Комментарии могут оставлять пользователи, купившие или арендовавшие продукт
Kasper Hytonen  
Tested XAUUSD M1 with raw spread account using default settings and every tick based on real ticks 
Syamsurizal Dimjati  
Kasper Hytonen #:
Tested XAUUSD M1 with raw spread account using default settings and every tick based on real ticks 
Hi Kasper,

Here's an in-depth analysis of why this happens and the fundamental differences in the MT5 system:

1. The Difference Between "Every Tick" and "Every Tick Based on Real Ticks"
This is the main reason why backtest results can vary drastically:

Every Tick (Simulation): MT5 creates artificial ticks based on M1 data (OHLC). The movements are smooth, and the spread is usually assumed to be constant (fixed). This often produces "too good" results because there are no price surprises.

Every Tick Based on Real Ticks (Real Data): MT5 uses per-second price movement data from the broker's server that has occurred in the past. This includes:

Wider spreads: During high volatility, spreads can increase 5-10 times. The RUHM EA may try to open when the spread is very wide, resulting in a large loss.

Slippage: Price jumps (gaps) that prevent your execution from being accurate.

Micro-Noise: Small movements that can hit the RUHM Stop Loss before the price moves in the right direction.

2. Why Does the EA Experience Losses? (Signal Analysis)
There are several possible errors in the detection logic when dealing with Real Ticks:

Candlestick Patterns are "Lagging" Indicators: Hammers or Haramis are only confirmed after the candle closes. In a highly aggressive Gold market, by the time the candle closes, momentum has often run out or reversed.

Lack of Trend Filters: Hammer patterns in the midst of a strong downtrend (Bearish) are often just a "short break" before the price drops further. If an EA buys simply because of a Hammer without considering the larger trend (e.g., the 200 EMA), it is very risky.

Risk/Reward Ratio: If your TP and SL are too close (e.g., TP 3000, SL 5000), then your win rate must be above 65% just to break even. With pure candlestick patterns, it is difficult to achieve such a high win rate without additional filters.

3. Do Brokers Hide Data?
No, real tick data is not confidential, but it is very cumbersome. Many brokers only provide real tick history for the past few months. MT5 differentiates this because testing with real ticks requires significantly more computing power. If your broker grants this access, it means the data is their actual market recording.

(That's why real tick data is "available" in the paid testing area.)

4. Problems with EAs Reading Real Ticks
The EA doesn't "fail" to read, but it may be too sensitive or lack protection. In a real market, prices don't move linearly. If the RUHM EA enters based on patterns without considering volume or trend strength, it will often be hit by fakeouts (false signals).

In the next version, I'll add a strategy with two trend filters and a protective spread filter (during news). You can update this in the upcoming version 1.1. Thank you for your feedback.

patrickdrew  

Your graph looks amazing. 

I look forward to testing on demo. 

yoshiyuki shoji  
This EA seems rather good.
Though you won't get a clean curve fitting unless you optimise it for each broker.

It works with real ticks, so it should be viable for live trading too.

I'll run it for a while and review it if it performs well.
Syamsurizal Dimjati  
yoshiyuki shoji #:
This EA seems rather good.
Though you won't get a clean curve fitting unless you optimise it for each broker.

It works with real ticks, so it should be viable for live trading too.

I'll run it for a while and review it if it performs well.
Hi, Yoshiyuki, Thanks

Any individual who contributes through test result analysis (like you), reporting execution anomalies such as fake or false signals, identification of sideways market conditions with compressed volatility, or other relevant technical insights, will be granted free access to this Expert Advisor in the form of the latest updated and production-ready release.

This Expert Advisor is planned to undergo continuous development aimed at improving accuracy, robustness, and adaptive capability in response to evolving market structure dynamics.

Current development is primarily focused on XAUUSD, however, the system remains flexible and instrument-agnostic, allowing deployment on other forex instruments with competitive spreads, in alignment with the EA’s internal architecture and trading logic.

Syamsurizal Dimjati  
patrickdrew #:

Your graph looks amazing. 

I look forward to testing on demo. 

Hi, Patric, Thanks
Continue testing and provide further information
Syamsurizal Dimjati  
Cristian-bogdan Buzatu #:
Let's see the new version.
Hi, Cristian, Thanks
Continue testing and provide further information
yoshiyuki shoji  
I tested it for a full day yesterday, but slippage and fees still seem likely to be issues.
Even as it stands now, I reckon it might just about be profitable, but I suspect it'd be tough if there were a significant drawdown.
Theoretically, it should be profitable, but I feel countermeasures against the broker are needed.
It might be worth preparing separate measures for slippage and fees.
Even then, there's no guarantee it'll be profitable though 💦


Incidentally, the demo account was generating stable profits.

patrickdrew  

On a RoboForex demo account. 

