Topic: Live Demo vs Back test issue

Hi guys,

I have created an EA (EURUSD, M1, ST/TP 65/110, wanted to make multiple trades per day) and put it on the MT5, Live Demo test and after one day no Trade was taken.
I did a backtest exactly for the same period and found the EA was taking three trades.
Can anyone help why it is not taking any order on the Live Demo?
Would it be possible to happen on Live Real account too?

Thanks

Re: Live Demo vs Back test issue

Where's the strat?

Re: Live Demo vs Back test issue

I have observed some of the behavior that you mention with some EA's'.

Some EA's in MT4 exported from FSB produce 8 trades per day , while on MT5 the same EA only produce 1 trade per day  .

I do not know why is this happening , so far I have nothing but study the Phenomena .

It will be good to get some feedback from other users with more experience in these topic.

This topic is very important since it opens the door of effectiveness of FSB in translating EA's to MT4 and Mt5.

The statistics of the EA alone will not be enough to make conclusions, only a complete EA' will allow a comprehensive analysis.

FSB creates wonderful EA's but when transferred to MT4 and MT5, the confidence level may change substantially.

After several searches in the Forum I fond some answers to these topic, but I could not find anything conclusive.

it will be good to post your EA , so the example can  be researched completely.

Re: Live Demo vs Back test issue

We can summarize the most common reasons for the discrepancy:
- using different historical data in FSB Pro and MT
- different length of the data series
- different market settings
- using indicators based on Exponential or Weighted Moving Average
- using volume indicators
- using "accumulative" indicators
- different account state (flat, long or short) on the start of the test

Re: Live Demo vs Back test issue

Thanks all for your responses,

I am still wondering why EA should take different actions on, for example, Live today data and offline tomorrow backtest data, actually the data and price pattern is not changing.
When an EA should take action based on an indicator, anytime the indicator output sends a long/short signal it should take that position. anyway, I attache here the EA, its a very simple EA with two indicators Stochastic and MA crosswover.

Post's attachments

10001001 Scalp.mq5 250.54 kb, 2 downloads since 2020-07-13 

10001001 Scalp.xml 13.26 kb, 8 downloads since 2020-07-13 

You don't have the permssions to download the attachments of this post.

Re: Live Demo vs Back test issue

It's really a big problem, to be honest.
I put another EA for M5 timeframe on the live test and had only one trade yesterday. When I did today the backtest offline for yesterday I saw there were two trades.

Is it a problem with MT5 or FSB?

Re: Live Demo vs Back test issue

I agree with you .
As Mr. Popov points out in the thread there are issues with the EA's : including indicators with exponential or weighted MA , what makes any body exporting an EA to MT5 to be extremely careful in the process , because the probabilities to make a mistake are very high.
- I downladed your EA's and noticed that :
1- you are using exponential and weighted MA.
  2- Try a new EA with defaults only ( since defaults will translate better to MT5)
you can try something like:
Strategy Overview
Strategy name: New Strategy
Forex Strategy Builder Professional v3.8.8
Date: 8/14/2020 17:02:59

Table of Contents
Description
Logic
Strategy Properties
Indicator Slots
Statistic Information
Additional Statistics
Description
None.

Logic
Opening (Entry Signal)
Open a new long position at the beginning of the bar when all the following logic conditions are satisfied:

Stochastics* (Exponential, 5, 3, 3) - Slow %D rises; and
Moving Averages Crossover* (Close, Simple, Weighted, 13, 21, 0, 0); Fast MA is higher than Slow MA; and
TEMA (Open, 14, 0) rises.
Open a new short position at the beginning of the bar when all the following logic conditions are satisfied:

Stochastics* (Exponential, 5, 3, 3) - Slow %D falls; and
Moving Averages Crossover* (Close, Simple, Weighted, 13, 21, 0, 0); Fast MA is lower than Slow MA; and
TEMA (Open, 14, 0) falls.
Closing (Exit Signal)
Close an existing long position when the market rises 200 points from the last entry price.

Close an existing short position when the market falls 200 points from the last entry price.

Handling of Additional Entry Signals**
Entry signal in the direction of the present position:

No averaging is allowed. Cancel any additional orders which are in the same direction.
Entry signal in the opposite direction:

No modification of the present position is allowed. Cancel any additional orders which are in the opposite direction.
Trading Size
Always trade a constant number of lots.

Opening of a new position - 0.01 lots.
Protection
The strategy does not provide a permanent loss limitation.

The strategy does not use a Take Profit.

--------------
* Use the indicator value from the previous bar for all asterisk-marked indicators!
** The averaging rules apply to the entry signals only. Exit signals close a position. They cannot open, add or reduce one.

