About strategies and optimization

New comment
 
What are the factors that make a strategy a one that does not fit all symbols but needs optimizations?

And the idealistic question, what can be the factors and principles for a strategy that has a chance to fit all symbols without adjusting?

If someone thought about it and wants to share and open discussion.
 

Differences in spread and liquidity make for different patterns on different symbols, so generally speaking there is no one shoe that fits all feet without optimization.

 

Your questions will open a very huge debate...

Theorically it's impossible to create a strategy that can perform well on multiple assets. Forcing a strategy to work on a specific asset optimizing it (trying best parameters) is, most of time, useless.

There are a lot of people that think that a simple MA cross EA can work well if they found "magic" parameters for each asset, but it's not true for a lot of reasons:

  • optimization of parameters is, always, an overfitting of the EA performance in the historial period tested. 
  • adding 2-3 months to test will alter best parameters.
  • best parameters found cannot perform in the same way in the future.

In addition to this we need to talk about how assets (also in same market) are very different: think about how EURUSD and EURCHF move and how DE30 and US30 move, for example... they are same type (forex pairs the first, indexes the seconds) but they need different trading approach and strategies.

The main goal of a good optimization is, in my opinion, optimizing concepts and not parameters: don't waste time in searching best parameters for an EA but think about if a function can improove general performance.

 
Enrique Dangeroux:

Differences in spread and liquidity make for different patterns on different symbols, so generally speaking there is no one shoe that fits all feet without optimization.

The spread is something that not really hinder us because its in your control to not trade at certain high spreads.
Liquidity is out of our control, but what if we try to tell what is it that liquidity affects that we can see? Is it volatility? 

Let' go step by step to try to understand the causes and effects.
 
Amir Yacoby:
The spread is something that not really hinder us because its in your control to not trade at certain high spreads.
Liquidity is out of our control, but what if we try to tell what is it that liquidity affects that we can see? Is it volatility? 

Let' go step by step to try to understand the causes and effects.

You miss the point. A symbol with an average spread of 4 ticks shows different patterns vs a symbol with an average spread of 20 ticks for example. This is somewhat related to liquidity. Typically spreads are lower on assets with high liquitidy. This and this alone will require slight adjustments (optimisation) given the same strategy on different symbols.

 
Fabio Cavalloni:

Your questions will open a very huge debate...

Theorically it's impossible to create a strategy that can perform well on multiple assets. Forcing a strategy to work on a specific asset optimizing it (trying best parameters) is, most of time, useless.

There are a lot of people that think that a simple MA cross EA can work well if they found "magic" parameters for each asset, but it's not true for a lot of reasons:

  • optimization of parameters is, always, an overfitting of the EA performance in the historial period tested. 
  • adding 2-3 months to test will alter best parameters.
  • best parameters found cannot perform in the same way in the future.

In addition to this we need to talk about how assets (also in same market) are very different: think about how EURUSD and EURCHF move and how DE30 and US30 move, for example... they are same type (forex pairs the first, indexes the seconds) but they need different trading approach and strategies.

The main goal of a good optimization is, in my opinion, optimizing concepts and not parameters: don't waste time in searching best parameters for an EA but think about if a function can improove general performance.

Yes, I mentioned I am looking for factors in either way, not conclusions.

About optimizing concepts, are we talking about concepts that are true to every asset? Concept is supposed to be so, isn't it? Are there such concepts and if so, what makes it theoretically impossible to be multi asset without optimizing parameters?

I like the step you took discarding the parameters and replacing with concepts. Still the question is understanding the why's in order to maybe get some practical guidelines

 
Enrique Dangeroux:

You miss the point. A symbol with an average spread of 4 ticks shows different patterns vs a symbol with an average spread of 20 ticks for example. This is somewhat related to liquidity. Typically spreads are lower on assets with high liquitidy. This and this alone will require slight adjustments (optimisation) given the same strategy on different symbols.

Can you tell what type of patterns are likely to show on 4 pips vs what kind on 20 ticks? If so, then its proved and also controlable, meanning we can utilize it. If not, how do we know that the spread is responsible? 
 
Amir Yacoby:
Yes, I mentioned I am looking for factors in either way, not conclusions.

About optimizing concepts, are we talking about concepts that are true to every asset? Concept is supposed to be so, isn't it? Are there such concepts and if so, what makes it theoretically impossible to be multi asset without optimizing parameters?

I like the step you took discarding the parameters and replacing with concepts. Still the question is understanding the why's in order to maybe get some practical guidelines

Optimize concepts can be testing if filter entries using an indicator it's better than don't do that, optimizing parameter is finding parameters for that indicators that procuded biggest profit in a particular past period.

Concepts are strictly related to strategies and strategies are strictly related to assets and historical period, so it can be that your strategy  that performs well until now, from tomorrow need to be stopped and throw away... 
 
Fabio Cavalloni:
Optimize concepts can be testing if filter entries using an indicator it's better than don't do that, optimizing parameter is finding parameters for that indicators that procuded biggest profit in a particular past period.


From your experience, does a concept optimized differently on different assets? Filter using an indicator as example. I guess you need to optimize in one or two symbols to draw conclusions for all assets. This is what I call a factor that supports multi asset strategies. Those concepts or factots, and listing them here maybe by whoever interested. The question then remains: supoose we find many more concepts, and deep ones, so what is the reason that we havent created a multi asset steategy? Because optimizig parameters is certainly one reason, but still not the only one.
 
Amir Yacoby:
Can you tell what type of patterns are likely to show on 4 pips vs what kind on 20 ticks? If so, then its proved and also controlable, meanning we can utilize it. If not, how do we know that the spread is responsible? 

Try for yourself. Take your strategy and run it on the same symbol same parameters but 2 different brokers. One with low spread, one with high spread. Your results will be different.

 

Let's correct myself to clarify my point of view.

There are a lot of strategies that can perform on multi assets, but the most important thing is analyze markets and understand when a specific asset is in the right market cycle to use that strategy.

For example. I can create a trend following strategy and say that it works only on EURUSD because, THEORICALLY, it's a strong trending pair, but if I used that strategy in last 2 years, probably I lost a lot of money.

A strategy can be "universal", intended to work on a lot of assets, but behind of using that there is a need of market analysis to understand if it's a good moment to use that or not.

Universal strategies, intended to be used 24/24hrs on a lot of pairs, do not exists... or better... I never found one.

But probably, I gone a bit off topic with this post.
12
New comment