How to Source and Evaluate Algorithmic Trading Strategies
Summary
The article outlines a process for finding, screening, and preparing algorithmic trading ideas for backtesting. It begins with practical fit: a trader’s discipline, available time, research commitment, capital, programming skills, and income needs all affect which strategy frequency and markets are realistic. It then recommends building a steady idea pipeline from books, blogs, forums, and academic research, while evaluating ideas by evidence rather than asset-class preferences or enthusiasm.
Before implementation, researchers should check whether suitable historical data is available and affordable, and whether the strategy’s details can be replicated. Academic results may omit transaction costs, slippage, spreads, liquidity limits, or order mechanics, so the article advises independently reproducing them with realistic costs. It also highlights the infrastructure demands of data storage and strategy implementation. The piece offers a research framework rather than a tested trading strategy, and its capital examples and technology recommendations reflect the author’s context; no performance evidence is presented.
Key ideas
- Choose strategy frequency and markets to fit your time, capital, technical skills, and tolerance for drawdowns.
- Maintain a systematic pipeline for sourcing and rejecting ideas, and guard against preferences that bias evaluation.
- Treat published strategies as hypotheses that require independent replication and realistic transaction costs.
- Assess historical data access, quality, and expense before committing to a strategy.
- Plan for continuing research and reliable data and technology infrastructure.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.