AlgoStrategy Lab

Understand it. Code it. Backtest it.

Systematic trading treated as a research discipline rather than a promise. Every strategy is stated as a hypothesis, coded openly, tested with costs included, and reported with its failures intact.

Research

Hypotheses tested with costs, point-in-time data and out-of-sample evaluation.

Market research

Does RSI Actually Work?

Testing the classic RSI(14) mean-reversion rule across 20 years and 500 instruments, with costs included and multiple-testing accounted for.

IntermediateDifficulty: IntermediatePython · Polars · DuckDB
NotebookDataset

Strategies & backtests

Systems built and tested openly, with the failures reported.

Trading Systems

Algorithmic Trading Backtest Lab

A backtesting framework with transaction costs, slippage, walk-forward analysis and survivorship-bias handling — designed to stop a strategy looking better than it is. Educational research only.

AdvancedDifficulty: Advanced~24h buildPython · Polars · DuckDB
Code

Tutorials & explainers

The methodology behind honest testing.

AlgoStrategy

Why Most Backtests Lie

Backtests fail in a consistent direction — optimistic. Here are the five structural reasons, with a detection method for each. Educational research only.

IntermediateDifficulty: Intermediate2 min readPython · Polars

Tools

Risk arithmetic and backtest statistics, computed in your browser.

Trading researchFree

Backtest Metrics Calculator

Paste a return series and get the risk-adjusted metrics that matter, with the assumptions stated. Educational only.

930 runs
Trading researchFree

Position Size Calculator

Fixed-fractional and volatility-scaled position sizing, with the resulting portfolio exposure shown. Educational only.

750 runs

Datasets & notebooks

Starting points you can run today.

NotebooksFree

Backtesting Starter Notebook

A Jupyter notebook with point-in-time data handling, realistic costs and walk-forward validation already wired in.

180 downloads
Email required

How this Lab works

Every result is reported with what it could not establish

In-sample and out-of-sample results appear side by side. Cost assumptions are stated. The number of variants tested is disclosed. If a strategy only works in one regime, the write-up says so — because a backtest that only ever flatters is not research.