A winner trend system

 

Good morning everyone!!

One of my teachers used to say to the students: "The class is wise", referring to the knowledge of all the class is enormous. The same feeling I perceive in .

That’s why I want to propose you the creation of a trend automatic system. With the experience of everyone who wants to participate in this challenge: the creation of a winner system.

For example we can start using the system of minimum and maximum breakpoints. If the price exceeds the maximum price of the last n candles, opens a buy.

It will close when the price is less than n candles. As the next example chart:

In this case if the price exceeds the maximum price of the last 100 candles, opens a buy order. And if the price closes below the minimum of the last 40 candles, the order gets closed.

I developed the system for MetaTrader 4 and the example is a one-hour timeframe and EURUSD from 2000 to 2009.

The results are shown in the following graph:

The report is:

What do you think of this system? I think it’s mediocrity, beacuse the benefit and profit factor are scarce, and the maximum drawdown of 13.58% is quite high, no one would like to lose 13%.

It's great because you can improve this system a lot, for example the benefit operations are only 41.20%, so it would be nice to add more filters to the system input, and then go testing them.

Espero vuestra colaboración!! Seguro que se puede mejorar muchísimo añadiendo filtros de RSI, Stochastic, etc.…

I’m waiting for your ideas!! I’m sure we can improved this system greatly by adding filters like RSI, Stochastic, etc....

See you soon!!

 
I dont speak spanish or whatever language is the statement written on, but i can tell that the red text at the mode of rendering is probably set to "control points" or inancurate testing.You should backtest it with "every tick" method which is the most precise, on like atleast 1000 trades on atleast 3 different years, so for example 2008- 2009 (a challenging period),2009-2010 and 2012-1013, with 1000 trades each backtest, and try to vary it, for example test it on winter and spring sessions also because the season change does some changes in the market too.And if its holding after this then you can give it a forward test, on a demo account of course!
 

you can forward test an EA on the live price chart, just substitute the ordersend with a method of recording the prices at which the orders would have been opened and closed.

 

Congratulations, you just discovered the Turtle Trading Strategy.

https://www.mql5.com/en/code/10727

 
I wouldn't get too excited about a backtest with low modelling quality and no slippage that makes 2.7% per year .
 

Hi, I’m back!

I included the smooth moving average in the system, for better detection of the main trend. After that change the 1 hour chart has no visible results. The results do not vary substantially even if modifying the period of the moving average.

I had to reduce the test period to 15 minutes and the moving average filter has worked much better.

The parameters I used are the following:
Number of candles of resistance line 80
Number of candles of support line: 50
Period Smoothed Moving Average: 230
Stoploss: Not applicable. (by now)
TafeProfit: Not applicable. (by now)


And the results from 1/1/2000 to 1/1/2010 are:

The profit chart has improved but there are still periods of Great Falls:

The year 2008 was terrible for many automated systems, it seems that in our case too. The following graph shows three consecutive failures that occurred from 13 to November 19:

In my opinion the winner trend system is that we have greatly increased the benefit and the number of operations, but we haven’t reduced the maximum loss and we haven’t improved the profit factor.

EURUSD 2000-2010

Break Resistance

Breaks and Moving averege

Benefit

2734

4155

Profit Factor

1.25

1.19

Operations

233

867

Maximun Drow Dawn

1600

1613

Benefit operations

41.20%

43.02%

What do you think the system? Do we add more indicators, such as a filter for false signals, add the StopLoss?

I’m waiting for your ideas!!

Of course if someone wants to test the system, send me a message and I’ll send you by email the latest system. It’s developed with MetaTrader 4.
Thanks and regards!

 
4% over 10 years is not good. You would do much better in a simple interest bearing account.
 
GumRai:
4% over 10 years is not good. You would do much better in a simple interest bearing account.
Not in the UK, 2% is about the norm here.
 
RaptorUK:
Not in the UK, 2% is about the norm here.


Sorry, I was not clear :(

The backtest showed 4,000 profit on a 100,000 account over 10 years, 4% profit, 0.4% per year

 
GumRai:


Sorry, I was not clear :(

The backtest showed 4,000 profit on a 100,000 account over 10 years, 4% profit, 0.4% per year

Ah . . .
 

Hi

You can start with 2000$ account, so the profit will be 200% :)

This EA needs more filters, any suggestion??

regards!!