Strategy tester and real account completely different

[Maurizio Lucini]

I bought many experts on the MT5 market who gave good results in testers but put in VPS on a demo and real account they gave completely different results but, not only, by testing a month in real life and at the end of the month by repeating the test over the same period the results are always completely different states.

I'm a technician and I teach computer programming and I find this disconcerting so I'm looking for a detailed explanation that can make me understand how a test done in retrospect with respect to the values actually obtained can give different results. (same broker, same platform, same parameters and therefore in my opinion same input data but completely different outputs)

I thank all those who will work to enlighten me and help me finally understand why as a technician I cannot accept what cannot be rationally explained unless it is malware deliberately inserted by the developers and in this case I have to ask myself why such subjects are not removed from the market.

Subject has been edited by moderator to replace ALLCAPS

[William Roeder]

maurizio lucini: STRATEGY TESTER AND REAL ACCOUNT COMPLETELY DIFFERENT

Don't SHOUT at us, that is very RUDE.

[Fernando Carreiro]

Please don't create topics randomly in any section. It has been moved to the section: Expert Advisors and Automated Trading

[Fernando Carreiro]

Please read the following ... Tips to avoid being scammed …

Please note, that I'm not saying you were scammed. That is not the point.

The point is that EAs are tools, and like any tool you need to know how to use them properly.

Just like a brush is a tool which in the hands of a master you get beautiful art, the same brush in the hands of someone without knowledge or skill, you get rubbish.

Before you can use EAs, learn to trade by gaining the necessary knowledge and skill, not only be able to use EAs properly, but also be able to understand what is a good EA and what is a bad EA.

[Maurizio Lucini]

Fernando Carreiro #:

Please read the following ... Tips to avoid being scammed …

Please note, that I'm not saying you were scammed. That is not the point.

The point is that EAs are tools, and like any tool you need to know how to use them properly.

Just like a brush is a tool which in the hands of a master you get beautiful art, the same brush in the hands of someone without knowledge or skill, you get rubbish.

Before you can use EAs, learn to trade by gaining the necessary knowledge and skill, not only be able to use EAs properly, but also be able to understand what is a good EA and what is a bad EA.

Thanks Fernando, I understand that you love trading, just like I, an old computer programming professor, love algorithms. I am grateful for your suggestions that will help me understand. I can tell you that I totally share what you told me about the absolute need to possess competence and in this regard I can claim to have over 5000 hours of study in technical analysis and programming in MQL therefore, without sin of immodestness, I can say that I know how to use the tools correctly and to refer to your good example of the brush, which I personally use with my students, I would like to add that if the color container says red and then contains undefined colors inside, even Da Vinci he could'n have painted from Gioconda with her magnetic gaze following you.

Thank you for the time you dedicate to me.

A  greeting from Maurizio

Edited by moderator. Product recommendations or advertising is not allowed in the forum.

[sbower1976]

This is interesting as back testing is assuming you open and close positions to historical data. In the real world you might not get to trade at those volumes and prices and it will slip and in some cases you can slip alot. 

I will admit i am totally new to coding and mql5 but i have experience in the markets and also studied countless hours of technical data. I am now learning the coding part ot create my own Bot to trade for me. The reason i stopped trading may years ago is cause i realized that to make a 100,000 you need at least a million in the pot. so to trade with nothing will yield even less.  But if i could create a BOT to do exactly what i would like i believe i could perform so many more trades and compile the amount to eventually make a passive income. Not going to get rich fast but even if it yields 1000 dollars a month ill be happy.

Edited by moderator. Self-promotion or advertising is not allowed on the forum.

[Yip Sin Hang]

It's generally accepted that a test account should not be identical to a live one. The data used in a testing environment typically originates from a single broker, but in reality, users may opt for different brokers.

This is similar to currency exchange outlets, where each outlet provides distinct quotes.

Additionally, can you predict with absolute certainty what time you'll have lunch tomorrow, or in the upcoming weeks?

Such is the nature of trading.

Referring to the Principle of Reflexivity, no event will occur in exactly the same way twice, regardless of how similar the circumstances may be.

[[Deleted]]

here are a few factors that could contribute to the differences in results:

1. Market Conditions: The foreign exchange market (Forex) is highly dynamic and can experience significant volatility and liquidity variations. Market conditions during the testing period may not accurately reflect the conditions during the live trading period, leading to different outcomes.

2. Slippage: Slippage occurs when the execution price of a trade differs from the expected price. In real trading, slippage can happen due to various factors, such as latency, market gaps, and order execution speed. Slippage can impact the performance of trading strategies, especially when dealing with high-frequency or low-latency trading systems.

3. Execution Quality: The quality of order execution can differ between demo and live accounts. While demo accounts aim to simulate real trading conditions, they may not always replicate the precise execution environment of a live account. Real accounts may experience delays, requotes, or order rejection, which can affect strategy performance.

4. Data Quality: The accuracy and completeness of historical data used for testing can have an impact on strategy performance. It's essential to ensure the historical data used for testing is reliable, properly adjusted for splits, dividends, and other market events, and covers a sufficiently long period to capture different market cycles.

5. Over-optimization: Strategies that perform well during backtesting but fail to deliver consistent results in live trading may have been over-optimized. Over-optimization occurs when a strategy is excessively tailored to fit historical data, making it less robust to changes in market conditions. This phenomenon is commonly known as curve-fitting.

6. Broker Differences: While you mentioned using the same broker for both demo and real accounts, it's worth considering that even small variations in pricing, spreads, commissions, and trading conditions can impact strategy performance. These differences can influence the entry and exit points of trades, potentially leading to different outcomes.