Systems research · October 2026

A backtest is a starting point.

I built a trading system. The most useful part was learning how easily a test can answer the wrong question.

The problem I kept coming back to

If a replay uses different entry, exit or risk logic from the live system, its result does not tell you how the live system behaves. The two can look similar and still make different decisions.

What I wanted to inspect

• The same decision logic in replay and execution.
• Signals based on information available at the time.
• Fees, fills and the order of events.
• Risk during a trade, beyond its closing result.
• Saved data and scripts so a result can be reproduced.

This project sits here as a record of systems thinking and testing discipline. Its code and trading strategy remain private.

No return figures or profitability claims are presented. This is a process note, not a public trading product.