The history quality of backtest data

New comment
 

Hello.

I'm pretty new to this EA stuff and I got a question. Recently, when I was testing a strategy on demo I noticed that history quality was very red (20%) even though the other results were great and ended up with nice profit!
So my question is: how the history quality can affect my future trading results?

 
Augustinas Dabužinskas:

Hello.

I'm pretty new to this EA stuff and I got a question. Recently, when I was testing a strategy on demo I noticed that history quality was very red (20%) even though the other results were great and ended up with nice profit!
So my question is: how the history quality can affect my future trading results?

History quality is a very important factor when backtesting, especially with scalping or short term strategies where every tick counts.

You should always use every tick or every tick based on real ticks data when testing, whenever this is available and you should never trust results with lower than 90 something quality (99% or 100% are considered trustworthy).

 
Eleni Anna Branou #:

History quality is a very important factor when backtesting, especially with scalping or short term strategies where every tick counts.

You should always use every tick or every tick based on real ticks data when testing, whenever this is available and you should never trust results with lower than 90 something quality (99% or 100% are considered trustworthy).

Got it! Thanks!
 
Eleni Anna Branou #:

History quality is a very important factor when backtesting, especially with scalping or short term strategies where every tick counts.

You should always use every tick or every tick based on real ticks data when testing, whenever this is available and you should never trust results with lower than 90 something quality (99% or 100% are considered trustworthy).

Oh and also.. Is there a way I can improve history quality on the strategy I'm testing?
 
Eleni Anna Branou #:

History quality is a very important factor when backtesting, especially with scalping or short term strategies where every tick counts.

You should always use every tick or every tick based on real ticks data when testing, whenever this is available and you should never trust results with lower than 90 something quality (99% or 100% are considered trustworthy).

Only useful for debugging. Optimizing using this mode leads to tester-grails because of the way the tick are generated.

 
Enrique Dangeroux #:

Only useful for debugging. Optimizing using this mode leads to tester-grails because of the way the tick are generated.

That's why I said testing and not optimizing Enrique.

 
Augustinas Dabužinskas:

Hello.

I'm pretty new to this EA stuff and I got a question. Recently, when I was testing a strategy on demo I noticed that history quality was very red (20%) even though the other results were great and ended up with nice profit!
So my question is: how the history quality can affect my future trading results?

Recently I encountered a history quality issue with IC Markets, it continues to show a low quality even though I use same set, same period of time, and same " every tick based on real ticks data" option. I even tried backtesting on newly installed MT5 platforms, on different MT5 versions and on other PCs. Same bad results. Other brokers though show ~100% history quality. I've written to IC Markets and will wait for their response.
Anyways, gladly, I've saved the older backtests, and the difference between results is huge. Those with 100% history quality were showing ~1300$ profit, over 2015.01-2022.04 period. Now, same EA, on same period of time, with same fixed lot size, same DD, shows ~2500$ profit, 10% history quality.
While I'd be glad to have my EA improved by such impressive ~200% profit, I realize the results way too unrealistic.
 
Alexandru Casian #:
Recently I encountered a history quality issue with IC Markets, it continues to show a low quality even though I use same set, same period of time, and same " every tick based on real ticks data" option. I even tried backtesting on newly installed MT5 platforms, on different MT5 versions and on other PCs. Same bad results. Other brokers though show ~100% history quality. I've written to IC Markets and will wait for their response.
Anyways, gladly, I've saved the older backtests, and the difference between results is huge. Those with 100% history quality were showing ~1300$ profit, over 2015.01-2022.04 period. Now, same EA, on same period of time, with same fixed lot size, same DD, shows ~2500$ profit, 10% history quality.
While I'd be glad to have my EA improved by such impressive ~200% profit, I realize the results way too unrealistic.

Hey I have the same issue but I realised that after a month. Do you have any response from them ?
Could you please tell me which broker has the best data. Thanks 

 
@Stan #: Hey I have the same issue but I realised that after a month. Do you have any response from them? Could you please tell me which broker has the best data. Thanks 

Broker recommendations or suggestions are not allowed on the forum. Please do your own research.

 
Stan #:

Hey I have the same issue but I realised that after a month. Do you have any response from them ?
Could you please tell me which broker has the best data. Thanks 

The issue with ICMarkets data is on the real account servers only. Use a demo MT5 account instead, and you'll get better history quality.
New comment