Overview
This article explains the roadmap we follow to implement our approach of a self-learning and self-adaptive automated trading system. Through this article, we provide insights into the transformative stages our project will undergo as it develops into an automated system the consumer sector has never seen before. Our final goal is to deliver a fully automated trading system that intelligently adapts to changing market conditions in real-time. This will be achieved through sophisticated machine learning processes and artificial intelligence that will both establish the foundational structure of our trading systems and continuously refine and optimize their performance.
We invite investors and contributors who share our vision to connect with us about potential involvement in this innovative project.
The Current State of Trading Automation: A Broken System
For nearly two decades, MQL4 and MQL5 have dominated the landscape of automated trading software development, particularly in the retail trading sector. Their easy accessibility and widespread integration with MetaTrader 4 and MetaTrader 5 platforms have established these programming languages as the go-to solution for implementing trading automation strategies.
However, this approach only works for traders with significant capital to pay experienced developers hundreds of hours, or for those who already know exactly what they want to build and can code by themselves. Since coding typically takes months or years to master, the freelancer market is dominated by semi-professional developers who often reinvent the wheel with each project. In 2025, this principle is highly outdated.
The typical problem with this process is that traders often get trapped in a continuous loop of readjusting to changing market conditions. Depending on the trader’s coding skill level, this leads to either increasing costs for modifications or, in cases of semi-professional attempts, the risk of losing significant funds due to software bugs and functionality issues.
It’s well established that the vast majority of retail traders lose money, whether they trade manually or automatically. This vicious cycle of constant readjustment combined with semi-professional development helps explain why.
The core problem is clear: Traders cannot access a ready-to-use, user-friendly solution that enables them to completely avoid the coding step in their journey toward trading automation. This gap represents a significant market opportunity for a solution the industry has never seen before.
The Solution: Bringing Modern Software Principles to Trading
Modern software development is built on abstraction, simplification, and reusability of thoroughly tested components. This principle governs almost every aspect of our digital world. When you drive a car, you unconsciously rely on dozens of safety systems that make your journey safer, you don’t need to understand how anti-lock brakes work to benefit from them.
Yet in trading, it’s accepted that people must navigate a trial-and-error journey with their hard-earned money, often relying on poorly developed, custom-coded solutions.
We target a solution that addresses these fundamental problems: a centralized tool that enables traders to create and implement automated trading strategies without coding or complicated workflows. Our new product, the Algo Strategy Center, uses a sophisticated software framework that fully encapsulates and abstracts trading complexity into simple object oriented configuration parameters. Instead of writing code, traders focus on what actually matters, the trading logic itself. For example, if you want to use a technical indicator, you simply configure its input parameters rather than adding code lines to an already complex software project.
Through rigorous simplification and abstraction of the MQL programming language and trading in general, we’ve eliminated the compilation step that typically creates barriers for traders. Despite this simplification, there are virtually no functional limitations.
The end result is elegant: traders work with a pre-compiled, universal trading algorithm that’s fully customizable through input parameters rather than copy-and-paste coding. This approach delivers professional-grade automation without the traditional complexity, costs, or risks associated with custom development.
New opportunity - Stage 1 - Algo Strategy Center
FEA Trading already offers two products that fundamentally establish this system, though they’re less accessible than we’d like them to be. Our current solutions, Expert Advisor Builder and Custom Expert Advisor, already utilize the universal trading algorithm introduced earlier, which can be configured through input parameters rather than individual code lines.
This universal algorithm abstracts the entire trading logic and both MQL4 and MQL5 programming languages. However, configuring the input parameters currently requires one of two sub-optimal approaches:
- Manual configuration of an input file using a text editor
- Platform-based configuration through MetaTrader’s input dialogue system
For experienced traders, this already opens a wide range of automation possibilities and has become well-established and accepted. However, it doesn’t feel as user-friendly as it could be. This usability issue stems primarily from the limitations of the underlying trading platform itself. MetaTrader can only handle a limited number of static input parameters, leaving no viable path to enhance the system without partially moving beyond MetaTrader’s constraints. Our new Algo Strategy Center will take a user interface approach that allows traders to add and modify only the trading components actually needed for their specific strategy. This eliminates unnecessary complexity while maintaining full functionality.
The image below shows our first mockup of the Algo Strategy Center interface, demonstrating how we’re transforming strategy configuration from a technical task into an intuitive, visual process.
The Algo Strategy Center will be based on state-of-the-art web technology, which will easily allow us to offer both a web-based version and a desktop-based version.
- Web-based: A lean user interface accessible through our website. This tool will not require any login or registration. Its sole purpose is to generate trading modules which can later be downloaded and inserted into the MetaTrader trading platform. Through simple drag-and-drop actions, it will also be possible to load already existing trading modules and easily modify them.
- Desktop-based: This version requires the initial step of installation. Only by relying on a locally available version is it possible to access the computer’s file systems, thus loading and inserting the trading modules automatically into the installed MetaTrader platforms. On top of this convenience, the local version will include an analytics section to visualize data such as automatically produced trading journals, extended market analysis, and much more.
What is common for both versions is that each configured trading module will be exported as a set file. Depending on the number of trading modules/set files used, two different MetaTrader applications will be used to actually perform the trading idea defined within them.
