Simple EMA mechanical system!

Thank you for your analysis.

In your opinion is it incorrect ?

I need more trades to determine if it's profitable or not, if you have time, help me code an ea to facilitate the backtesting process

I already have an EA that can give 50% WR for a 1:1 R:R it's what I used to create the chart I posted . . . here is the data from the EA

What about your total pips with this ea? (R:R=1:1)

My points were made to made you think.
1. 2. "I don't read my backtest file, do you? There are times that I lose 5 trades in a row because the ranging market. But at last, I'm still being profitable"
Sorry I didn't understand that you actually meant "so far it looks like the strategy is profitable". No it's not. Yes I read you backtest excel sheet.
Quoting RaptorUK: "So a R:R of 1:1 and a Win Rate of 52% ? that is very slightly better than a coin toss, the "slight better" bit could simply be due to your small number of trades . . ."
He's correct the small number of trades make you believe you have a net win of 3 trades. That's not important. The ratio win/(win+loss) is important. The difference between the win and the loss can become any long, in both directions.
Have a look at the animation on this page http://en.wikipedia.org/wiki/Gambler%27s_fallacy
Also your win/loss ratio IS NOT 52%. See http://www.forexfactory.com/showthre...52#post6623952

3.
Capital preservation is much more important than the profits. Each time you lose your trading power reduces. This impairs your ability to overcome a drawdown.

4. "A profitable system doesn't mean that it will always be profitable. I hope that it could be profitable in 2-3 years."
Yes if it is profitable, it is profitable. People believing systems stop working just don't have a profitable system.

5. "34 and 12 because I spent a lot of time to choose the best filter for the system. At last, I think they are the best."
This is called overfitting. This is one of the reasons why people believe "systems stop working"... The system actually never worked.

6. M15 and GBPUSD because the GBPUSD often has very wide daily range and it has also low spread. M15 allows SL smaller and more trades.
If it works only on G/U it doesn't work. G/U could start behaving like E/U does. This is another reason why people believe "systems stop working"...
If it works only on low spread pairs it doesn't work. A profitable system is not a system which aims at only being able to survive the spread.
M15 DOES ABSOLUTELY NOT allow smaller SL than higer TF. It is the excact opposite. In absolute value the stop is smaller but relatively to the volatility it is much bigger than H4 or daily.
The more trades the more cost. Better to make 100 pips in one trade than 100 pips in 10 time 10 pips trades. It's better not only for the spread cost but also for the risk taken. 10 entries=10 risks.

7. So I just want to respect the author's mental rules. That's why I kept 34 EMA close too.
You also said you modified it a lot. This EMA is just plain useless. Did you notice that the channel forms some sort of Schmitt trigger?

8: 10000$ with 1 lot because if you want to blow up your account, you have to lost 10000 points in a row.
So a bad value for a bad reason.

9: You know that the ranging market happens most of the time on the market. So I use this ratio for safe
It is exactly the reason why this ratio is UNSAFE. Your win/loss ratio is awfully bad you must have a high R:R.

Regarding the entry, I realize that I looked at the wrong place. If I had been in front of this chart at that time I would have shorted targetting... your SL!
I though you were talking about the narrow range breakout trade.



"Because most of my lost trade is due to the ranging period. And most of my winning trade is due to trending period"
So let the winning trades grow!!!

It's a Coin Toss . . . it will never make money consistently.

Why don't you subcribe in this thread and check it out

Sorry, I don't understand what you mean

My points were made to made you think.
1. 2. "I don't read my backtest file, do you? There are times that I lose 5 trades in a row because the ranging market. But at last, I'm still being profitable"
Sorry I didn't understand that you actually meant "so far it looks like the strategy is profitable". No it's not. Yes I read you backtest excel sheet.
Quoting RaptorUK: "So a R:R of 1:1 and a Win Rate of 52% ? that is very slightly better than a coin toss, the "slight better" bit could simply be due to your small number of trades . . ."
He's correct the small number of trades make you believe you have a net win of 3 trades. That's not important. The ratio win/(win+loss) is important. The difference between the win and the loss can become any long, in both directions.
Have a look at the animation on this page http://en.wikipedia.org/wiki/Gambler%27s_fallacy
Also your win/loss ratio IS NOT 52%. See http://www.forexfactory.com/showthre...52#post6623952
=> why don't you forget about the win rate? just see the pips gained. Just subcribe my thread and see how this strategy will go on.
3.
Capital preservation is much more important than the profits. Each time you lose your trading power reduces. This impairs your ability to overcome a drawdown.
=> it's right. I agree with you about it.
4. "A profitable system doesn't mean that it will always be profitable. I hope that it could be profitable in 2-3 years."
Yes if it is profitable, it is profitable. People believing systems stop working just don't have a profitable system.
=> because this strategy is pure mechanical, not discretional. So the parameters should change according to market condition.
5. "34 and 12 because I spent a lot of time to choose the best filter for the system. At last, I think they are the best."
This is called overfitting. This is one of the reasons why people believe "systems stop working"... The system actually never worked.
=> 34 and 12 parameters are best until now. If I choose the others, the result is not as good as now.
6. M15 and GBPUSD because the GBPUSD often has very wide daily range and it has also low spread. M15 allows SL smaller and more trades.
If it works only on G/U it doesn't work. G/U could start behaving like E/U does. This is another reason why people believe "systems stop working"...
=> This problem is just like 34 and 12 above, I think this pairs and time could fix the system the best.
If it works only on low spread pairs it doesn't work. A profitable system is not a system which aims at only being able to survive the spread.
=> Right now I only backtest manually, this action takes a lot of time and energy. If I have a robot, it could help me faster. I can test on any timeframe, and pairs. Now I only focus on the GBPUSD and M15
M15 DOES ABSOLUTELY NOT allow smaller SL than higer TF. It is the excact opposite. In absolute value the stop is smaller but relatively to the volatility it is much bigger than H4 or daily.
=> I agree with you this opinion.
The more trades the more cost. Better to make 100 pips in one trade than 100 pips in 10 time 10 pips trades. It's better not only for the spread cost but also for the risk taken. 10 entries=10 risks.
=> It seems that you are following James16 thread. That's right! But I have to accept this truth.
7. So I just want to respect the author's mental rules. That's why I kept 34 EMA close too.
You also said you modified it a lot. This EMA is just plain useless. Did you notice that the channel forms some sort of Schmitt trigger?
=> Do you agree that with this EMA the chart seems to be more beautiful

