is no small time frame. J, it have 25% quality.
Resume to have the best modeling quality in 1H, you must use 1M time frame! For data in 1H modeling.
This is from MT4 help:

· “Modeling quality — the quality of ticks modeled during testing, in per cents. Modeling is shown elementarily as a band in the next line of the report. The band can be of one of three colors:
1. Gray — this part of available data was not involved into testing. Gray color can appear if there was a date range specified in the testing settings;
2. Red — modeling was not performed in this space because of missing available data of a less timeframe. At that, only data of the timeframe selected in the testing settings were used;
3. Green — modeling was performed in this space. And the brighter is the color, the higher modeling quality was. For example, at testing on the H1 period, the deep-green band means that data of M30 period were used for testing, and the brightest one does that M1 data were used;
Attention: If the fastest method ("Open prices only") was selected in the testing settings, the entire band will be red. At that, "n/a" (no modeling was performed) will be written in the "Modeling quality" field;
“
OldSchool
  • #9
  • Jun 29, 2007 9:21am Jun 29, 2007 9:21am
Hi there
I honestly recommend everybody who have EA to run, to test them in 1M time frame.
I discover yesterday during release of USA interest rate statement, that my broker don’t change spread in big news release, what I know, but they change stop levels!, normally it is 10pips, and during big news it comes to 31pips, as I work with 22pips stop loss, what happens is that my system don’t produce a series of orders in that timesJ. #invalid stops.
As consequence I have made some adjustments, and surprise J system produces 25 trade, instead of 18, and surprise of surprises all positive, and today he produces more 4 trades all positive J
I join results (until yesterday, what I can back test), and appreciate commentaries, I must confess I’m looking for a positive ratio of 10/2, and this month, that I have some more time for forex, is completely exceed my expectations.
Good Weekend, and good trades for all.
Attached File(s)
File Type: zip meva.zip   14 KB | 541 downloads
OldSchool
  • #10
  • Jul 8, 2007 10:38am Jul 8, 2007 10:38am
Hi there
I decide to put here one update for those who wont to check
I must confess in 1M time frame the thinks are harder but profitable! Why?
In my opinion is simple 1M time frame is very rapid to test your strategies, if you can define one that make strategy trades, running in 24h/24n and system is not a loser, is because your strategies is good and it will work for all time frames!
For example if you have a strategies for 1H, I’m sure that it could be a very good opportunities but when is better to enter? One open next candle? In next minute? Ten minutes? 30minutes? ??? I sincerely recommend find it in 1M, you will get better entries and better exits.
For example in last day my system don’t get no entries J, for a 1M system its fantastic, with price action that have been in GBPUSD, but in resume all as the start! What’s your opinion?
I have pass last days study/optimize my system of 1H, to is original idea of 1M and I’m very pleased with results! And MT4 is not perfect but I have to confess is a fantastic tool for traders.

PS: I have made some adjustments in Money management, and in entry points of my system, so results are little different (for better in my side), I diced it because of elevate ratio of winners/losers.
Attached File(s)
File Type: zip M E V A.zip   14 KB | 529 downloads
OldSchool
  • #11
  • Jul 8, 2007 11:35am Jul 8, 2007 11:35am
Can u give us your details on this system...I would like to try this...
  • #12
  • Last Post: Jul 18, 2007 2:36pm Jul 18, 2007 2:36pm
Quoting alsup1955
Can u give us your details on this system...I would like to try this...

Hi there
Sorry i think FF forum team had remove this thread, i only notice today that they have move it to here.
alsup1955, sorry for now I can’t give /share more details, it will come up in a professional way latter this year, then I will have all pleasure to give all details, but take note I will never sell it, I wont it to come up in a innovation way (professional).
For those who interested in check how it goes I join one updated, and i can ensure, the performance is very similar (only one trade is different) of MEVA for the real account where I have it run, exactly with same conditions as the back test, the only differences are in 2/3 pips of few open/close trades, because of requote from my broker.

GOOD TRADES FOR U ALL.
Attached File(s)
File Type: zip MEVA.zip   18 KB | 612 downloads
OldSchool
0 traders viewing now