Waiting list for SharpeEdge backtesting software:
https://forms.gle/P5GafhzAawTUNkSv6

I tested whether a simple machine learning model can actually find a trading edge, starting with linear regression on the S&P 500. It found nothing robust, tested across 10 instruments and 6 asset classes. Logistic regression, with the right target definition, did. This video walks through the full process: not just the model, but how I checked the result was real before trusting it.

In this video:
– Why machine learning makes sense for trading
– The risks: look-ahead bias, overfitting, non-stationarity
– How logistic regression works
– How to setup the problem
– Python Data Analysis: feature engineering, target definition, correlation analysis, comparison with naive classifier, walk forward analysis, replication across other instruments, regularization using cross validation
– Backtesting the model signal in my own trading software

Resources:
Jupyter notebook: https://github.com/pietrobelli26/sharpe_edge_code

Interested in seeing this ported to an MT5 Expert Advisor? Let me know in the comments.

🔔 Subscribe for algorithmic trading, quant research, and Python strategy development.
📌 Previous video — How to Code a Trading Bot in MetaTrader5:
https://youtu.be/bVbqQzrxxZo

Disclaimer: This video is for educational purposes only and is not financial advice. Trading involves risk, and past performance (backtested or otherwise) does not guarantee future results.

Timestamps:
00: 00 Intro
00: 29 Why ML for Trading
00: 56 Dangers
02: 12 Logistic Regression
03: 03 Regularisation
03: 51 Problem Setup
04: 43 Case Study
05: 20 Python Data Analysis
15: 29 Backtesting the Model

#AlgorithmicTrading #MachineLearning #Python #QuantitativeTrading #LogisticRegression #TradingStrategy