Chan Theory Fractals Quantified — Step-by-Step Guide & Transcript

以前嗤之以鼻，现在逐字学习！缠论量化代码大公开！| 全网独家！用Python量化缠论，一键自动生成交易信号 | 缠论基础知识讲解【量化投资邢不行啊】

0h 09m video Published May 9, 2024 Transcribed Sep 12, 2026 量化投资邢不行啊

5.1K views Recent velocity 0.0 views/hour View full performance history →

Intermediate 5 min read For: Traders and quantitative analysts interested in technical analysis, particularly those familiar with A-share markets and basic Python programming.

AI Trust Score 65/100

⚠️ Average / Some Fluff

"Delivers a solid, practical introduction to quantifying Chan Theory fractals, though the title's promise of 'quantifying Chan Theory' is only partially fulfilled (covers only the fractal layer)."

AI Summary

This video introduces the concept of 'Chan Theory' (禅论), a technical analysis framework popular among Chinese A-share investors, and demonstrates how to quantify its core 'fractal' (分形) concept using Python. The host, Xing Zhixing, explains the theory's origins, simplifies K-line processing rules, and provides code to automatically identify top and bottom fractals across all A-share stocks.

Chapters

1 Introduction to Chan Theory 00:00 2 K-line Processing in Chan Theory 01:42 3 Automating K-line Processing with Python 04:19 4 Fractal Definition and Detection 06:48 5 Conclusion and Next Steps 08:35

[00:00]

Introduction to Chan Theory

The video aims to quantify Chan Theory, a unique technical analysis system in A-share investing, acknowledging that interpretations vary among practitioners.

[00:37]

Origin and Author

Chan Theory was created by 'Chan Zhong Shuo Chan' (likely a pseudonym) and published on June 7, 2006, based on articles from the stock forum 'Gu Shu Li'. It claims applicability to any market with price fluctuations, including stocks, warrants, and futures.

[01:03]

Mathematical Claims

The theory claims to be the only trading theory that can be proven using mathematical methods, deriving price structure from geometric principles. The host expresses skepticism about this claim.

[01:16]

Author's Disappearance and Identity Rumors

The author stopped posting in 2008 and never reappeared. Rumors suggest he was Li Biao, the trader behind Yian Technology, the first 100-yuan stock on the ChiNext board, but this is unverified.

[01:42]

Four-Layer Structure of Chan Theory

The theory is divided into four layers: fractals (分形), strokes (笔), line segments (线段), and central pivot (中枢), which together form the basis for buy/sell signals. This video focuses only on the fractal layer.

[02:27]

Chan Theory K-lines Differ from Standard

In Chan Theory, K-lines are simplified to only include the highest and lowest prices, ignoring open, close, and shadows, which are considered irrelevant.

[02:52]

Inclusion and Non-Inclusion Relationships

Two adjacent K-lines can have either a trend relationship (non-inclusion) or no trend relationship (inclusion). Inclusion occurs when one K-line's high and low are both within the other's range.

[03:04]

Processing Inclusion Relationships

To simplify analysis, inclusion relationships are eliminated by merging two K-lines into one. The rule is 'rising takes high, falling takes low' (上升取高下降取低).

[04:19]

Manual Charting is Inefficient

Over 90% of Chan Theory traders manually draw on charts, which is time-consuming, especially for stocks with long histories. The host provides Python code to automate K-line processing.

[04:46]

Provided Data and Code

The host offers a dataset containing daily OHLC data for all 5,000+ A-share stocks, including delisted ones, and a Python script to process and plot Chan Theory K-lines.

[05:22]

Technical Indicator Testing Package

A bonus package includes historical data, a table of 125 technical indicators with formulas and usage, a coding guide for beginners, and Python code to calculate and test indicators.

[06:36]

Result: Clean K-line Chart

After processing, all K-lines have clear rising or falling relationships, matching Chan Theory's requirements. The chart is simplified and contains only essential price data.

[07:07]

Defining Fractals

A fractal consists of three adjacent K-lines that show a reversal pattern, not a continuous trend. There are top fractals (顶分形) and bottom fractals (底分形).

[07:34]

Significance of Fractals

A bottom fractal signals a potential price rebound and is a buying/accumulation signal; a top fractal signals a potential peak and is a selling signal. These are foundational for more advanced buy/sell points.

[07:58]

Automated Fractal Detection

The provided Python code can identify all historical top and bottom fractals for any stock, marking them on the chart (red for top, green for bottom).

[08:35]

Acknowledgment of Subjectivity

The host reiterates that Chan Theory is open to interpretation, and this is just one quantification approach. He invites discussion and feedback.

[09:00]

Next Video Teaser

If the video gets 5,000 likes, the next video will cover the next layer of Chan Theory: strokes (笔), which are built on fractals.

Mentioned in this Video

Python

tool

A-share historical data (all stocks)

tool

Technical indicator testing package

tool

Chan Zhong Shuo Chan (author)

person

Li Biao (rumored identity)

person

Tutorial Checklist

1 02:27 Understand that Chan Theory K-lines only use high and low prices, ignoring open, close, and shadows.

2 02:52 Identify inclusion relationships between adjacent K-lines (one K-line's high and low are within the other's range).

3 03:04 Merge inclusion K-lines using the rule 'rising takes high, falling takes low' to eliminate all inclusions.

4 04:46 Use the provided Python code and A-share data to automatically process K-lines and plot the simplified chart.

5 07:07 Define fractals as three adjacent K-lines showing a reversal pattern (top or bottom).

6 07:58 Run the provided code to identify all historical top and bottom fractals for any stock.

💡 Key Takeaways

📊

Mathematical Proof Claim

Highlights the theory's bold claim of being the only mathematically provable trading theory, which is a key selling point and source of debate.