Back test starting from $100 to over 180 million | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Back Test vs Forward Test 6 replies

High Leverage Trading, Going For A 100 Million, 100% A Week 149 replies

High Leverage Trading, 100k To 100 Million, 100% A Week 23 replies

Why EA test is forward test better than back test?? 1 reply

Cash & Carry (Grid) took over 180 pips out of the market yesterday and under 1% DD 1 reply

Attachments: Back test starting from $100 to over 180 million
Exit Attachments

Back test starting from $100 to over 180 million

  • #1
  • First Post: Edited May 16, 2015 5:47pm May 15, 2015 8:44pm | Edited May 16, 2015 5:47pm
EDIT - One of my fellow traders discovered a flaw in my testing. So it is most likely that this back test is flawed and would never work on a real account. I have posted the EA in another forum to see if an folks in FF can make it real. http://www.forexfactory.com/showthread.php?t=540091

If was a fun exercise though!

I've been on this forum a long time and seen alot of strange things. Learning different things from so many different people from all walks of life. Using those tips and tricks and what I have gathered on my own research, I started creating my own MT4 expert advisers. To date, I have designed EA's, indicators, systems, etc.

I will be the first to admit that each of the trading systems I have designed failed at some point. Martingale Trading systems eventually blow an account, Time Based Trading does not hold up to the test of time (no pun intended). Lately, I started playing with the basic indicators that are included with MT4.

I settled on the MACD. With it's crossover, I created the 4th version of my Expert Adviser which I called FX Rogue Wave. I was designing a horizontal channel breakout EA and thought the name was fitting. While playing around with the settings, I came across the perfect storm of settings (at least for me).

Low starting balance. Back test is starting $100.00.
Should work on any broker. I tested on FXCM at 1:50 leverage with NFA and FIFO rules in place.
No Martingale
5 year back test results with solid equity curve
This is a 1m timeframe so 25% is the best I could get. I can move to other time frames and it works also, but fewer trades due to longer time frames but still very profitable.

I'll let the results speak for themselves. The Strategy Tester report from MT4 was 168mb in size. So I am just posting a screen shot here.
Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 204 KB


To be honest, I was surprised by the results. So I thought I would share them here. I am not sure what I am going to do with this. Most of us know that back test data does not always correlate to forward test data and forward test data rarely will correlate to Live real money accounts.

So, beginning soon, I am forward testing to see how it does and what additions I need to make this work (at least in some form) in a real money account. Results like this I can't give up easily on. Would you not agree if you were me?
  • #2
  • May 15, 2015 9:54pm May 15, 2015 9:54pm
hi
it seems great result for backtesting.
would you share Your EA here?
thanks
fred
  • #3
  • May 15, 2015 10:00pm May 15, 2015 10:00pm
Following...
  • #4
  • May 15, 2015 10:10pm May 15, 2015 10:10pm
Quoting fredgold3434
hi it seems great result for backtesting. would you share Your EA here? thanks fred
Hi Fred,

Eventually, I might. I was so excited when I saw these results and tested them over and over and over again on various brokers that I jumped to share these results. Since I am still testing and this is a work in progress, I am not sure of it's stability on a forward testing - real money environment. So for the sake of version control and remaining in focus, I am keeping the EA itself close to the vest. If this turns out to be something that even does a percentage of these results on a consistent basis, I will have to do some serious soul searching on the best way to move forward. Share it, sell it, offer subscriptions, who knows. But there is enough evidence to continue work on this project. As of today, it's still too early in the process. But, a forum has to start somewhere.

I think in the near future, I would be willing to post the ex4 file for people to corroborate my findings.

Thanks for stopping by!

Scott
  • #5
  • May 15, 2015 10:20pm May 15, 2015 10:20pm
i think if you see your EA is profitable on demo account,your next step would b to open small live account with some broker and start testing it with lowest lot size like 0.01 to see how does it perform in live forward test.
luck with your test
fred
  • #6
  • May 15, 2015 10:30pm May 15, 2015 10:30pm
Quoting fredgold3434
i think if you see your EA is profitable on demo account,your next step would b to open small live account with some broker and start testing it with lowest lot size like 0.01 to see how does it perform in live forward test. luck with your test fred

Thanks Fred,

Well, I have already discovered long ago that back test results do not immediately correlate to real trading conditions. For example, in back test results, there are no requotes, slippage or communication latency issues. When trading on a demo account, I immediately started getting losses as the broker had minimum distances for Take Profit and Stop Loss modifications. So, I had to find a broker that did not have these limits. Then, once I had a broker with no limitations and started trading on demo, I started getting requotes. So, I added filters to help clean those up. Today, I opened a new demo account with the same settings as shown in the back test and here are those results.

