Chaps,
I’m very interested in how many of you are consistently making money from auto trading.
As we all know, PRT has some major short comings when it comes to backtesting and the very limited data for lower time frames.
What are your experiences and thoughts?
Thanks for that Grizzly! 😉 Would be nice to know how long it took you to become profitable and how long you have been profitable.
How many strategies you are running?
What markets?
What are your rules for tweaking strategies?
Do you use any kind of walk forward test/ Monte Carlo etc?
Do you rely solely on PRT or do you use other data for backtesting?
gm74Participant
Average
I feel I have successfully mitigated the shortcomings of the backtester, by limiting the type of systems I employ:
1
So I have entered conditions for when to close trades. These conditions are always based on closing prices/lows/highs.
this then mitigates the shortcomings of backtesting, where it only looks at low,high,open,close. I do not Exit with a stop loss/trailing stop, but rather if the closing price is below the EMA20 (for example).
2
I also flat out all orders each night to ensure that no overnight interest is charged.
3
I also include the spread (or commision) of the product I am running my system on.
I have substantially limited the type of systems I can now run. But at least I know (or I think I know) that there will be no surprises when I run it live.
Now just to design/find that perfect system.
Thanks GM, number 1 is useful to know that you are doing it this way. You’d hope that most people are aware of 2 and 3 already…
I make a little pocket money at the moment. I’ve had systems that made profit but decided to turn them off live until they proved them selves a little more. I focus on fx with specific systems for specific pairs due to difference in volatility and spreads. It’s taken a long time (a few years on and off) but i guess that’s relative?
Hi Cosmic,
Auto trading USDJPY only.
My code is as generic as possible. Very little room for curve fitting.
I have run pseudo walk forward tests in Excel & tweak the code every few weeks. And by tweak i mean add in risk management layers.
I’m testing Nicolas’ code for CAC breakout on different markets at the moment due to the fact that with 200k units and the 15TF you can get 8-9 years of data for increase confidence + the fact of no 0 bars/no fake profits. Prior to this I was using 1min TF to eliminate 0 bars/fake profits on some auto trades that I had made for the major indices but with only 18 months of data, I lost confidence.
I’ve made a separate post in the forum for my values today.
dwgfx, grizzly. Care to share the things that you are runnning?
Hi Cosmic1,
I run more than 10 strategies at the moment, most of them on DAX-CFD, but also on EURUSD and Gold.
And yes – I am profitable.
I use 1 hour or 10 min – charts.
Due to the fact that there is not as much data history I tested the strategies with another software before I traded live.
In 1 hour chart you have about 10 years history with prt – this is not so bad, I think.
Well this is good to hear that people are making money.
Did you test with MT4 with longer data?
Are these strategies ones you have made yourself or the one’s posted in the library here with tweaks etc?
I tested with Tradesignal and data back to 2002
All the strategies are scripted by myself.
Most of them are own ideas, but there are some with adapted ideas of colleagues, friends or partners.
I wasn’t a programmer before but come from market side. The programminng I learned by reading the probuilder-guide or using the construction kit from prt and checking the codes in it.
I think it’s better to have trading experience and ideas and learn to program thoose as to be a brilliant programmer with no trading ideas.
WingParticipant
Veteran
I’ve only been live for a month, but results are so far meeting my expectations. You could say they have paid off the hours I have put in at least. For me it is about finding a robust set of algo trading systems and then slowly scaling up the size. I currently trade higher timeframes (15m-1H), CFDs with lowest possible spreads compared to ATR (some only consider spread), and of course good backtesting and Monte Carlo. The only thing I have struggled with is to actually trust the systems. I’ve taken a few losses from intervening in trading done by systems. I have a lot of systems running, all are intraday and most are low-frequency.
gm74Participant
Average
Do people mind posting or sharing their successful strategies? or is that too personal? My opinion is, sharing is caring, I am not costing you anything by also sharing and implementing your strategy.
I am a novice, and writing a couple strategies a week, but I have yet to find something I solidly trust to run live. I would love to share in successful systems of others.
I will thank you with compliments.
Have a lekker weekend peeps.
I am definitely of the same thinking gm74. I am happy to post anything I find successful with others but understand if others do not share this view point. These strategies do and can take a long time to get to a point where confidence is high. I have lost many hours/days learning and testing various things.
I will post the ITF’s for breakout in the other topic later that I believe will be profitable going forward that I have put some time in to.
For Example:
have a look at the bollinger reversal strategy I have postet in the library.
http://www.prorealcode.com/prorealtime-trading-strategies/bollinger-reversal-strategy/
