Third Edition · Out now
Machine Learning
for Trading
A structured workflow for building systematic trading strategies. From hypothesis formulation through production deployment.
Foundations
Process & Financial Data
Ch 1-5
Features
Research & Features
Ch 6-10
Models
ML Pipeline to Causal ML
Ch 11-15
Strategy
Backtest to Execution
Ch 16-20
Advanced AI
RL, RAG & Agents
Ch 21-24
Production
Deploy & Operate
Ch 25-27
Foundations
Ch 1-5
Features
Ch 6-10
Models
Ch 11-15
Strategy
Ch 16-20
Advanced AI
Ch 21-24
Production
Ch 25-27
More than a book
An integrated learning system: structured content, production software, and AI-powered research tools.
27 Chapters
Six parts from foundations to production. Covers data infrastructure, feature engineering, ML models, backtesting, GenAI, and live deployment.
9 Case Studies
End-to-end strategies across equities, ETFs, crypto, options, futures, forex, and commodities. Each case study is a complete, runnable implementation.
6 Python Libraries
Production packages covering data acquisition, feature engineering, finance-native model development and diagnostics, event-driven backtesting, and live trading with broker integrations.
112 Primer Topics
Foundational concepts in machine learning, statistics, and quantitative finance. Glossary, prerequisites, and background for each chapter topic.
61 Agent Skills
Autonomous workflow tasks with built-in guardrails against lookahead bias, data leakage, and multiple testing errors. From data fetching to strategy evaluation.
Agent Lab
AI-powered research environment where agents generate forecasts, analyze signals, and surface market insights. Try live forecasting in real time.
The Insights newsletter
Twice-weekly notes from the author
A new paper, a book passage worth revisiting, a result that changes the picture. Book-led and current-led pieces on machine learning for trading — and applied AI more broadly — delivered Tuesdays and Fridays.
Read & subscribeFree • Hosted on Substack • Unsubscribe anytime
From hypothesis to production
Each part of the book maps to a stage of the ML4T workflow.
Introduction & Financial Data
Process & Financial Data Ch 1-5ML4T workflow, data infrastructure, and evaluation protocols
Research Design and Feature Engineering
Research & Features Ch 6-10Alpha factors, text features, and label construction
Model Development
ML Pipeline to Causal ML Ch 11-15Time series, boosting, deep learning, causal inference
Strategy Implementation
Backtest to Execution Ch 16-20Backtesting, portfolio, risk, and strategy synthesis
Advanced AI
RL, RAG & Agents Ch 21-24Reinforcement learning, RAG for finance, knowledge graphs, autonomous agents
Production
Deploy & Operate Ch 25-27Live trading, MLOps, and systematic edge
Purpose-built Python libraries
Six packages covering the full workflow, from data acquisition to finance-native modeling and live trading.
ML4T Data
Data Docs →Unified market data acquisition from 19+ providers
ML4T Engineer
Signal Docs →Features, labels, alternative bars, and leakage-safe dataset preparation
ML4T Models
Models Docs →Finance-native latent factors, SDFs, direct prediction, and portfolio learning
ML4T Diagnostic
Evaluation Docs →Feature validation, strategy diagnostics, and Deflated Sharpe Ratio
ML4T Backtest
Strategy Docs →Event-driven backtesting with realistic execution
ML4T Live
Deployment Docs →Production trading with broker integrations