Geepy Scalper

A Scalping Strategy Born from Real Trading Experience

Geppy Scalper is the product of a personal journey, one that began not with code, but with chart time, and a ton of hours focused solely on trading GBP pairs. The idea was simple: master one segment of the market deeply. GBP pairs, particularly GBPJPY, are unique for their distinctive volatility price behavior. Over time, it became clear that this asset class offered powerful trending opportunities.

After 2 years of manual trading, a consistent and profitable strategy emerged, built not on theory, but on experience. Every entry and exit rule was shaped by live execution, repeated observations, and deep familiarity with the pair’s behavior. Once the method proved robust across various market phases, a question naturally followed: what if this could be improved, scaled, and automated?

Not being a developer myself, I initiated a collaboration with a ML & algorithmic trading expert to bring this idea to life. The goal was not to reinvent the logic, but to faithfully reproduce and improve the core manual approach and explore how data-driven methods could enhance it, without compromising its original foundation.

Geppy Scalper is not designed to chase perfect equity curves or impress with optimized backtests. It was built to mirror the structure and logic of a working manual strategy, and to preserve the discipline that made it viable in live markets.

Development Philosophy

This EA was not developed by a programmer seeking to sell algorithms, but by a trader who collaborated with a dev specialized in machine learning to translate a functioning, proven manual strategy into code. The project began as a personal initiative, long before any commercial goals were considered. 

Throughout the development process, specific and strict requirements were set, including:

• The reproduction of custom indicators and logic used during the manual phase.
  
  

• The use of machine learning only as a supporting tool, not as the core of the strategy.
  
  

• A strict and transparent validation process using out-of-sample data only.
  
  

Geppy Scalper integrates a machine learning component that was trained using data up to March 2021. All performance from 2021 onward represents strictly out-of-sample results, with no re-training, no parameter tuning, and no adaptation to new data. As expected and as is healthy in any realistic system, the out-of-sample period shows slightly lower performance compared to the in-sample phase. This is a natural and desirable outcome. It indicates that the model was not overfit to historical data and that its logic generalizes across time without relying on curve-fitted parameters, as expected from a solid strategy backing the entire system.

Trading Logic

Geppy Scalper is a scalping strategy. It is designed to capture short-term market inefficiencies based on price action and volatility patterns that were originally identified and traded manually. Key characteristics of the strategy:

• Wide stop-losses, in line with the manual trading style, allowing trades to breathe in volatile environments.
  
  

• Medium trade frequency, focusing on quality over quantity.
  
  

• No martingale, no grid, no averaging down.
  
  

• High discipline, with strict rule-based trade entries and exits.
  
  

The system is not intended to trade all market conditions. It is selectively active based on predefined technical conditions and internal confidence thresholds, ensuring that trades are only taken when the statistical edge is present.

Performance 

• All tests use real spreads and real tick data.
  
  

• Testing periods include a full out-of-sample window from 2021 to present, with no retraining or parameter adjustment.
  
  

• The EA is compatible with all brokers with a low spread.
  
  

Important

• This EA is not designed for high-frequency or ultra-tight scalping.
  
  

• Default parameters are provided for GBPJPY H1.
  
  

• The algorithm uses a moderate risk profile by default. Lot size, risk percentage can be adjusted.
  
  

How To Run

• Place the EA on GBPJPY H1.
  
  

• Adjust Risk & News, a 2 hours restriction before and after is a minimum

Geppy Scalper represents a realistic approach to algorithmic trading. Rather than overpromising or overselling, it offers a system based on real experience and refined logic. It is a tool designed to apply a consistent edge with discipline and restraint.

Market conditions evolve, and strategies must adapt accordingly. While the current version reflects a stable and tested approach, future updates will be deployed to ensure the strategy remains aligned with changing regimes, shifts, or emerging inefficiencies. Updates may include indicator refinements, model re-training (with full transparency), or risk adjustments, always with the original philosophy and discipline preserved. Any update will be driven by careful analysis, not by pressure to improve backtests.