Multi-Chain State Modeling

1. Data Input Dimensions

TradingRazor integrates multi-dimensional data inputs in a unified framework, including:

Market Microstructure Data: order book depth, spread structures, bid-ask distribution, transaction rhythm, and potential impact costs.

Contract and Risk States: funding rates, margin utilization, liquidation density, and leverage limits.

On-Chain Behavioral Data: whale transfers, cross-chain capital migration, address clustering behavior, and changes in capital concentration.

Macro and Sentiment Data: macro event proximity, long-short sentiment indicators, and volatility regime identification.

Through multi-dimensional data integration, the system can accurately perceive market conditions across different time scales, forming a high-precision foundation for decision-making.

2. Multi-Timeframe Windows

The system maintains:

Minute-Level Windows: capturing microstructure shocks and high-frequency events.

Hourly-Level Windows: volatility pattern analysis and mid-term strategy decision-making.

Daily-Level Windows: trend structure analysis and long-term risk assessment.

Longer Cycles: macro regime states used for dynamic strategy weight switching.