18 years Amazing back test result | Page 4 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

AMAZING Strategy!!! 3,266 replies

Back Test vs Forward Test 6 replies

Unknown Indicator and amazing result 28 replies

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

Firebird V1.0 an amazing EA ! 33 replies

Attachments: 18 years Amazing back test result
Exit Attachments

18 years Amazing back test result

  • #61
  • Jul 5, 2018 11:29am Jul 5, 2018 11:29am
quality data 25% ?? You dont know how to get propper data for backtesting?

Results will be much diferent in live account as slippage etc...

And backtests in MT4 are generated from OHLC of 1 minute candle not the ticks inside so the backtest is useless. Results will be much different.

Its nice to see this Equity but you will never achive it in reallity believe me. I spend with EAs many years.

Good luck
Adapt useful, reject useless, add what is specifically your own.
David Perk CLS Strategy 1 All Time Return: 3,591.0%
  • #62
  • Jul 5, 2018 11:38am Jul 5, 2018 11:38am
"
quality data 25% ?? You dont know how to get propper data for backtesting?
Results will be much diferent in live account as slippage etc...
And backtests in MT4 are generated from OHLC of 1 minute candle not the ticks inside so the backtest is useless. Results will be much different."

EXACTLY.

Stay safe,

FF
www.sme-fx.com
  • #66
  • Jul 5, 2018 12:03pm Jul 5, 2018 12:03pm
Just IGNORE or BLOCK this TROLL.

And do NOT trust your backtest results. They are COMPLETELY WRONG!

FF
www.sme-fx.com
  • #68
  • Jul 5, 2018 6:30pm Jul 5, 2018 6:30pm
Quoting DaPerk
quality data 25% ?? You dont know how to get propper data for backtesting? Results will be much diferent in live account as slippage etc... And backtests in MT4 are generated from OHLC of 1 minute candle not the ticks inside so the backtest is useless. Results will be much different. Its nice to see this Equity but you will never achive it in reallity believe me. I spend with EAs many years. Good luck
I am not a scalper.
I use min 5 min chart.
Therefore although the price is random in a bar, it does not affect me.
All my parameters are using median that is (H+L)/2 so the data from mt4 is good enough.
I have attached a photo to explain the difference between mt4 data vs 99% quality data.
Attached Image (click to enlarge)
Click to Enlarge Name: tick_vs._mt4.png Size: 47 KB
  • #69
  • Jul 5, 2018 6:37pm Jul 5, 2018 6:37pm
Quoting dukas_trader
{quote} 1) fix Sl and TP dont help anything. only for connection problems, the backtest will be wrong same. 2) 5 min is same problem then 1 min, 5 min is simple calculated from 1 min bars. 1 min is even the best. 3) real account can be extrem in minus and same time backtest can win millions, that happens very often,. because of the big problems in backteste, same happens when trading in demoaccount, there can also be millions and live account later losses. because demoaccount has many many advantages against real account. and your calculation is...
If I uses 3x lot size, the drawdown will definitely be bigger.
If I spread it out to 3 different strategies or pair, the drawdown can be the same or even lesser.

Let me explain the architecture of my system.
For each pair, although I uses the same 3 strategies, but it is tweak in a way that it's not exactly the same.
This is like the same person, wearing different clothes for different occasions.
It is design such that the TP and SL is also quite different for the same strategies on different pairs and within the same pair.
This helps to create a diversity in entry and exit points therefore if I run on 2 pairs, it's actually 3+3=6 "strategies" and for 3 pairs it is 3+3+3="9 strategies"

If you look at my demo result so far, although I have GBPUSD and EURUSD, it doesn't trigger entry at the same time.
It is important for the system not to overtrade in order to minimise risk.
  • #70
  • Jul 6, 2018 4:53am Jul 6, 2018 4:53am
I'm currently working on the optimization of EURGBP to complete the permutation of EUR, GBP and USD since i'm already done with EURUSD and GBPUSD.
So far, it's been quite challenging as the behaviour of EURGBP is different and returns is also not as good.
One main difference is that I will need to use a longer weekly period to determine the trend.
Nevertheless, it still profitable with low DD.

After EURGBP, I will be adding JPY into the permutations.
I expect it to be different as well but it's a test for me to show that my strategies can work for all pairs.
  • #71
  • Edited 5:28am Jul 6, 2018 5:14am | Edited 5:28am
Quoting DonaldDuckie
{quote} If I uses 3x lot size, the drawdown will definitely be bigger. If I spread it out to 3 different strategies or pair, the drawdown can be the same or even lesser. Let me explain the architecture of my system. For each pair, although I uses the same 3 strategies, but it is tweak in a way that it's not exactly the same. This is like the same person, wearing different clothes for different occasions. It is design such that the TP and SL is also quite different for the same strategies on different pairs and within the same pair. This helps to...
the drawdown can be higher or smaller, depends what currency you choose! then win can be higher or smaller trading 1 against 3, depends what currency you choose.
when all your backtests have near same results, like you told.

