All articles
The full archive of Out of Sample, grouped by topic. Each article explains a method, a piece of research or an institution, with its sources at the foot of the page. For a guided path, start with the open research series.
All articles are educational and none is investment advice.
Systematic trading
How rule-based trading systems are designed, tested, automated and run, from a spreadsheet crossover to fully automated execution.
- Building a simple moving average crossover strategy in Microsoft ExcelColumns, formulas and pitfalls for a moving-average crossover backtest built entirely in a spreadsheet.
- Getting started building a fully automated trading systemData, signals, risk, orders and monitoring: the parts of an automated trading system and how they fail.
- Inside the black box: the parts of a quantitative trading systemAlpha, risk and cost models, portfolio construction and execution: how a quant system is put together.
- QuantInsti and the EPAT programme: what an algorithmic trading course covers, and how to evaluate oneWhat QuantInsti's EPAT programme is, and a checklist for judging any algorithmic trading course.
- Resistance to high-frequency trading: the debate, the evidence and the market designs it producedFlash Crash, speed bumps and batch auctions: the case against HFT, the case for it, and the evidence.
- Why people trade forex, and what the numbers say about retail resultsThe largest market in the world, the strategies studied in it, and what regulators found about retail traders' results.
Momentum & factors
Cross-sectional and time-series momentum, dual momentum, momentum crashes and the factor models used to explain returns.
- Market timing models for a momentum strategyMoving-average filters, market states and regime models as overlays for momentum, and what they cost.
- Momentum crashes: why winners-minus-losers can lose so much so fastRare, sudden and concentrated in rebounds: what drives momentum's worst losses and how researchers respond.
- Momentum on dual momentum portfolios: rotating between whole strategiesDual momentum across asset classes, and the idea of rotating between whole portfolios by their own momentum.
- Momentum strategies: what the evidence says about buying winnersHow momentum portfolios are built, what three decades of research found, and what erodes the paper returns.
- Optimal stock quantity, selection and weights for momentum investingHow many winners to hold, which ones, and how to weight them: diversification versus signal strength.
- The Fama-French five-factor asset pricing modelMarket, size, value, profitability and investment: how the factors are built and how to use them to test a strategy.
Financial machine learning
Labelling, meta-labelling, fractional differentiation and the research tools that grew up around machine learning for markets.
- A laboratory for machine learning in finance: bars, labels and honest validationDollar bars, triple-barrier labels, purged cross-validation and feature importance: the pipeline step by step.
- Advances in Financial Machine Learning: a map of the techniques and the tools that implement themThe book's five parts and twenty-two chapters, which techniques matter most, and how implementations evolved.
- Does meta-labelling add to signal efficacy? How to test it honestlyBaselines, metrics and validation for judging whether a meta-labelling layer really improves a signal.
- Dreaming, thinking machines: from automata to algorithms in the marketsAutomata, Descartes, Turing, ELIZA and the Chinese room: what the history of machine intelligence says about markets.
- Fractional differentiation: stationary features that keep their memoryDifferencing by a fraction: how to make a price series stationary while keeping as much memory as possible.
- Meta-labelling: a toy example to build intuitionA primary model chooses the side, a secondary model decides whether to trust it: a worked, hypothetical example.
Risk & portfolios
Value at risk, volatility targeting, regime models and how diversified strategies are combined into a portfolio.
- An introduction to value at riskHistorical, parametric and Monte Carlo VaR, how to backtest it, and why regulators moved to expected shortfall.
- Choosing your risk: volatility targets, drawdown limits and position sizingVolatility targets, Kelly sizing, drawdown limits and leverage: deciding how much risk a strategy should run.
- Combining diversified alpha to deliver a superior Sharpe ratioThe arithmetic of combining uncorrelated strategies, the fundamental law of active management, and its traps.
- Hidden Markov models for market regimes: an introductionMarkov chains, hidden states and the forward, Baum-Welch and Viterbi algorithms, applied to market regimes.
Inside a hedge fund
Prime brokers, custodians, auditors, seed capital and the investors and contracts that surround a fund.
- Accredited investors and qualified purchasers: who may invest in a hedge fundIncome, net worth and professional tests: the rules that decide who may invest in private funds.
- Asset custodians: who actually holds a fund's securitiesSafekeeping, settlement and segregation: the institutions that hold securities on behalf of funds.
- Bridging the gap into hedge funds: roles, skills and routes inQuant researcher, developer, trader, risk and operations: the skills each role needs and how people get in.
- Capital introductions: how prime brokers connect hedge funds with investorsHow prime brokers introduce funds to investors, the marketing rules that apply, and realistic expectations.
- Clawback clauses: taking back pay and profits that were not really earnedGP clawbacks, high-water marks, malus and the SEC's executive pay recovery rule, explained.
- Family offices: how wealthy families manage money, and why they matter to hedge fundsSingle- and multi-family offices, how they invest, how they are regulated, and the Archegos debate.
- Fund auditors: what a hedge fund audit checks, and what it cannotWhat the annual audit of a fund covers, the rules that require it, and why auditor quality became a red-flag check.
- Prime brokers: what they do for hedge funds and why it matters who holds the assetsFinancing, securities lending, custody and reporting, and the lessons of Lehman and Archegos.
- Seed capital for hedge funds: how seeding deals workWhy first money is hard to raise, who seeds new managers, and the revenue shares and lock-ups that come with it.
Open research
Open-source tools for quantitative finance and the idea of researching a systematic fund in public.
- Creating a systematic strategy step by step, in the openFrom hypothesis to monitoring: a nine-stage research workflow with the checks that keep it honest.
- Getting started with open source for quantitative financeLanguages, core libraries, backtesting frameworks and data sources, and habits that keep research reproducible.
- Running a research project in the open: a practical guideRepository layout, licences, reproducibility, review and governance for an open quantitative research project.
- What is an open-source hedge fund?Open research, crowdsourced platforms and tournaments: what openness can and cannot mean for a fund.
- mlfinlab on the PyPI index: packaging financial research code, and what happened nextHow research code is packaged and published on PyPI, and what the mlfinlab licence change teaches about dependencies.