Yeah, I asked myself the same question before deciding to learn MQL in order to program Expert Advisors. Do Forex robots really work?
Well, I guess the ones selling power points based systems will say they don't work. But think about it,
simple logic. If someone trades based on a system, with specific concrete rules, including money management, position sizing, entry an exit criteria, well defined rules for stop losses and target points, etc, and that person makes profitable trades more often than not, why then cant a robot implementing all those same rules be profitable?
Additionally, if someone has an idea that considers to be a good trading idea, isn't it useful to automate it in order to backtest it with already available data and see if it works? Saving you months of entering trades manually, day after day, and keeping track of the results on an excel spread sheet just to realize after 6 months of wasted time that "no", the idea doesn't work.
In my opinion, expert advisors are useful. For validating new ideas quickly with some degree of accuracy, and for, yes, automating someone's profitable manual system. Many of them are scams? Yes, sure they are. And the reasons are among the following:
1- Selling Expert Advisors to the public that win 40% - 50% of the time (which is the real nature of real profitable methods) wouldn't work for sellers, as the vast majority of buyers would ask for their money back seeing an expert not winning 9 plays out of ten. This leads us to reason number two.
2- Given that the only way to make buyers happy is to give them a substantial greater amount of winners, then EA sellers, focus on strategies with very small TargetPoints and rather large StopLosses. This trick alone, just based on random probabilities, makes any EA have more winners than losers. Which doesn't mean that the EA will be profitable anyways, as one loser will often wipe out the profit achieved by several winning trades.
3- Given that most EAs are scalpers (so that they can be sold to the naive public), the simulations are not valid at all. Scalpers depend on tick by tick data, which is not stored in historical data sets. Historical data-sets go at most to the opening and closing price of the minute bar. This makes back-testing totally unreliable, and prone to interpolation errors.
4- Unsound tactics present in most commercial EAs. Tactics such as Martingales, D'Alembert, Grid trading, are proven unsound tactics, Time bombs that will inevitably explode at some point.
5- Lack of flexibility in the trades management. For example aspects such as fixed TP in pips, fixed Stop Losses distances, or fixed Position sizes will result in poor money management and the inability of the Expert to adjust itself to different volatility environments.
Finally a piece of advice.
If an EA seller doesn't provide you with at least 6 - 8 years of back-test evidence. Plus at least 6 months of live test with Investor (read-only) access to their account through a third party source such as myfxbook.com then it means they are probably hiding something or they don't trust their own expert to trade their money.
I would also add as an advice, never never never trust any back-test performed on a time-frame lower than H1.
Also, never trust affiliate reviews. Learn to spot affiliate reviews and don't trust them.
Automated trading is possible, not only in Forex, but also in commodities, stocks, bonds etc. In fact automated trading is common nowadays and represents the majority of the trading volumes. Get away from over hyped EAs, analyze the trades history thoroughly, evaluate the "evidence" the seller is trying to sell you and don't ever buy anything that doesn't meet your standards. The equity curve alone, its shape tells you a ton of things about the nature of the robot being sold to you. Read this article written by me about the meaning of different forms of equity curve growth, Tips for reviewing Expert Advisors I believe it is a crucial piece.
I would say the best way to trade with an automated software is using a strategy designed by you and hopefully programmed by yourself or someone you hire in case you don't have the programming skills. The evidence is out there, for example in the stats published every month on the Barclay's systematic traders index, only to name a serious source.
Related Articles
Tips for reviewing Expert Advisors
LT Trend Sniper - a Forex strategy that works
The Turtles Trading system automated (Expert Advisor for download)
Well, I guess the ones selling power points based systems will say they don't work. But think about it,
simple logic. If someone trades based on a system, with specific concrete rules, including money management, position sizing, entry an exit criteria, well defined rules for stop losses and target points, etc, and that person makes profitable trades more often than not, why then cant a robot implementing all those same rules be profitable?
Additionally, if someone has an idea that considers to be a good trading idea, isn't it useful to automate it in order to backtest it with already available data and see if it works? Saving you months of entering trades manually, day after day, and keeping track of the results on an excel spread sheet just to realize after 6 months of wasted time that "no", the idea doesn't work.
In my opinion, expert advisors are useful. For validating new ideas quickly with some degree of accuracy, and for, yes, automating someone's profitable manual system. Many of them are scams? Yes, sure they are. And the reasons are among the following:
1- Selling Expert Advisors to the public that win 40% - 50% of the time (which is the real nature of real profitable methods) wouldn't work for sellers, as the vast majority of buyers would ask for their money back seeing an expert not winning 9 plays out of ten. This leads us to reason number two.
2- Given that the only way to make buyers happy is to give them a substantial greater amount of winners, then EA sellers, focus on strategies with very small TargetPoints and rather large StopLosses. This trick alone, just based on random probabilities, makes any EA have more winners than losers. Which doesn't mean that the EA will be profitable anyways, as one loser will often wipe out the profit achieved by several winning trades.
3- Given that most EAs are scalpers (so that they can be sold to the naive public), the simulations are not valid at all. Scalpers depend on tick by tick data, which is not stored in historical data sets. Historical data-sets go at most to the opening and closing price of the minute bar. This makes back-testing totally unreliable, and prone to interpolation errors.
4- Unsound tactics present in most commercial EAs. Tactics such as Martingales, D'Alembert, Grid trading, are proven unsound tactics, Time bombs that will inevitably explode at some point.
5- Lack of flexibility in the trades management. For example aspects such as fixed TP in pips, fixed Stop Losses distances, or fixed Position sizes will result in poor money management and the inability of the Expert to adjust itself to different volatility environments.
Finally a piece of advice.
If an EA seller doesn't provide you with at least 6 - 8 years of back-test evidence. Plus at least 6 months of live test with Investor (read-only) access to their account through a third party source such as myfxbook.com then it means they are probably hiding something or they don't trust their own expert to trade their money.
I would also add as an advice, never never never trust any back-test performed on a time-frame lower than H1.
Also, never trust affiliate reviews. Learn to spot affiliate reviews and don't trust them.
Automated trading is possible, not only in Forex, but also in commodities, stocks, bonds etc. In fact automated trading is common nowadays and represents the majority of the trading volumes. Get away from over hyped EAs, analyze the trades history thoroughly, evaluate the "evidence" the seller is trying to sell you and don't ever buy anything that doesn't meet your standards. The equity curve alone, its shape tells you a ton of things about the nature of the robot being sold to you. Read this article written by me about the meaning of different forms of equity curve growth, Tips for reviewing Expert Advisors I believe it is a crucial piece.
I would say the best way to trade with an automated software is using a strategy designed by you and hopefully programmed by yourself or someone you hire in case you don't have the programming skills. The evidence is out there, for example in the stats published every month on the Barclay's systematic traders index, only to name a serious source.
Related Articles
Tips for reviewing Expert Advisors
LT Trend Sniper - a Forex strategy that works
The Turtles Trading system automated (Expert Advisor for download)
Go to the bottom of this page in order to see the Legal Stuff
No comments:
Post a Comment