Top

Strategy Properties
Handling of Additional Entry Signals
Next same direction signal behavior - Does nothing
Next opposite direction signal behavior - Does nothing

Trading Size
Maximum number of open lots - 0.05
Number of entry lots for a new position - 0.01

Protection
Stop Loss - None
Take Profit - None
Break Even - None

Top

Indicator Slots
The slots show the logic for the long positions only. Forex Strategy Builder automatically computes the proper logic for the short positions.

Strategy Properties
Same direction signal Nothing
Opposite direction signal Nothing
Stop Loss None
Take Profit None
Break Even None
Bar Opening
Enter the market at the beginning of the bar
Base price Open
Stochastics
Slow %D rises
Smoothing method Exponential
%K period 5
Fast %D period 3
Slow %D period 3
Level 20
Use previous bar value Yes
Logical group A
Signal shift 0
Signal repeat 0
Indicator symbol Default
Indicator period Default
Moving Averages Crossover
Fast MA is higher than Slow MA
Base price Close
Fast MA method Simple
Slow MA method Weighted
Fast MA period 13
Slow MA period 21
Fast MA shift 0
Slow MA shift 0
Use previous bar value Yes
Logical group A
Signal shift 0
Signal repeat 0
Indicator symbol Default
Indicator period Default
TEMA
Moving Average rises
Base price Open
Period 14
Shift 0
Use previous bar value No
Logical group A
Signal shift 0
Signal repeat 0
Indicator symbol Default
Indicator period Default
Take Profit
Exit at Take Profit level
Take Profit 200

Top

Statistic Information
Input Parameters Account Statistics
Profile name Default profile Acceptance criteria Fulfilled
Data source name IG Net balance 1096.89 USD
Strategy name New Strategy Intrabar scanning Not accomplished
Chart EURUSD M1 Ambiguous bars 0
Data bars 50000 Backtest quality 100 %
Time of updating 2020-08-12 18:49 Profit per day 1.98 USD
Time of beginning 2020-06-25 00:31 Max consecutive losses 0
Spread 10.00 points System quality number 100.00
Swap long 0.05 points Sharpe ratio 36.30
Swap short -0.43 points Average HPR 0.18 %
Commission 0.00 points Profit factor 96.89
Slippage 0.00 points Tested bars 49976
Interpolation method Pessimistic Minimum balance 1000.00 USD
Initial Account 1000 USD Maximum balance 1096.89 USD
Leverage 100 Max balance drawdown 0.00 USD
Account exchange rate Not used Max equity drawdown 11.35 USD
Auto scan Switched on Max equity drawdown 1.12 %
Trade until Margin Call Switched on Max stagnation 5 days
  Average profit 3.80 USD
  Average loss 0.00 USD
  Executed orders 103
  Win/loss ratio 1.00
  Time in position 99%

Top

Additional Statistics
Parameter Long + Short Long Short
Initial account [USD] 1000.00 1000.00 1000.00
Account balance [USD] 1096.89 1072.22 1024.67
Net profit [USD] 96.89 72.22 24.67
Net profit [%] 9.69 7.22 2.47
Gross profit [USD] 96.89 72.22 24.67
Gross loss [USD] 0.00 0.00 0.00
Profit factor ∞ ∞ ∞
Annualized profit [USD] 736.73 549.17 187.56
Annualized profit [%] 73.67 54.92 18.76
Minimum account [USD] 1000.00 1000.00 1000.00
Minimum account date 6/25/2020 6/25/2020 6/25/2020
Maximum account 1096.89 1072.22 1024.67
Maximum account date 8/12/2020 8/12/2020 8/12/2020
Absolute drawdown [USD] 0.00 0.00 0.00
Maximum drawdown [USD] 0.00 0.00 0.00
Maximum drawdown [%] 1.12 0.00 0.00
Maximum drawdown date 6/25/2020 6/25/2020 6/25/2020
History bars 50000 50000 50000
Tested bars 49976 49976 49976
Bars with trades 0 0 0
Bars with trades [%] 0.00 0.00 0.00
Number of trades 51 38 13
Winning trades 51 38 13
Losing trades 0 0 0
Win/loss ratio 1.00 1.00 1.00
Maximum profit [USD] 1.90 1.90 1.90
Average profit [USD] 1.90 1.90 1.90
Maximum loss [USD] 0.00 0.00 0.00
Average loss [USD] 0.00 0.00 0.00
Expected payoff [USD] 1.90 1.90 1.90
Average HPR [%] 0.18 0.18 0.19
Geometric HPR [%] 0.18 0.18 0.19
Sharpe ratio 36.30 48.57 118.94

Top

Re: Live Demo vs Back test issue

Thanks Maximus1;

I think the problem is more complicated than just using exponential or weighted MA.
For someone like me with very limited experience in EAs, the software should work in its best.
I have another experience with an EA that uses only MACD, RSI and Oscillator of ROC. after a week with zero trade, I just found that there were six trades in the backtesting.

I am really wondering if there is anybody who has used FSB EA on a real account?