Yiheng Quant Tech

01 / System framework

Core Quant System

A modular research-to-execution infrastructure for quantitative investing.

Integrating data, research, models and execution into a unified system. Each module has a clear purpose, well-defined interfaces, and can be developed and evaluated independently.

  1. Data
  2. Context
  3. State
  4. Setup
  5. Position
  6. Portfolio
  7. Execution
MODULE / 01Research infrastructure

Data Layer

What it does

  • Collects and standardizes multi-source market data
  • Maintains local-first historical datasets and caches
  • Supports whole-market research workflows

Key capabilities

  • A-share daily market data
  • ETF / index inputs
  • Selected global market inputs
  • Data validation
  • Refresh workflow
  • Universe filtering

Output

Clean standardized research data · Validated universes · Reusable historical datasets

MODULE / 02Research infrastructure

Macro / Market Context

What it does

  • Collects China macro evidence and market context
  • Synthesizes liquidity, policy, growth and risk conditions
  • Provides shared context to downstream modules

Key capabilities

  • Macro evidence collection
  • Daily market context
  • Market-state classification
  • Shared market context
  • Cross-asset / risk backdrop

Output

Regime / context classification · Macro features and market backdrop

MODULE / 03Research infrastructure

Signal & State Engine

What it does

  • Classifies stock character and technical state
  • Tracks market activity states
  • Filters and ranks actionable setups

Key capabilities

  • Stock character classification
  • Technical state
  • Volume / volatility / trend structure
  • Activity-state classification
  • Setup classification
  • Opportunity ranking
  • Catalyst / event context where available
dead_bearishdead_quietcompressed_readyexpandingoverheated

Output

Structured signal states · Opportunity list · Setup classification · Confidence / ranking context

MODULE / 04Decision infrastructure

Position State Machine

What it does

  • Separates stock selection from position management
  • Manages position lifecycle through explicit states
  • Controls entry, hold, warning and exit decisions

Key capabilities

  • State transitions
  • Thesis validity checks
  • Price / signal / risk interaction
  • Re-entry logic
  • Position-level decision rules
BUY
HOLD
WARNING
EXIT
← Re-entry after renewed validation

Thesis validity · Signal state · Risk checks

Output

Target position state · Trade intent · Hold / reduce / exit decisions

MODULE / 05Decision infrastructure

Portfolio Engine

What it does

  • Combines new opportunities with existing holdings
  • Compares positions on a relative basis
  • Determines replacement and target allocation

Key capabilities

  • Opportunity ranking
  • Position sizing
  • Exposure control
  • Concentration awareness
  • Existing-holding comparison
  • Portfolio replacement logic

Output

Target portfolio · Target weights · Replacement decisions · Exposure summary

MODULE / 06Decision infrastructure

Risk & Execution

What it does

  • Converts portfolio decisions into executable trades
  • Applies pre-trade risk and order checks
  • Monitors orders and positions

Key capabilities

  • Trade guard
  • Order validation
  • Broker execution interface
  • Market / limit orders
  • Cancel / order status
  • Position monitoring
  • Simulated and live workflows

Output

Executable trade intents · Order / execution status · Monitored portfolio state