How to Test a Trading Edge Before Risking Money
Year: 2026
Type: preprint
Abstract: If you are considering trading CFDs with your own money: this article is not a recommendation to do so, and its purpose is not to discourage you. It is to point out things people tend to skip or overlook. This article describes how I search for an edge in CFD markets, and what happens to most ideas once they are tested seriously. It focuses on methodology rather than results: how to treat data as suspect, how to lock in hypotheses, when counting attempts becomes critical, and what calibrating against noise reveals. Topics include data integrity, realistic transaction costs, pre-defined research protocols, multiple testing and noise calibration. I tested many families of strategies based on price and on external data, and found no edge that met the criteria I set in advance. This is a statement about what was tested only, not a claim about the market. The article closes with a 13-point checklist of what a serious check requires before real money is put at risk, including: clean and verifiable data, measured (not historical) costs, hypotheses written before testing, counting every attempt, calibrating the winner against noise, an untouched out-of-sample window, and verifying that live execution matches research. This is a methodology article and a personal experience report, not investment advice. It has not been independently reviewed. Comments, critiques and corrections are welcome: [email protected]
Author: D, Avi
Language: English
Subfield: Finance
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7236707113
