OPEN-SOURCE SCRIPT

Multi-Crypto Principal Component Analysis

93
Version 0.2

## 📌 Multi-Crypto Principal Component Analysis (PCA) — Indicator Summary

### 🎯 Purpose
This indicator identifies **cryptocurrency assets that are behaving differently** from the rest of the market, using a simplified approach inspired by Principal Component Analysis (PCA). It’s designed to help traders spot **cross-market divergences**, detect outliers, and improve asset selection and correlation-based strategies.

### ⚙️ How It Works
The indicator analyzes the **log returns** of up to 7 user-defined assets over a configurable lookback period (default: 100 bars). It computes the **z-score** (standardized deviation) for each asset’s return series and compares it against the average behavior of the group.

If an asset’s behavior deviates significantly (beyond a threshold of 1.5 standard deviations), it’s flagged as an **outlier**.

- Each outlier is plotted as a **colored dot horizontally spaced** above the price bar
- Up to **3 dots per bar** are shown for visual clarity

This PCA-style detection works in real time, directly on the chart, and gives you a quick overview of which assets are breaking correlation.


### 🔧 Inputs
- 🕒 **Lookback Period**: Number of bars to analyze (default: 100)
- 🔢 **Assets 1–7**: Choose any 7 crypto symbols from any exchange
- 🎨 **Colors**: Predefined per asset (e.g. BTCUSDT = red, ETHUSDT = yellow)
- 📈 **Threshold**: Internal (1.5 std dev); adjustable in code if needed

### 📊 Outputs
- 🟢 Dots above candles representing assets that are acting as outliers
- 🧠 Real-time clustering insight based on statistical deviation
- 🧭 Spatially spaced dots to avoid visual overlap when multiple outliers appear

### ⚠️ Limitations
- This is a **PCA-inspired approximation**, not true matrix-based PCA
- It does **not compute principal components or eigenvectors**
- Sensitivity may vary with asset volatility or sparse trading data
- Real PCA requires external tools like Python or R for full dimensional analysis

This tool is ideal for traders who want real-time crypto correlation insights without needing external data science platforms. It’s lightweight, fast, and highly visual — and gives you a powerful lens into market dislocations across multiple assets.

Declinazione di responsabilità

Le informazioni ed i contenuti pubblicati non costituiscono in alcun modo una sollecitazione ad investire o ad operare nei mercati finanziari. Non sono inoltre fornite o supportate da TradingView. Maggiori dettagli nelle Condizioni d'uso.