Trading Systems Leaders in this forum (automated trading systems) are winning more than 3000 pips in a month (30000$ investing one lot every time). Click here to register and get more information
I've had similar results to John so far on EURUSD and GBPUSD here is the picture. I'm still developing...
Looks like it really depends on when you start this thing. Started at the right time you really get going on some good trades. Started at the wrong time it ends trying to keep itself out of a whole. Guess I'll have to give it a little better intelligence on choosing its entries.
Interesting to note that even the backtest on this using the current data shows the hit, we'll see if it recovers as I'll be continuing the forward test.
There is an obvious weakness in ranging markets...I'm going to be abandoning the newsfilter, as I believe there is money to be made all throughout the day.
Here are some of my future developments...let me know any of your ideas.
- Use better indicators
- Use different take profit targets during active times and less active times (auto adjusts)
- Double down to increase gains instead of just to recoop losses
- Program will internally keep track of gains and losses and adjust risk based on successes
- Program will automatically adjust for sweet spots, if target is unrealistic (above a all time high or above a logical support/resistance it will adjust it)
Still working on this...I'll likely be starting a new thread soon as the deviation from 10points3 is substantial at this point. I'll keep you posted.
If you "double down" on both wins and losses, how do you know when to reset?
In strong trends price often blows through support and resistance, especially on a daily basis. It might be good to give special attention to weekly or monthly S/R levels.
Any further thoughts to operating differently when strong trend versus ranging market is encountered? And what indicator would be used to determine the difference.
Best of luck to you. Based upon your prior work I have every confidence you will give this effort a good run. All we can do is keep trying things.
Here are my results with the old version eur/usd had no losses,
there is some losses on this chart it it is from the usd/chf wich I am going to stop testing after tommow.
once again disreguard the gold trade
If you "double down" on both wins and losses, how do you know when to reset?
In strong trends price often blows through support and resistance, especially on a daily basis. It might be good to give special attention to weekly or monthly S/R levels.
Any further thoughts to operating differently when strong trend versus ranging market is encountered? And what indicator would be used to determine the difference.
Best of luck to you. Based upon your prior work I have every confidence you will give this effort a good run. All we can do is keep trying things.
saintmo
I'm thinking the EA will keep an internal win/loss count, if the win/loss count reaches a certain level the EA can choose to go in at double. Just an initial idea...still brainstorming.
Right now my thoughts on operating differently during strong trends versus ranging markets is to have a dynamic take profit. It may take profit of 3 during slow times and 8 during active times. This will not help avoid major trend shifts but it will probably lock in better positions during slow hours.
I'm also experimenting with creating some of my own indicators for this EA, possibly and EA on RSI or CCI, or maybe some new stuff I'll describe later. I've been keeping track of baba_masters indicator development because I think it will have some great value My new indicators !
XPMA and Turbo_JRSX are also on my to-test list...other suggestions would be appreciated. Bascially I need aggressive stuff that can detect fast. I'm going to try having the EA look at smaller time frames as well.
I've replaced the RSI indicator on the latest version with the JRSX and also removed the time restriction function and it seems to test smoother on backtest. I will report at the end of next week with forward test results. A very big THANK YOU to neta1o for his contributions.
Here is my redesigned version of neta1o. my version tries to use 4H and 1H timeframe positive divergences of TURBO-JMA. If 4h TURBO-JMA and 1H TURBO-JMA have the same direction then open position.
I think that maxtrade set to 3 is ok. So i resigned from changing this parameter. Another important thing: i belive that in order to balance loosing money there should be adjusted in more precious way THE LOTS PROGRESSION. My expierience indicates that the biggest part of money should be engaged into the most volatile part of movement during the trend. THE TREND IS LIKE A CORRIDOR WHICH HAS BOTTOM, TOP AND THE MIDDLE PART. So if we have 3 maxtrades we use 7*0.1 in order: 0.1 0.2 0.4. The most volatile part of TREND CORRIDOR is the MIDDLE PART. So i set progression: 0.1 0.4 0.2. This gives more profit, look at the raports, the most transactions are made with the middle lots.
I have made many test on different broker with different quality modellnig. Im really depressed and i feel very bad the even data from alpari with 90% quality modelling gives creazy results. I noticed that it is almost impossible to test any EA !!! All kinds of data are crazy.
My observations indicates that the most successful in not loosing money is not trading. So all EAs must treat equally not trading with selling and buying! Even in the strongest trends we can see backmoves of prices, so we never know if it is the beggining of trend reversal or good occasion to enter.
I propose to test systems in FEBRUARY 2007, that month was very creazy for GBPUSD. In that month you have the mixture of 2 kinds of market behaviour: waves and breakouts. 10pointsXXX is kind of wave(trend) following system, so it has many problems during breakouts. So i think that breakouts are very big problems, i think that there are needed very special exepctions for system to cut loosing money in breakouts.
Trying to adjust EA to ranging and trending market is very lagged. I think that more usable conceptions is to have eg. 4 lots for trending and 3 for ranging or make more elastic distance changing between lots. In stead of 15 pips there could be calculations of levels from CCi eg. 0.1 lot over 0.00, another 0.1 over 100, another 0.1 over 200.
I have done some initial testing and this ea produces less of a negative for 2007 than the other versions (for eur/usd 1HR TF, all of 2007) . Using lot size of .1 for 2007 from Jan 01 to current it is down only a couple thousand dollars.
But for period July 2003 through current (about 4 years) it produced $179,000. So overall this may be a winner. Will do more testing. If it can produce very little drawdown during the difficult periods and good profits during the good periods this could be a real winner.
My testing produced similar results but the most recent trend change caused a big hit. Although the EA is digging itself out of the drop right now. Here is my most current results.
This thing isn't there yet because live is much different then backtesting for various reasons. But I have some other developments that could get this thing performing even better.
Note: This EA was not optimized for GBPUSD, only for EURUSD, I ran it on GBPUSD anyway just for testing quality. Note these results are forward tests.
Linear Mode
