Independent quantitative research · Global equitiesATLASVECTOR / METHODOLOGY
03 / METHODOLOGY

A transparent path from question to evidence.

Our methodology emphasizes data lineage, reproducibility, realistic assumptions and clear separation between research results and investment decisions.

RESEARCH PIPELINE

Six-stage validation framework

Question

Define the economic intuition and falsifiable hypothesis.

Universe

Set eligibility, liquidity and survivorship rules.

Data

Validate sources, timestamps, adjustments and missing values.

Model

Specify signal, sizing, rebalance and constraint logic.

Simulation

Include costs, delays, capacity and realistic execution assumptions.

Review

Evaluate robustness, failure modes and operational limits.

VALIDATION CONTROLS

Controls against false confidence

01

Look-ahead controls

Information is aligned to when it was actually available.

02

Survivorship controls

Delisted and inactive securities are retained where the study requires them.

03

Cost assumptions

Fees, spreads, slippage and turnover are explicitly modeled.

04

Robustness checks

Parameters, sub-periods, markets and regimes are challenged.

Important limitation

Backtests and simulated results are hypothetical. They do not represent actual trading and cannot guarantee future performance. Research conclusions remain conditional on data quality and model assumptions.