Aller au contenu
Jefferson Gonçalves
← Retour aux projets

Ai Quant Book

#AI Quantitative Trading: From Zero to One | AI量化交易从0到1

Build production-ready quantitative trading systems with multi-agent architecture

用多智能体架构构建可落地的量化交易系统

License: CC BY-NC-SA 4.0

Author 作者: Wayland Zhang

#Language Versions | 语言版本

Language Status Progress Link
🇨🇳 中文 ✅ Complete 22 lessons + 30 background + 4 appendices 阅读中文版 →
🇺🇸 English ✅ Complete 22 lessons + 30 background + 4 appendices Read English →

#What's This Book About?

Not a strategy holy grail, but teaching you to build production-ready quant systems.

Most quant tutorials stop at:

  • API translations of backtesting frameworks
  • Stacking technical indicators and parameter optimization
  • Showcasing overfitted "magic strategies"

These let you "play" quant, not "do" quant.

Real quant systems need to answer:

  • Where does data come from? Rate limits, missing values, adjustments, timezones
  • How not to fool yourself in backtests? Lookahead bias, overfitting, transaction costs
  • Why isn't one model enough? Regime changes, signal conflicts, risk diversification
  • How to control risk? Stop loss, position sizing, factor exposure, circuit breakers
  • How to go to production? Execution slippage, monitoring, disaster recovery

This book uses multi-agent architecture to answer these questions: different agents handle different responsibilities (signals, risk, execution), collaborating to make trading decisions.

#这本书讲什么？

不是策略圣杯，而是教你构建可落地的量化交易系统。

市面上的量化教程大多停留在：

  • 某个回测框架的 API 翻译
  • 技术指标的堆砌和参数优化
  • 过拟合的"神奇策略"展示

这些只能让你"玩"量化，不能让你"做"量化。

真正的量化系统需要回答：

  • 数据从哪来？ API 限流、缺失值、复权、时区
  • 回测怎么不骗自己？ 未来函数、过拟合、交易成本
  • 为什么单一模型不够？ Regime 变化、信号冲突、风险分散
  • 如何控制风险？ 止损、仓位、因子暴露、熔断机制
  • 怎么上生产？ 执行滑点、系统监控、故障恢复

这本书用多智能体架构回答这些问题：不同的 Agent 负责不同的职责（信号、风控、执行），协作完成交易决策。

#Content Overview | 内容概览

The book has 5 parts, 22 lessons | 全书分为 5 个部分、22 课：

Part Topic 主题 Lessons 课程 Core Content 核心内容
1 Quick Start 快速体验 1 Quant landscape, multi-agent intuition 量化全景图、多智能体直觉
2 Fundamentals 量化基础 7 Markets, statistics, strategies, data, backtesting 市场、统计、策略、数据、回测
3 Machine Learning 机器学习 2 Supervised learning, from models to agents 监督学习、从模型到Agent
4 Multi-Agent 多智能体 7 Architecture, regime detection, LLM, risk control 架构、Regime、LLM、风控
5 Production 生产实战 5 Costs, execution, operations, projects 成本、执行、运维、实战

Plus 30 background articles and 4 appendices | 另有 30篇背景知识 和 4篇附录

#Core Architecture | 核心架构

Multi-Agent Architecture

#Target Readers | 目标读者

Reader Type What You'll Get
Developers → Quant Complete path to build trading systems
Quant Researchers Multi-agent architecture, production risk control
Investors/PMs Understand quant system capabilities and risks

Prerequisites:

  • Required: Basic programming (Python)
  • Helpful: Statistics, financial markets
  • Not needed: ML/DL background
读者类型 你会获得什么
程序员想转量化 从零构建交易系统的完整路径
量化研究员 多智能体架构、生产级风控设计
投资者/PM 理解量化系统的能力边界和风险

前置要求：

  • 必需：基本编程概念（Python）
  • 有帮助：统计学基础、金融市场常识
  • 不需要：机器学习或深度学习背景

#Repository Structure | 目录结构

ai-quant-book/
├── manuscript/
│   ├── cn/                      # 中文版 (Complete 已完成)
│   │   ├── Part1-快速体验/
│   │   ├── Part2-量化基础/
│   │   ├── Part3-机器学习/
│   │   ├── Part4-多智能体/
│   │   ├── Part5-生产与实战/
│   │   └── Resources & Links/
│   ├── en/                      # English (Complete)
│   │   ├── Part1-Quick-Start/
│   │   ├── Part2-Quant-Fundamentals/
│   │   ├── Part3-Machine-Learning/
│   │   ├── Part4-Multi-Agent/
│   │   ├── Part5-Production/
│   │   └── Resources-Links/
└── README.md

#Quick Start | 快速开始

#English Readers

→ Read English Version

Recommended Path:

Reader Path
Complete beginner Part 1 → All Part 2 → Part 3-5
With coding background Lesson 01 → Skim Part 2 → Parts 3-5
With quant background Lesson 01 → Lesson 08 → Parts 3-5

#中文读者

→ 阅读中文版

推荐路径：

读者类型 路径
零基础入门 Part 1 → Part 2 全部 → Part 3-5
有编程基础 第01课 → Part 2 快速浏览 → Part 3-5
有量化基础 第01课 → 第08课 → Part 3-5

#Risk Disclaimer | 风险声明

Quantitative trading involves risk. Invest carefully.

量化交易有风险，投资需谨慎。

This book is educational and does not constitute investment advice.

  • Strategies are for learning only, no profit guaranteed
  • Fully understand risks before live trading
  • Never trade with money you can't afford to lose
  • Past performance ≠ future results

本书是教育性质，不构成投资建议。

  • 书中策略仅供学习，不保证盈利
  • 实盘交易前请充分理解风险
  • 永远不要用你输不起的钱交易
  • 历史表现不代表未来收益

#License | 许可证

This work is licensed under CC BY-NC-SA 4.0.

本书内容采用 CC BY-NC-SA 4.0 协议。

"Whatever happens in the stock market today has happened before and will happen again." — Jesse Livermore

"We search through historical data looking for anomalous patterns that we believe are predictive of future price action." — Jim Simons

Nouvelle version disponible.