What is considered "good" in a trading strategy?

 

This may sound like a stupid question, but what is the definition of a good strategy?

The ultimate fantasy would be a strategy trades at a massive Profit Factor, with almost 0 drawdown, that has many trades, and works on all Forex pairs on multiple timeframes.

But is that realistic? I have found strategies that work great on one pair, only to flop massively on another pair. Is testing on data from 1980 and 2019 the same?


Are there experienced Forex traders here that can comment on how often, if ever, they changes strategies?

 
Quint Boney:

This may sound like a stupid question, but what is the definition of a good strategy?

The ultimate fantasy would be a strategy trades at a massive Profit Factor, with almost 0 drawdown, that has many trades, and works on all Forex pairs on multiple timeframes.

But is that realistic? I have found strategies that work great on one pair, only to flop massively on another pair. Is testing on data from 1980 and 2019 the same?


Are there experienced Forex traders here that can comment on how often, if ever, they changes strategies?

Good strategies are strategies that can earn good profits without suffering big drawdowns.

For this reason, profit factor and drawdown from strategy tester are useless values because you don't see the most important thing: floating equity and distance between balance and equity.

An EA that closes only trades in profit will have a very good profit factor (nearly +infinite) but it pretend to close only profit trades, leaving loosing ones opened for undefinied time (generating big drawdown and burning accounts).

Usually a good strategy can perform well on different forex pairs, except particular conditions (EA that trades inefficiency founds only on a SPECIFIC currency, for a SPECIFIC reason.)

If you have an EA that trades with 2 moving averages and performs well (for example) only on EURUSD H4 from 2010 to 2019, this will not be a good strategy but only a statistical force of your strategy, that you found peforming well only in a particular scenario that cannot be replicated in the future, of course.

At the end, strategies need to be continously studied and improved, and also leaved alone when they stopped to work as expected, for this reason there will be not Holy grail EA but only some EA that can perform well in a particular scenario and particular time period.

 

I think you may find by doing a walk forward that each pair has its own requirements and that you can not assume that one expert will suit all purposes.

I feel that you will have to work with each currency pair individually, although some pairs will have similar results for  a particular ea.

 
david mackay:

I think you may find by doing a walk forward that each pair has its own requirements and that you can not assume that one expert will suit all purposes.

I feel that you will have to work with each currency pair individually, although some pairs will have similar results for  a particular ea.

A strategy that don't work on everything and I mean everything from currency to stocks to commodity is not a successful strategy but thinking each instrument have it's own requirements is absolutely wrong because it's not a human being to understand it's feeling, it's just numbers going up down and sideways and a simple fact everyone keeps forgetting, when strategy is working it's calculating these numbers going up down and sideways, based on these calculations we get a buy, sell or don't trade, if you figured that out you won big time. 
 
 I think you will always have advantages and disavantages on every strategy. The perfect strategy in my opinion is a strategy that works on a quiet and on a volatile markets, a strategy that works for all sessions, a strategy without a big drawdown (exact numbers, exact tp and sl), and I think the most important one is that it needs to bring you money. You need self-discipline to not over-trade, you need to close the trades before the end of the day, important news and weekends etc. Good luck in creating one!
 
In my point of view, a good strategy is one that has proper money management with very small drawdown. I prefer this strategy so I would control my emotions better and so I would be able to follow my overall strategy... Less emotions too will make your trading less stressful
 
Fabio Cavalloni:

Good strategies are strategies that can earn good profits without suffering big drawdowns.

For this reason, profit factor and drawdown from strategy tester are useless values because you don't see the most important thing: floating equity and distance between balance and equity.

An EA that closes only trades in profit will have a very good profit factor (nearly +infinite) but it pretend to close only profit trades, leaving loosing ones opened for undefinied time (generating big drawdown and burning accounts).

Usually a good strategy can perform well on different forex pairs, except particular conditions (EA that trades inefficiency founds only on a SPECIFIC currency, for a SPECIFIC reason.)

If you have an EA that trades with 2 moving averages and performs well (for example) only on EURUSD H4 from 2010 to 2019, this will not be a good strategy but only a statistical force of your strategy, that you found peforming well only in a particular scenario that cannot be replicated in the future, of course.

At the end, strategies need to be continously studied and improved, and also leaved alone when they stopped to work as expected, for this reason there will be not Holy grail EA but only some EA that can perform well in a particular scenario and particular time period.

Very well put Fabio. This is what i am finding as well.

 
SilentForex:
A strategy that don't work on everything and I mean everything from currency to stocks to commodity is not a successful strategy but thinking each instrument have it's own requirements is absolutely wrong because it's not a human being to understand it's feeling, it's just numbers going up down and sideways and a simple fact everyone keeps forgetting, when strategy is working it's calculating these numbers going up down and sideways, based on these calculations we get a buy, sell or don't trade, if you figured that out you won big time. 

That's what i thought as well. I'm finding there is market conditions, caused by trading intentions, that change the math tremendously. There doesn't seem to be a "one size fits all" for Forex.

My recent self revelation can be explained like a car and road. A race car does very well, but only when it's on a race track. It will never be suited for an off road course.

And between different forex pairs, and even changing conditions in a single forex pair over time, a single EA simply will not work.