(H+L)/2 will get problems in news, rollover, weekend open/closing, misquotes in history..... and can bring total crazy results....... how many levels of controlling the quotes for correct possibility you have implemented? at least any big difference you should write in file and control later manuell. or program that it will be corrected inside program... and write file to check for mistake,or write EA and let file read and recalculate your backtester statement..... or or or... many possible.but i would first work on the correct method how to test in backtester or you end like fractafreak, he dont know whats backtester can do because he has no programming skills and write then his bullshit and misleading persons total. take him as a best example of a failed trader..... will you end like him, in the end starting a fraud like him to cheat persons later to get money back you lost in trading?

always basics first: how to backtest correct, when you want save time then learn only how to backtest for your startegy, but better is to learn how all is possible to do and whats total impossible and to avoid to test.
you will learn so much more when you learn basics of the tester and programm strategy first that trades like you did manuell and comapre results of backtester to your real results and see the differneces and reprogramm until its same results or you see if its impossible to reprogram. i cant say , i dont know your strategy, but thats the smart way.... start a point one, your are at point 100 with optimatztion of different currencies and did not really did step 1 with controlling.

you dont have to believe me, but i write EA since the beginning of mt4 , and you can lose so much time doing it not the easy logic way, and when you have time think about backtest helper programs: like automated optimizer (that in 1 backtest step you get all important variables optimized and results in file, and so on., correcter of wrong quotes or news/weekend quotes....... you can save years with this helper programs.....you can use many as "include" parts for later easy integrating inside all your programs.
  • #72
  • Jul 6, 2018 5:22am Jul 6, 2018 5:22am
Quoting dukas_trader
{quote} the drawdown can be higher or smaller, depends what currency you choose! then win can be higher or smaller trading 1 against 3, depends what currency you choose. when all your backtests have near same results, like you told. (H+L)/2 will get problems in news, rollover, weekend open/closing, misquotes in history..... and can bring total crazy results....... how many levels of controlling the quotes for correct possibility you have implemented?
For errors, it can be split into either systematic errors or random errors.
For random errors, if I have test more pairs, it should eliminate the risk because if there is any, the result will be very different.
For systematic errors, I can only rely on the demo results to filter them out at the moment before going live.
  • #73
  • Jul 6, 2018 5:29am Jul 6, 2018 5:29am
Quoting DonaldDuckie
{quote} For errors, it can be split into either systematic errors or random errors. For random errors, if I have test more pairs, it should eliminate the risk because if there is any, the result will be very different. For systematic errors, I can only rely on the demo results to filter them out at the moment before going live.
no, it can be easy programmed that filtered out... if not you need to go thourh any trade by hand--- any bar this trade is open!
thatswhy more easy writing in file at least the "crazy" moments of the backtester quotes, where some problems could happen.
  • #75
  • Jul 6, 2018 6:05am Jul 6, 2018 6:05am
Quoting dukas_trader
{quote} no, it can be easy programmed that filtered out... if not you need to go thourh any trade by hand--- any bar this trade is open! thatswhy more easy writing in file at least the "crazy" moments of the backtester quotes, where some problems could happen.
Thanks. Do you mean this? I googled for automated optimiser and below link came out.
https://www.mql5.com/en/articles/1467

I can add a code to trade based on the hour in a day, set it to optimise and generate a table based on profitability every hour.
I read this somewhere but the idea on the post that I read is to trade only on profitable hours which I feel is curve fitting.
But I can use the same idea to stop opening trades at low volume hours just to see any difference in result.
  • #77
  • Jul 6, 2018 7:01am Jul 6, 2018 7:01am
Quoting DonaldDuckie
{quote} Thanks. Do you mean this? I googled for automated optimiser and below link came out. https://www.mql5.com/en/articles/1467 I can add a code to trade based on the hour in a day, set it to optimise and generate a table based on profitability every hour. I read this somewhere but the idea on the post that I read is to trade only on profitable hours which I feel is curve fitting. But I can use the same idea to stop opening trades at low volume hours just to see any difference in result.
ok, looks like you trade in statisitical edge , not fundamental, so you can get problem that it will not work any time. i dont trade this , so cant speak about this. to risky for me.
  • #78
  • Jul 6, 2018 8:38pm Jul 6, 2018 8:38pm
Ended 2 weeks of my EA with 80% win with 1 open trade over the weekend (I hope it gaps up on Monday).
So far the result from my demo account is better than my backtest.
Who knows, maybe the backtest result is understating the potential of my EA?
Only time will tell.

Attached Image (click to enlarge)
Click to Enlarge Name: Trade Explorer Week 2.PNG Size: 126 KB
  • #79
  • Jul 8, 2018 2:27am Jul 8, 2018 2:27am
I have decided to opening a live MT4 account just for the sake of testing my EA live. Initial funding will be $300.
Once the account is ready, I will link it to the trade explorer.
  • #80
  • Jul 8, 2018 2:40am Jul 8, 2018 2:40am
Quoting DonaldDuckie
I have decided to opening a live MT4 account just for the sake of testing my EA live. Initial funding will be $300. Once the account is ready, I will link it to the trade explorer.
in order, really, since you already link to zulutrade on your profile...https://www.zulutrade.com/trader/372674/trading...
would think the right order is to prove things live before presenting to others.
does the 300$ live shows the level of commitment and trust in the backtest data?
some would think yes. ( me included).

that said, best of luck!
there is always, always another trade!!
0 traders viewing now