8: 10000$ with 1 lot because if you want to blow up your account, you have to lost 10000 points in a row.
So a bad value for a bad reason.

9: You know that the ranging market happens most of the time on the market. So I use this ratio for safe
It is exactly the reason why this ratio is UNSAFE. Your win/loss ratio is awfully bad you must have a high R:R.

Regarding the entry, I realize that I looked at the wrong place. If I had been in front of this chart at that time I would have shorted targetting... your SL!
I though you were talking about the narrow range breakout trade.
=> I use the fibonacci retracement to illustrate my trade entry, my sl, my tp, there are comments on each line, you could see that easily... Just subcribe to this thread. If you have any advices. Just tell me. I have a dream that someday, I can make profit from forex automatically. Hope this time come true

I mean that I think why don't you subcribe to this thread to check if a 50% winrate RR=1:1 strategy will make profit or not

I guess you don't understand the chart and table I posted.

Any Strategy with a R:R of 1:1 and a Win Rate of 50% will break even at best, if you haven't accounted for the spread then it will lose money . . . there is no debate or argument about this, it is simple mathematics. Of course anything can make a few winning trades over a short period of time . . . but the aim is to be consistently profitable, isn't it ?

As said RaptorUK, 50% winrate and RR 1:1 will never make money. If you know very little math, you understand that.

Don't need to wait for this train wreck. Simple grade school arithmetic will tell you in a few minutes what must happen and it isn't pretty.

"why don't you forget about the win rate? just see the pips gained."
Because pips gains are not a valid metrics. The only two things you have to focus on are the R:R (R times your SL) and the winrate (p, probability of winning). Altogether. One doesn't come without the other. They are linked by the relation E = (1 + R)p - 1, the expectancy of your system.

If p=50%=0.5 and R=1/1=1, E=(1 + 1)*0.5 - 1 = 0. You expect an average of 0 pips per trade.


"So the parameters should change according to market condition"
Of course!! Mechanical doesn't mean stupid. Why do you think I bother finding an algo to find the S/R levels? (http://www.forexfactory.com/showthre...16#post6898616)

"34 and 12 parameters are best until now. If I choose the others, the result is not as good as now."
Take two random numbers one bigger than the other and do your BT again. You will be very surprised!

"Right now I only backtest manually, this action takes a lot of time and energy. I can test on any timeframe, and pairs. Now I only focus on the GBPUSD and M15"
In this case put your energy on the very bad situations. GBP/SEK, USD/MXN. Trading those is like hugging a cactus you always feel better when you stop. But they are the acid test to a strategy! If it works here it just works!

"If I have a robot, it could help me faster."
This is THE mistake is see again and again in FF; First you design a statistically valid methodology and AFTER you code. Not the other way around.

"It seems that you are following James16 thread. That's right! But I have to accept this truth."
No I don't. I've read several times that this is a good thread but the first page is an ad insert for his paying service and after reading 10-20 pages I still don't understand what it is really about. It's now too big to skim out the useless from the useful. I would certainly have read it if I had been around in 2005.

"Do you agree that with this EMA the chart seems to be more beautiful"
Yes it balances the EMAs high and low in a pure feng shui fashion. Yet it is useless for a robot!

"I can make profit from forex automatically. Hope this time come true"
I share this dream. My method is too complex for an abvious programming. But I'm a computer engineer. This challenge is to me more important than making profit!

BTW, replying to your thread subscribed me ;-)

I think, there's way to make a strategy with R:R=1:1 and Winning rate = 50% can be profitable. I will illustrate it



Well, we enter 2 positions at the same volume and TP level and different SL.

For example: buy 0.5 lot @ 1.54674, SL @ 1.54445, TP @ 1.55132
buy 0.5 lot @ 1.54674, SL @ 1.54216, TP @ 1.55132

WHat do you think about my Solution?

I will backtest again and summit the result here as soon as possible

You are scaling out half of the position when it is in the red. I show in Expectancy Management this is worse -in the long term- than letting the original trade run. The original trade is a breakeven-er therefore, with no need for BT, this variant is a loser.

No way. This is 100% obvious that you don't understand maths. No need to look at chart to understand that it isn't possible.

What you are showing here isn't RR 1:1.

Oops example values are so wierd that I didn't even check it was 1:1

Manque de concentration