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 40 KB


So, the testing will continue to see if I can spot anything in demo testing that can be used to ensure similar results in a live account. Then, I will run a demo account side by side with a real money account to spot differences there that can be programmed in or out.
  • #7
  • May 15, 2015 10:49pm May 15, 2015 10:49pm
Can you at least explain exactly what is logic behind of your EA ?
i mean if i wann try it manually what is the entry and exit points,SL and TP ?
what indicators should we use to try it out? of course TF and what pairs that EA works on?
regards
Fred
  • #8
  • May 15, 2015 11:18pm May 15, 2015 11:18pm
Quoting fredgold3434
Can you at least explain exactly what is logic behind of your EA ? i mean if i wann try it manually what is the entry and exit points,SL and TP ? what indicators should we use to try it out? of course TF and what pairs that EA works on? regards Fred
Sure, I will run a quick backtest and take some screen shots to try to explain in a nutshell.

Ever since I started trading forex with EA's like A Simple London Breakout V.2 and 7am-9am Big Dog USD Breakout Strategy, I was always fascinated with breakout strategies. What I found recently was that a 1-1 or 2-1 risk/reward ratio was not always the most profitable. Sometimes just being accurate is. For my lack of accuracy, I embedded recovery strategies like the well known Martingale. They would run GREAT for months on end, just to blow an account in a single day wiping out 100's of percent in growth.

So, I stepped back and looked at how to increase accuracy. I came up with targets that were more easily attainable. Like most scalping methods do.

Here is a summary of the parameters.

1m chart - EURUSD
Indicator: MACD set to default values (12,26,9)
When there is a MACD cross over, a channel is defined with several bars in the past.

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 10 KB


If the price crosses a BLUE line, a long order is placed. If a RED line is crossed, a short order is placed. When an order is placed, SL is before the opposing side of the channel and TP is 10-20% above the entry point as defined by the channel size (shown by yellow lines).

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot2.png Size: 47 KB


Of course, I have added many filters to process out communication latency issues, requotes, etc. But that is the gist of it. Obviously, I really do not want to spend my days and nights staring at a 1m chart. So, I designed the EA.

And...Tadaa! We have a winner. A small one to be sure, but multiply that by thousands of trades a year with a 80% accuracy and money management increasing the lot size based on account balance. And there you have it. The theory is simple. Reality, we are still working on.

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot3.png Size: 52 KB
  • #9
  • Edited May 16, 2015 12:13am May 15, 2015 11:45pm | Edited May 16, 2015 12:13am
Quoting PipsExpress
since you don't share, and because of: https://www.google.gr/search?q=FX+Ro...m=122&ie=UTF-8 (includes: http://www.fxroguewave.com/ http://www.myfxbook.com/community/tr...-mod3/746218,1 http://www.forexfactory.com/showthread.php?t=482710 http://alpari.com/en/investor/pamm/317762/ ) this thread should be moved to the commercial dept. of the forum as you are only promoting your sales
Hi Pips,

As I said in my first post, I am the first to admit that earlier versions did not pan out and have gone from version 1-3. I would like to see where this one goes and would be happy to share if it turns out to not be another failure. I've never had results like these. And the Alpari website, is not mine.

If, in fact I choose to make this commercial, then I completely agree that a commercial forum would be the best place for it. But, at this time, it is not the case. I am just happy with my results.

Also, I had not mentioned my website (thanks for the plug). But, yes, I have a website and it has no pricing on it as I have not decided what to do with the EA yet. I have learned that I do not wish to be in the business of selling EA's. But there is value in the experience I have received from this site and there may be benefits to sharing the EA in a more public forum as there are brighter minds than mine that could improve this EA even more.
  • #10
  • May 15, 2015 11:52pm May 15, 2015 11:52pm
happy to see this thread going
  • #11
  • May 16, 2015 12:08am May 16, 2015 12:08am
Quoting Ducdaovn
happy to see this thread going
Thanks for stopping by!
  • #12
  • May 16, 2015 12:13am May 16, 2015 12:13am
hi
i really dont understand that how that channel(yellow lines) and blue and red lines created. i read your post above several times and till not got it clearly.
if you explain more details would be great.
thanks anyway
fred
  • #13
  • Edited 12:32am May 16, 2015 12:17am | Edited 12:32am
Quoting fredgold3434
hi i really dont understand that how that channel(yellow lines) and blue and red lines created. i read your post above several times and till not got it clearly. if you explain more details would be great. thanks anyway fred

