About Out of Sample
Out of Sample is an independent English-language publication about quantitative and systematic investing. It is written for students, self-taught researchers and people starting out in the industry who want the reasoning and the references behind an idea, not a sales pitch.
The name
In quantitative research, "in sample" is the data used to build and tune a model, and "out of sample" is the data held back to test it. Most strategies that look brilliant in sample disappoint out of sample. The name is a reminder of the standard every article tries to apply: a result is interesting only if it survives a test it was not fitted to.
Coverage
- Systematic trading: designing rules, backtesting them in a spreadsheet or in code, automating execution, and the debates around high-frequency trading.
- Momentum and factors: the academic record on momentum, its crashes, timing overlays and the Fama-French models.
- Financial machine learning: labelling, meta-labelling, fractional differentiation and validation methods suited to market data.
- Risk and portfolios: value at risk, volatility targeting, regime models and combining strategies.
- Inside a hedge fund: prime brokers, custodians, auditors, capital introductions, seed capital, investor eligibility and contract clauses.
- Open research: open-source software and the practice of documenting a systematic research programme in public.
Editorial method
- References over opinions. Claims about returns, risk or regulation are tied to a peer-reviewed paper, a regulator's text, an official dataset or a software project's own documentation, linked in the article and listed at its foot.
- Examples are labelled. Where an article uses numbers to illustrate a method, it says whether they are hypothetical or taken from a cited study. We do not present invented backtests as real results.
- Rules carry a date. Regulations on investor eligibility, custody or audits differ by country and change over time. Articles state the jurisdiction they describe and the year of the rule.
- Collective byline. Articles are written and edited by the editors of Out of Sample and published under the publication's name. There are no individual bylines, invented experts or testimonials.
No investment advice
Out of Sample is educational. It does not know your circumstances and does not tell anyone what to buy, sell or hold. Strategies discussed here are described to explain how they work and how they have been studied; past results, simulated or real, do not predict future ones. The site carries no advertising, sponsored content or affiliate links, and receives nothing from any broker, course provider or software vendor it mentions.
About this domain
Out of Sample is a new publication, started in 2026 and run independently. The web address it uses was previously held by a different project: an older blog on quantitative finance and hedge funds. After that project stopped, the address passed through other hands and for a while carried unrelated pages. Out of Sample has no link with any of those earlier owners or authors. It does not reuse their name, logo, articles or images, and it does not speak for anyone who wrote there.
Because the earlier blog covered similar ground, some of our article addresses deliberately match topics readers were looking for, such as momentum, meta-labelling or how prime brokers work. The articles at those addresses are new texts, researched and written for this site.
Corrections
If an article contains an error, a broken reference or a rule that has since changed, the editors correct it and record the date of the update at the top of the article. The legal notice describes how requests reach the editors, and the contribute page explains how to suggest a topic or a source.