How do I know how much risk I am taking? In one market and in many markets
Upload your EA as mql5 code (because this forum is for Metatrader 5) and members will tell about
The code is all in Java and its two years worth of code (a lot). Plus I can't release it (because of the company that is funding my PhD).
I can see how that could be a problem when trying to evaluate what I've done, but this is all I can give sorry. The equity curve (in accumulated pips) is all 'out of sample' trading so my system didn't see this data when it was created, that might help.
I hope I haven't misplaced my question in the wrong forum, I thought a questions about risk and my automated trading system's performance would fit into mql5's 'General' section.
What do you think of the below? Any insights/thoughts/criticisms?
Your Short position Win Rate is about as good as a coin toss, your long position win rate would probably be the same if you had more data, perhaps 100 times more data, try your backtest again but over 3 or 4 years across maybe 6 symbols . . .
Thanks, yeah I need to trade the market with MetaTrader for a longer period of time and in a couple of different markets. The couple of days I traded was manually done every hour, I was slow getting trades in, I missed some opportunities and I didn't really bother on Friday. So I decided on Thursday to get my Java application to make trades in MetaTrader by itself without my intervention. This will solve my inexperienced trading errors and also get me in trades whilst I'm asleep.
I don't think I can do a backtest with MetaTrader because its Java based (I should really investigate it further, perhaps implement backtesting in my Java program if I'm unable to do it in MetaTrader)
However, the equity curve at the bottom of my post is all out of sample data (its sort of like a backtest but really its my Java program's performance on 7249 hours worth of trading without any knowledge of the data beforehand. So it was never given the 7249 hours of EURUSDpro data to create itself).
7249 hours is about 1.16 years worth of data and that equity curve is what I would of gotten if I traded with my Java application.
What do you think about the equity curve at the bottom? What do you think about the risk questions at the start?
- www.mql5.com
Thanks, yeah I need to trade the market with MetaTrader for a longer period of time and in a couple of different markets. The couple of days I traded was manually done every hour, I was slow getting trades in, I missed some opportunities and I didn't really bother on Friday. So I decided on Thursday to get my Java application to make trades in MetaTrader by itself without my intervention. This will solve my inexperienced trading errors and also get me in trades whilst I'm asleep.
I don't think I can do a backtest with MetaTrader because its Java based (I should really investigate it further, perhaps implement backtesting in my Java program if I'm unable to do it in MetaTrader)
However, the equity curve at the bottom of my post is all out of sample data (its sort of like a backtest but really its my Java program's performance on 7249 hours worth of trading without any knowledge of the data beforehand. So it was never given the 7249 hours of EURUSDpro data to create itself).
7249 hours is about 1.16 years worth of data and that equity curve is what I would of gotten if I traded with my Java application.
What do you think about the equity curve at the bottom? What do you think about the risk questions at the start?
Post your data from your Java "backtest" as you get from MT4/5 . . . including average win, average MAE and actual Win Rate.
You can risk whatever you like, that is upto your risk tolerance, if you have 3 EA running and each allows one trade at a time and you risk 1% per trade then your total risk is 3% . . . or your EA could look at the risk on any other open trades and limit itself to a basket global risk %age . . . there are many ways to do it, the way you end up doing it is part of your strategy and is up to you. Poor risk control won't make a winning strategy lose it will just make it grow slower . . . just keep your global risk beelow 50%.
Post your data from your Java "backtest" as you get from MT4/5 . . . including average win, average MAE and actual Win Rate.
I'm not able to get a backtest from MetaTrader 4 or 5 because my trading system is all in Java but I should investigate it further. But I've knock up a figure from the equity curve if this helps,
Profit/Loss ratio: 1.231864
The trading system always buy or sells every hour and holds a buy or sell trade usually for a few hours, so I dont think the 'Win rate' / average win can be extracted from the equity curve. I don't know what 'average MAE' is sorry. But i think I'm going to need to get these performance figures
You can risk whatever you like, that is upto your risk tolerance, if you have 3 EA running and each allows one trade at a time and you risk 1% per trade then your total risk is 3% . . . or your EA could look at the risk on any other open trades and limit itself to a basket global risk %age . . . there are many ways to do it, the way you end up doing it is part of your strategy and is up to you. Poor risk control won't make a winning strategy lose it will just make it grow slower . . . just keep your global risk beelow 50%.
Thanks :). That helped me :)
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Hey, I'm a bit confused about what size lot I should be trading.
I've heard that I should be risking 5%-2% of my account. So I decided that i'd be willing to risk 3-4% of my bankroll. So if I start with £3,000, I'm willing to lose £90-£120 on a trade.
Now then this is where it gets specific, for each trade I do, I set my stop loss 500 pips away (just to be safe). Each pip is worth 0.22p at lot size 0.30 therefore 500 pips = £110 which falls between my 3-4% risk level.
Is there anything flawed with the logic above? Most importantly, is the idea that my risk is set by my stop loss? and is 3-4% too much or too little?
I know trading in different markets can reduce the overall risk. But do I buy/sell a lot size of 0.1 in 3 different markets or do I buy/sell lot size of 0.3 in each of the 3 markets?
Based upon what I achieved in about 4 days of trading on a demo account, trading on the EURUSDpro on a hourly basis (Day 5 I didn't really bother because I decided that I'm going to automate all of this through a script (I'm almost done implementing it :D ). You can see I introduced Take profit levels as well which I wont be doing in the future). I know 4 days isn't really enough evidence so I've included a backtest at the bottom. Also I missed £44 on Thursday because I drove to work :(.
What do you think of the below? Any insights/thoughts/criticisms?
EDITED: The week of practical trading in MetaTrader isn't enough data to make any conclusions but feel free to comment of the equity curve at the bottom.
Theoretically on the EURUSDpro, over 52-60 weeks of trading I should be able to achieve this equity curve (this includes my 500 pip stoploss, trading costs and the loss obtained from the bid-ask spread).
EDITED: The equity curve is out of sample performance (like a back test but the trading system was not created/optimised on that data)
The Y axis is the accumulated number of pips (0.58 is about 58,000 EURUSDpro pips).
The X axis is trading time in hours (it goes up to 7247 hours and ignore the first hours where trades didn't occur).
Thank you in advance :).