What am I missing

[forexster]

Hi,

I'm an engineer by profession and have never traded real money in forex.
Out of curiosity I started going deeper in to it and learning more.

Here's the thing: I built a trading strategy and I've run it on the past 10 years historical data for the EUR/USD. The strategy has worked, without exception, for every single year it was run on.

Obviously this is strange and I'm wondering if I'm missing something. Are there any rules I'm don't know? Here are the boundary conditions I've considered for my strategy:

1. The commission is in the spread. For EUR/USD this is usually 3 pips.
2. Minimum limit for Take Profit or Stop Loss is 10 pips.

Thats its... Is there anything else I should be considering?

 

[Enivid]

How did you run the tests? I once ran a test of an interesting strategy that turned out particularly profitable. The problem was that my testing script (it wasn't EA, but a plain pip calculating script), by my coding mistake, was taking some of the data from the "future" bars - those that didn't exist at the moment of modeled trade decision.

 

[forexster]

Its running in VBA and no, I've double checked... it is not using future data, (I've set it up in such a way that that is not possible)

 

[capsmart]

Hi Forextester
May I ask if you curve fitted your system? Did you "play" with the parameters till you come up to a profitable result? This is the usual problem when developing automatic systems. That way few "fantastic" trades give so much profit that will overcome all the losing ones. If those conditions will not happen again you will end with a losing system.

 

[Adam FX]

3 pips is a way too much! Check also funding costs, execution speed, slippage etc. Take time to choose your broker. To making profit everything should be perfect

 

[forexster]

Hi Capsmart,
It uses some paramaters which I have tweeked...But the thing is... if it is tweeked for certain conditions and will not work in others, surely my results shouldn't be consistently good with ticker data over 10 full years?

 

[capsmart]

@forextester
In theory yes but not in the real world. The financial situation is changing. Some years ago we had "easy" days of 300 pip moves and now we "suffer" with 60 pip days. It also depends on your system. If it will trade the 5 min chart and you look for 10 pip profits, yes this situation can be produced often, but if you trail your stops and make hundreds of pips from trailing on daily or weekly charts then this situation might not happen again. Of course I am using extremes (5min - weekly charts) to emphasize my point. Just tell me if the system is profitable because of certain trades or if it produces a smooth equity uptrend

 

[Fxpipper]

Tests performed on historical data should serve more as an indication rather than an "actual test". Here's why, the market reacts to several metrics, in real time so any test or analysis built on historical data may not necessarily translate into actual money.
So you may want to test it, using a nano account or micro..just a suggestion.

 

[Petar_Serbia]

Hi,

I'm an engineer by profession and have never traded real money in forex.
Out of curiosity I started going deeper in to it and learning more.

Here's the thing: I built a trading strategy and I've run it on the past 10 years historical data for the EUR/USD. The strategy has worked, without exception, for every single year it was run on.

Obviously this is strange and I'm wondering if I'm missing something. Are there any rules I'm don't know? Here are the boundary conditions I've considered for my strategy:

1. The commission is in the spread. For EUR/USD this is usually 3 pips.
2. Minimum limit for Take Profit or Stop Loss is 10 pips.

Thats its... Is there anything else I should be considering?

Hello there,
Testing with historical data and real trading and making profits are 2 different worlds.You can test 100 years and see good results but once you start on real market that dont work on that way.
If you use robot,you will have spikes,slipages and also your broker know extacly what you trade where robot send request to their database.
If you trade manually you will have emotions when you start with real money so you must have very clear mind and very precise set of rules to make consistent profit and of course large knowledge and practice...

Best Regards,
Petar

 

[hayseed]

metatrader

Its running in VBA

//----

you might try converting the vba to mq4 and run a backtest with metatrader... or perhaps backtest with metastock, diff language of course......h