CPG - Institutional Premium Arbitrage System

Concept & Logic:
This strategy captures institutional sentiment by analyzing the Cross-Exchange Arbitrage Data between Coinbase (USD pair) and Binance (USDT pair). Instead of using raw price difference which is noisy, this script employs a Proprietary Dynamic Threshold Algorithm. It normalizes the premium data using a custom volatility-adjusted window to filter out retail noise and identify genuine "Whale Accumulation" zones.

Key Features:
Data Source: Real-time BTC/USD vs BTC/USDT spread analysis.
Signal Filtering: The proprietary algorithm (closed-source logic) dynamically adjusts upper and lower bands to prevent false signals during low liquidity periods.
Execution:

Bullish: When the premium breaks the dynamic upper threshold (Strong Institutional Buying).
Bearish: When the premium drops below the dynamic lower threshold (Institutional Selling).

Usage:
Note: The dynamic threshold algorithm is specifically calibrated for Bitcoin's unique liquidity structure. Extensive backtesting shows that this logic is NOT suitable for altcoins (like ETH or SOL). Please strictly use it on BTC pairs.

策略核心：
本策略透過分析 Coinbase (USD) 與 Binance (USDT) 之間的跨交易所資金流 (Arbitrage Data)，來捕捉機構投資者的動向。 原始的價差數據通常充滿雜訊，因此本腳本內建了一套**「獨家動態閥值演算法」**。該算法能對數據進行平滑處理與正規化，有效過濾市場雜訊，精準識別出機構大戶的資金流向。

功能特點：
數據源： 即時運算 BTC/USD 與 BTC/USDT 的溢價差。
獨家過濾： 閉源的動態演算法會根據波動率自動調整上下軌閥值，避免假突破。
交易訊號：

看多： 溢價突破動態上軌（機構強力買入）。
看空： 溢價跌破動態下軌（機構拋售）。

用法：
注意： 本策略的動態閥值演算法是針對比特幣的流動性結構進行嚴格校準的。回測數據顯示，此邏輯不適用於 ETH 或 SOL 等其他幣種。請務必僅在 BTC 圖表上使用。