ASR Research & Analyst Operating Model

The package is designed so that Research and Analyst teams can share one evidence chain instead of exchanging disconnected notebooks.

Responsibilities

Analyst team - collect and document data; - inspect market structure and descriptive diagnostics; - generate candidate observations; - map definitions and market conventions; - build figures/tables and interpret economic meaning; - document data limitations and availability lags.

Research team - formalize the question and falsifiable hypothesis; - map the closest literature and prior art; - select models and baselines; - derive/implement methods; - perform statistical testing, calibration and robustness; - decide which claims are supported.

Independent reviewer / reproducibility reviewer - rerun the work from clean inputs; - inspect look-ahead, leakage and specification risk; - challenge the novelty statement; - verify formulas, conventions and limitations; - approve/reject the public claim, not merely the code execution.

Shared A-to-Z API

import asrquant as asr

# 1. Discover
board = asr.discovery.weekly(data=data, domain="fixed_income", n=10)

# 2. Select and formalize
project = board.start(0)

# 3. Run the existing ASRQuant research workflow
# project.plan_data(...)
# project.attach_data(...)
# project.features(...)
# project.test_hypothesis(...)
# project.backtest(...)
# project.robustness(...)
# project.decide(...)

# 4. Package the Friday-to-Friday publication
cycle = asr.weekly_cycle(board, 0, launch_friday="2026-08-14")
cycle.publication_pack("WR-001")

The project manifest and publication pack are the hand-off between teams. The package does not replace scientific judgment or approval.