Research methodology
How research is done here.
Every study published by Quant Investing Lab follows the same five-step process. The same one a rigorous hedge fund research desk would follow before approving a strategy.
The process
Five steps. No exceptions.
This process governs the research library — the original quantitative studies. The shorter Notes are not backtests and do not run through it; what binds them is the sourcing standard below, and each one carries its sources at the foot of the page.
01
Hypothesis
Every research note begins with a clearly stated hypothesis traceable to academic literature. We do not explore data without a prior question, because that path leads to data-snooping bias and false discoveries.
02
Data
Data sources are disclosed on every study — FRED, BLS, the Conference Board, and public market data among them. Dataset construction (survivorship corrections, point-in-time data, look-ahead prevention) is described in plain language, never assumed.
03
Test
Every strategy is tested on in-sample and out-of-sample periods. Walk-forward analysis is the default. Costs, slippage, and execution assumptions are disclosed before results. No parameter optimization at scale.
04
Stress
Results are stress-tested across regimes: pre-2000, dot-com bust, 2008 crisis, 2010s low-volatility, 2020 pandemic, 2022 inflation shock. A strategy that only works in one regime is not a strategy. It is an artifact.
05
Publish
Every research paper states: hypothesis, data source, methodology, results, caveats, and open questions. Readers audit the reasoning. If the methodology fails scrutiny, the paper is revised.
Our commitments
Five promises to the reader.
We disclose before we conclude
Every assumption, parameter, and data choice is stated at the top of every study. No conclusion is presented without its inputs.
We cite to the primary source
Models and formulas trace back to a peer-reviewed paper or a named practitioner. Statutory figures trace back to the agency that sets them, not to another site quoting them. No anonymous authority.
We publish net of costs
All performance metrics are reported after realistic transaction costs, slippage, and bid-ask spreads. Pre-cost performance is flagged as illustrative.
We note what we don't know
Every study has a 'Limitations' section identifying what was not tested, what data was not available, and what assumptions could break the conclusion.
We refuse conflicts of interest
No sponsored content. No affiliate picks disguised as research. If we benefit financially from a product we discuss, the conflict is disclosed at the top of the article.
Editorial process
Human authorship, AI-assisted research.
Articles published by Quant Investing Lab are written by John Bergerat, founder of the Lab and holder of an MSc in Quantitative Finance.
AI tools play a supporting role. They accelerate source discovery across academic papers, regulatory filings, and institutional research, where manual literature review is slow and tedious, and they assist with drafting and copy-editing. What they produce is a starting point, not the published text.
Quantitative finance is a specialized field. AI models are generalists. They can gather information and shape a draft, but they lack the domain judgment to evaluate it: which source is rigorous, which assumption matters, which data pitfall can break a result. That judgment, the verification of every fact and figure, and the final text are John's.
AI is used as a research and drafting tool, not as a co-author. The accountability is not delegated.
What we don't do
The line we don't cross.
Quant Investing Lab is an educational research publication. It is not an investment advisor, a signal service, or a portfolio manager. Specifically:
- We do not publish “buy” or “sell” recommendations.
- We do not issue trade signals, whether live, delayed, free, or paid.
- We do not manage anyone's capital.
- We do not offer personalized investment advice.
- We do not claim our backtests predict future results.
What we do is publish research that helps serious individual investors understand how quantitative strategies work and what questions to ask of anyone who claims to run one.
Last updated: April 21, 2026