玄衡科技XUANHENG TECHNOLOGY
MENU

PROJECT 03 · QUANTITATIVE RESEARCH

Quantitative Research Platform

An internally developed quantitative research infrastructure.

Status: Internal Research System

FOCUS

  • Quantitative Research
  • Data Infrastructure
  • Research Engineering

01 · Overview

Overview

Building reproducible research infrastructure across data engineering, factor research, experimental validation and explicit execution boundaries.

The current system combines canonical data contracts, controlled ingestion and read-only access, an ETF universe, typed operators and formal factor research, with versioned baselines and validation reports preserving research evidence.

Backtesting, portfolio construction and execution remain separate stages of development; research output is not treated as a tradable conclusion.

02 · Capabilities

Capabilities

  1. 01

    Canonical Data Foundation

    Organising research data through shared contracts, controlled ingestion, publication checks and read-only access.

  2. 02

    Factor Research

    Maintaining typed numerical operators, formal factor definitions, registries and causality checks.

  3. 03

    Universe & Point-in-time

    Managing ETF scope, eligibility and human review with explicit timing and visibility constraints.

  4. 04

    Research Engineering

    Recording reviewable research through immutable baselines, validation reports and layered contracts.

03 · Workflow

Workflow

  1. 01

    Acquire & Validate Data

  2. 02

    Canonical Access & Alignment

  3. 03

    Universe & Factor Research

  4. 04

    Freeze Evidence & Deliver

04 · System View

System View

  1. 01

    Data Contract & Ingestion Layer

  2. 02

    Read-only Data Access Layer

  3. 03

    Universe, Operators & Factors

  4. 04

    Research Validation & Execution Boundary

05 · Current State

Current State

Implemented

Canonical data contracts, multi-dataset ingestion and read-only access, the ETF universe, formal Operator / Factor registries, and baseline validation are in place.

Current Focus

Improving daily and minute data paths, point-in-time visibility semantics, and consistency across research contracts.

Next Direction

Building backtest, portfolio and controlled execution loops only after timing, cost and tradability constraints are explicit.

06 · Principles

Principles

  1. 01

    Verify provenance, time alignment and data visibility before evaluating performance.

  2. 02

    Research definitions, run evidence and baselines must be reproducible and reviewable.

  3. 03

    Exploration, simulation and execution remain explicit boundaries.