Matt Dancho – Backtesting Algorithmic Trading Strategies with Python

Matt Dancho’s Backtesting Algorithmic Trading Strategies with Python is a hands-on course on portfolio-based quantitative trading. It uses projects, strategy files and reusable code templates to explore portfolio construction, risk management and backtesting.

The course welcomes beginners, including those without prior algorithmic trading or Python experience. Trading involves risk, and backtested results do not guarantee future performance.

Skills covered in Python backtesting

• 1. Understanding the basics of algorithmic trading.
• 2. Setting up the Python environment for backtesting.
• 3. Implementing various trading strategies in Python.
• 4. Analyzing backtest results for strategy effectiveness.
• 5. Optimizing strategies based on backtesting results.
• 6. Best practices for algorithmic trading development.

Build a Python quant trading lab

The course begins with a local Python Quant Lab: installing the Quant Stack software, organizing an algorithmic trading project and preparing the environment for testing, automation and data integration. Matt’s templates provide a structure for the project workflow.

Develop portfolio trading strategies

The strategy material covers four algorithmic portfolio strategies and introduces volatility targeting with automatic rebalancing in a style inspired by Ray Dalio. It also uses the Riskfolio-Lib Python library for portfolio optimization, minimum return thresholds and diversification.

The course includes code templates and strategy files for working with these techniques.

Backtest with Zipline Reloaded

Event-driven backtesting with Zipline Reloaded provides the framework for testing custom strategies. The workflow models rebalancing, slippage and commissions, then addresses debugging, performance optimization and reporting through a backtest engine and reporting pipeline.

Step-by-step course modules

Step 1: Quant Lab setup

Set up the Quant Stack Python environment, create an initial algorithmic trading project and use Matt’s templates to organize the work. This step also prepares the project for testing and data integration.

Step 2: Build a risk-managed strategy

Work with Matt’s volatility-targeting strategy and construct a risk-managed portfolio using Riskfolio-Lib. The module introduces the “Bridgewater Cheat Code,” inspired by Ray Dalio’s fund, alongside asset-class allocation methods used by quantitative funds.

Step 3: Event-driven backtesting

Use Zipline Reloaded to implement event-based tests with automated rebalancing and trading constraints. The module incorporates slippage, commissions and transaction costs and addresses five common backtesting mistakes.

Bonus modules: market data and volatility targeting

Premium data for US equities

Learn to ingest data covering more than 21,000 US equities, convert it into Zipline-compatible bundles and build a local backtesting database. This option uses an optional $50-per-month data subscription.

Free market data for backtesting

A separate bonus covers ingesting and converting free data sources, then using publicly available financial data to backtest strategies without a paid data subscription.

Three volatility-targeting variations

Explore three coded variations intended for different market conditions: Hierarchical Risk Parity (HRP), the CVaR risk measure, and a risk-factor approach using Principal Component Regression (PCR). Full code is included for these models.

About Matt Dancho

Matt Dancho founded Business Science, a data science education platform covering applications of Python and R in finance, analytics and trading. His background includes quantitative finance and algorithmic modelling.

Who the course is for

The material is suited to beginners, data analysts and programmers entering finance, and traders interested in moving beyond indicator-based approaches. It also addresses quant enthusiasts, investors and others interested in rules-based, risk-managed portfolio strategies.

Frequently asked questions

Do I need Python or algorithmic trading experience?

No prior experience is required. The course walks beginners through environment setup and the use of code templates.

What code and materials are included?

The course includes strategy files, templates, backtesting systems and data-ingestion code. The bonus volatility-targeting variations also include their full code.

Do I need to pay for market data?

No. A bonus covers free market data, while the premium data option for more than 21,000 US equities uses an optional $50-per-month subscription.

Does the course cover live trading?

Its focus is strategy development and backtesting rather than trade execution. The source materials describe adapting the systems and code for brokers that support Python APIs.