- Single Module Executor: As the name indicates, this application is designed to execute exactly one trading module. This application aims to be used by advanced and professional users. It also comes with the possibility of using MetaTrader’s built-in optimization feature to find the best input parameters.
- Multi Module Executor: This application can execute multiple (up to 1,024) individual trading modules simultaneously. It targets institutional traders with complex trading schedules who also seek the option of implementing self-adaptive feedback loops.
The illustration below provides a clear overview of the entire functional principle.
Both applications in common are that the trading modules / set files can be replaced during runtime. This ensures that long term performance and drawdown measurements won’t lose its history. Additionally, in the case of the bulk module executor, it allows for example to replace obsolete trading modules without interfering with the already active ones.
New opportunity - Stage 2 - AI enhanced Algo Strategy Center
By achieving stage 1, FEA Trading will already offer a solution that establishes an outstanding market position. Nevertheless, the journey won’t end there. Since artificial intelligence (AI) has reached a very sophisticated level, the next logical step is to include a chat-like assistant that can help create and modify trading modules. The AI assistant will take simple written ideas and transform them within minutes into fully functional trading systems.
While the technical details aren’t fully defined yet, the approach will almost certainly include implementing a Retrieval-Augmented Generation (RAG) AI framework. This will ensure that the underlying AI technology specifically understands how to configure trading strategies based on our trading framework.
The illustration below shows how the practical approach is planned to work:
Let’s assume a user wants a specific trading module. They will be able to simply instruct the AI companion with phrases like the following:
“Please create a trading module that utilizes three simple moving averages (SMA) across multiple timeframes: M15, H1, and H4, with periods of 14, 20, and 50 respectively. The system should execute a trade when the M15 SMA crosses below the H1 SMA, risking 1% of account equity per trade. Each position should be protected with a 20-pip trailing stop loss. This module should operate exclusively on the EURUSD currency pair.“
After a few seconds, this command will generate a new trading module, which may look like the example depicted below:
Advanced algorithmic traders might argue that this approach could lead to situations where many traders use identical trading logic, potentially causing market inefficiencies. However, the AI companion will rely on two technical principles to prevent this:
- Historical Awareness: The system maintains awareness of previously created trading algorithms and introduces a certain level of randomness to avoid the systematization of too many identical trading modules.
- Natural Variation: Due to the non-deterministic character of AI-generated output, trading modules will naturally vary slightly every time the AI assistant is requested to generate the same module, ensuring uniqueness even from identical prompts.
New opportunity - Stage 3 - Self-learning and self-adaptive automated trading system
The final stage is by far the most interesting. At the end of this phase, the previously introduced system will be able to take literal commands, create trading modules with enhanced AI technology, and actually oversee and optimize active trading modules in real-time. This level of automation and self-adaptation will give retail traders the option to create long-term stable automated trading solutions without the need for manual intervention.
The illustration below provides an overview of the entire functional principle.
Again, the details of the technical implementation are not fully defined yet. Since the field of AI is advancing rapidly, it’s possible that by the time of implementation, an advanced AI could take over the role of the machine learning-based algorithm as well.
However, as of now, the functional principle is planned as follows:
The user simply instructs the AI assistant to trade a specific asset, for example, “Trade EURUSD on M15 timeframe with maximum 15% drawdown.” Based on this simple target definition, the system will start creating trading modules, executing them, and optimizing them to achieve this target as quickly as possible. Due to its self-learning character, the system will learn from initially made mistakes, automatically disregarding underperforming strategies while promoting functional ones.
MetaTrader 5 is widely known as the go-to platform for Forex and CFD trading. However, since it’s technically capable of trading real stocks and commodities, the system illustrated above will be able to take over the tasks of entire portfolio management teams.
Technical systems like this already exist, but due to enormous complexity and upfront development costs, these systems are usually only available to large institutions like banks and investment funds. The main difference in our case is that we’re making this technology available to retail traders and smaller institutions who usually cannot afford to implement such a system independently. We’re essentially democratizing institutional-grade portfolio management technology for the broader market.
Timeline
Based on our currently available resources, the timeline for the previously outlined development stages is as follows:
If you’re interested in accelerating this timeline, we welcome support in any form. Due to rapidly growing market conditions, this represents an exceptional market opportunity for investors and contributors. Should you require additional information, we can also schedule a private consultation to discuss everything in greater detail and explore potential collaboration opportunities.
Whether you’re considering financial investment, strategic partnership, or technical contribution, we’re open to discussing how your involvement could help bring this innovative trading solution to market faster while creating mutual value for all parties involved.
Note for current customers
All current users of our Expert Advisor Builder and Custom Expert Advisor solutions will automatically receive a complimentary upgrade to the new Algo Strategy Center at no additional cost. This seamless transition ensures that our existing customers can immediately benefit from the enhanced user interface and expanded functionality without any interruption to their trading operations.
About FEA Trading
FEA Trading specializes in developing premium automated trading solutions for both MetaTrader 4 and 5 platforms, offering everything from assisted trading tools to complete automation systems. Our comprehensive resources blend rigorous technical accuracy with user-friendly explanations that traders can easily understand and implement.
At FEA Trading, we’re dedicated to elevating your trading capabilities and supporting your path toward becoming proficient in algorithmic trading strategies.