3 days of losing trades!? 

Any idea what is going wrong? I am using the default settings. 

Файлы:
Ruhm.png  10 kb
patrickdrew  
yoshiyuki shoji #:
I tested it for a full day yesterday, but slippage and fees still seem likely to be issues.
Even as it stands now, I reckon it might just about be profitable, but I suspect it'd be tough if there were a significant drawdown.
Theoretically, it should be profitable, but I feel countermeasures against the broker are needed.
It might be worth preparing separate measures for slippage and fees.
Even then, there's no guarantee it'll be profitable though 💦


Incidentally, the demo account was generating stable profits.

Were you on default settings? 

Which broker? 

Syamsurizal Dimjati  
patrickdrew #:

On a RoboForex demo account. 

3 days of losing trades!? 

Any idea what is going wrong? I am using the default settings. 

Hi Pattrick, Thanks

If the commission is greater than each order closed due to trailing (e.g., closing due to SL+/trailing per Lot 0.01 is 0.09, while the commission per Lot 0.01 is 0.15, then the order will incur a loss). Solution: Check the Account Commission (RAW). Check the Spread in RAW. In the EA, adjust the parameters for Trailing. See if the Spread allows for leeway to close profits while trailing is running. And/or find the trading time or market session to avoid severe sideways movements, which erode equity due to the EA being trapped in false signals.

I will improve with Market Session Time. for Ver. 1.2
Syamsurizal Dimjati  
yoshiyuki shoji #:
I tested it for a full day yesterday, but slippage and fees still seem likely to be issues.
Even as it stands now, I reckon it might just about be profitable, but I suspect it'd be tough if there were a significant drawdown.
Theoretically, it should be profitable, but I feel countermeasures against the broker are needed.
It might be worth preparing separate measures for slippage and fees.
Even then, there's no guarantee it'll be profitable though 💦


Incidentally, the demo account was generating stable profits.

In the process of being further improved
Jesús Jiménez  
Hi!! I´ve tested new version with raw account M1 with real ticks and results are not good ... Same Kasper told in first comment .... Is a problem with settings?? In this case, may developer send us a valid .set file, please?? Thanks in advance!!
yoshiyuki shoji  
This is just an idea, but it might be a good idea to set TP to ATR and SL to a fixed value.

I would appreciate your consideration.
Syamsurizal Dimjati  
Jesús Jiménez #:
Hi!! I´ve tested new version with raw account M1 with real ticks and results are not good ... Same Kasper told in first comment .... Is a problem with settings?? In this case, may developer send us a valid .set file, please?? Thanks in advance!!
Hi Jesus,

Check:
1. Look at the XAU price on your chart (e.g., 1).4225.330 or 2).4255.33).
2. If the XAU price is at number 1, you don't need to change the trailing value. However, if the XAU price on your chart is at number 2, then:
3. Change the trailing value to a multiple (e.g., if it was previously 500, change it to 50; if it was 10, change it to 1). This is related to the decimal point. XAU has 2 and 3 decimal places after the comma/price.

Make sure of this first.
Syamsurizal Dimjati  
yoshiyuki shoji #:
This is just an idea, but it might be a good idea to set TP to ATR and SL to a fixed value.

I would appreciate your consideration.
Hi Yoshiyuki,

I think this possibility is impossible, because there is already trailing for TP, if to bring the TP value closer, just change the ATR value for TP (shrink it)
patrickdrew  

Hi again. Can you please explain why you say it is NOT suitable for cent accounts?

Br,

Patrick

Syamsurizal Dimjati  
patrickdrew #:

Hi again. Can you please explain why you say it is NOT suitable for cent accounts?

Br,

Patrick

hi patrickdrew , Thanks

Typically, cent/micro accounts are commission-free, as this aligns with their wide spreads. Wide spreads are also common with standard accounts without commissions.

Wide spreads cause the EA's price calculations to be less accurate.

This EA is designed for fast entry and fast closing.

Profit-taking on this EA, due to a fast closing, already generates profit (even on a zero account, if the price fluctuates or reverses very slowly, reaching BEP). In this case, using a wide spread makes the trailing order incompatible with closing entries with margin. You must recalculate the SL+ value, minimum spread width, and spread protection.

if you read the parameters step by step, you will understand why the EA is not suitable for standard/cent/microcent accounts. Unless you set the trailing order and spread settings for a long distance, waiting for a long entry to close. I hope you understand this.

Комментарии могут оставлять пользователи, купившие или арендовавшие продукт