---
title: '【开源解读】券商金工复现框架'
source: 'https://www.youtube.com/watch?v=yRCv1PQMuOQ'
video_id: 'yRCv1PQMuOQ'
date: 2026-09-15
duration_sec: 81
channel: '量化Quantgirl'
---

# 【开源解读】券商金工复现框架

> Source: [【开源解读】券商金工复现框架](https://www.youtube.com/watch?v=yRCv1PQMuOQ)

## Summary

This video introduces an open-source project that reproduces over 100 quantitative trading strategies from top Chinese brokerages, including Guangda, Huatai, and China Merchants. The project, created by Yugo 2046 and hosted on GitHub, organizes strategies into four categories: timing, factor construction, quantitative value, and portfolio optimization. It provides code, original research PDFs, and notebooks to help users understand and apply these strategies.

### Key Points

- **Project Overview** [00:00] — The open-source project reproduces 100 quantitative strategies from top brokerages, published on GitHub by Yugo 2046, with over 5,400 stars.
- **Strategy Categories** [00:13] — Strategies are divided into four main categories: timing, factor construction, quantitative value, and portfolio optimization. The project is paper-style, replicating research reports from brokerages like Guangda, Huatai, and China Merchants.
- **Tech Stack and Quick Start** [00:25] — The project uses Python 3.8 and the README provides clear quick-start links, including project features, strategy statistics, and core dependencies.
- **Setup and Dependencies** [00:38] — To run, clone the repository with Git, install core dependencies including Pandas, NumPy, Qtlib, and Backtrader. Data sources can be optional: JoinQuant, QData, or Toucher Pro.
- **Project Structure** [00:52] — The repository is organized by strategy type into folders like timing, factor, value, and portfolio. Each strategy directory includes reproduction code and the original research report PDF for reference.
- **Educational Value** [01:06] — The main value is validating research ideas and learning quantitative trading. Included notebooks and visualizations help understand backtesting logic, and factor modules can be reused directly.

### Conclusion

This project is a valuable resource for quantitative trading enthusiasts, offering a comprehensive collection of reproduced strategies with original research materials, making it ideal for learning and validation.

## Transcript

跑通券商惊弓演爆吗? 这个开源项目复现了 100家个量化策略 作者Yugo 2046把它发布在Dithub上 目前已获超5400个新标
这篇论文式的项目 复现了光大华泰招商 等顶级券商的惊弓演爆 作者将策略分为 择时因子构建 量化价值和组合优化
四大板块 技术站涵盖Path 3.8 看项目首页Readme顶部 就给出了清晰的快速入口 你可以看到项目特色策略 数量统计以及核心依赖的说明
和上手运行 先用Gitclone拉去仓库 核心依赖安装 包括Pandas, NumPack, Qtlib和Backtrader 数据源可选聚宽J, Qdata或Toucher Pro
从仓库的文件数可以看到 项目结构基于策略类别来组织 分为则是因子 价值和组合等文件夹 每个策略目录里不仅包含复先代码 还附带了对应的原始研报PDF
供对照学习 仓库最大的价值在于 验证研报思路和量化入门策略 附带的notebook和可视化 能帮你快速理解回测逻辑 也能直接附用其中的因子模块