Hi Fred,

Once the MACD crossover occurs, the EA draws the channel (red and blue lines) based on the highs and lows of a previous number of candles in the past. This number of candles is user definable. In the screen shot, I am using 2 candles in the past. The yellow lines are the calculated Take Profit lines.
  • #14
  • May 16, 2015 12:33am May 16, 2015 12:33am
i got it thanks
but dont you think 2 pips for profit is very low since you have to consider spread which it normally between 1 to 3 pips could be?
regards
fred
  • #15
  • May 16, 2015 12:34am May 16, 2015 12:34am
Quoting ibthescottyb
{quote} Hi Fred, Eventually, I might. I was so excited when I saw these results and tested them over and over and over again on various brokers that I jumped to share these results. Since I am still testing and this is a work in progress, I am not sure of it's stability on a forward testing - real money environment. So for the sake of version control and remaining in focus, I am keeping the EA itself close to the vest. If this turns out to be something that even does a percentage of these results on a consistent basis, I will have to do some serious...
Hi scott,
I agree, backtest means nothing.
I tried the MACD flipped EA too. the backtest is good. also The RSI EA too. good result on backtest.
Then I tried on forward test got different result.

But I hope you could find the problem and fixed it to a consistent profitable EA.

The idea of ppl share the EA on forum is to help them found the bugs or get various advice to develop the EA to be a better one, until it become consistent profitable on forward testing. It's easier that way.

then maybe after that if they want to make money from it, they will stop the sharing and will put it on commercial content.
Fair enough, you got a free ppl helping you testing,find bugs, and some constructive advice, and you share them the EA.
But the one in commercial content is the latest version with more features,etc.

or maybe not. some ppl don't need money from the EA/signals if the EA/strategy he uses is profitable already, why?he already make decent money from it.
or they just share it as a "charity act". donation for the free EA. then they give the money to he orphanage,etc. I knows some guys doing that and he's still rich from trading.

But it's all up to you, good luck mate, I'll subscribe!
  • #16
  • May 16, 2015 12:39am May 16, 2015 12:39am
Quoting fredgold3434
i got it thanks but dont you think 2 pips for profit is very low since you have to consider spread which it normally between 1 to 3 pips could be? regards fred
Hi Fred,

I think it boils down to what you are looking for. 2 pips profit is nothing compared to large trending markets. But, if I can pocket 2 pips (beyond spread), thousands of times over and over with less than a 1% draw down? When the channel is drawn, it takes current spread into consideration and add or subtracts accordingly.
  • #17
  • May 16, 2015 12:47am May 16, 2015 12:47am
Quoting slianto
{quote} Hi scott, I agree, backtest means nothing. I tried the MACD flipped EA too. the backtest is good. also The RSI EA too. good result on backtest. Then I tried on forward test got different result. But I hope you could find the problem and fixed it to a consistent profitable EA. The idea of ppl share the EA on forum is to help them found the bugs or get various advice to develop the EA to be a better one, until it become consistent profitable on forward testing. It's easier that way. then maybe after that if they want to make money from it,...

Hi Slianto,

Oh, I agree. I started this forum wanting to see where it would go. The backtest results are proof enough for me that I should continue on this course. BUT, there is more work to be done if I want to use this for my own personal trading. Maybe it was premature to start the forum. I was just so excited that I wanted to share. But, I also knew that if I posted the EA now for people to download, I would spend all my time answering operational questions while I still have testing to do. Now that I am comfortable with the backtest, I can watch the forward tests during the week and make changes as it trades to see the best. My own personal downfall, is that I knew that I would not be able to keep track of my changes as well as the 100+ folks that would be making recommendations at the same time and expecting me to answer their questions immediately.
  • #18
  • May 16, 2015 12:49am May 16, 2015 12:49am
ok i got it clearly what are you hunting for in your EA.
thanks
fred
  • #19
  • May 16, 2015 12:51am May 16, 2015 12:51am
I noticed on your charts that the spread was "0", and on the back test the spread was 2, which I believe on the new versions of MT4 is actually .2pips. Either way the spread should be more realistic.
  • #20
  • May 16, 2015 12:56am May 16, 2015 12:56am
ok so please tell us here what can we followers do for you or what are you expecting from us to do for you to make your EA performs better?
regards
fred
1 trader viewing now