OU statistical arbitrage
An Ornstein–Uhlenbeck framework for examining cross-asset spread relationships, mean reversion and the stability of pair selection.
A map of the research: the market question, the technical approach and the limits of the evidence. These are studies and archived systems—not a list of strategies offered to investors.
An Ornstein–Uhlenbeck framework for examining cross-asset spread relationships, mean reversion and the stability of pair selection.
A multi-system extension of cross-asset mean reversion. The archive allows comparison of configurations without treating a retrospective replay as live performance.
A study of how listed closed-end funds trade relative to their reported net asset values, and whether apparent convergence survives realistic implementation.
Relative-value research into two share classes of the same issuer, including liquidity, voting rights and the practical cost of holding both legs.
Research into time-series momentum across asset classes, with volatility-aware sizing and explicit attention to market exposure.
An examination of currency carry alongside trend conditioning: a risk-premium question as well as a timing problem.
A study of market behaviour following unusually large price moves, where timestamp quality and event definition matter as much as the predictor.
Research into information in futures-curve shape, roll effects and the difference between a quoted contract and an investable return series.
An archived intraday index-system project examining opening-range structure, session definitions and execution constraints.
A study of rolling evaluation for intraday index models, examining how parameter selection changes across market conditions.
An archived gold-system project combining forecasting and risk controls. Later replay work revisited alternative historical configurations.
Research into directional-change and risk overlays on gold, separating an overlay’s contribution from the underlying market move.
An archived multi-sleeve research system combining macro-regime and trend inputs with portfolio-level risk allocation.
Cross-sectional research into relative equity performance, with an emphasis on point-in-time membership and investable comparisons.
An examination of low-volatility equity portfolios and the distinction between factor exposure, selection and incremental return.
Research into calendar structure in equities, including turn-of-period behaviour and the risk of selecting the attractive dates after the fact.
A study of equity valuation signals using fundamental information, where reporting lags and universe construction are part of the signal definition.
A symbolic-regression research project exploring machine-generated equity expressions and the challenge of evaluating an adaptive search.
Research into order-flow imbalance and limit-order-book representations, connecting microstructure prediction with the cost of an executable trade.
A follow-on order-flow project examining whether the available information can clear an execution-cost hurdle before model complexity is added.
A futures microstructure study focused on event timing, the order book and realistic interpretation of short-horizon signals.
A forecasting study connecting population-weighted weather measures with US natural-gas storage. Forecast accuracy and trading value are separate questions.
Research into shipping records as company information, including entity mapping, publication delay and the boundary between a pipeline test and market evidence.
A study of reported futures positioning and what can still be inferred after accounting for the delay between positions and public release.
Research into post-earnings-announcement drift, linking surprises to subsequent returns under point-in-time data and cost assumptions.
A study of text-derived macro information and its potential role as a market feature or risk indicator.
An archived event-driven gold research project examining political text and price behaviour. Price-only proxies are not equivalent to the original information feed.
A digital-asset cross-sectional research design examining the less-liquid tail, with explicit venue, survivorship and execution questions.
A machine-learning project for the Numerai setting, where raw predictive correlation and marginal contribution are distinct evaluation objectives.
A research-generation workflow for screening candidate mechanisms and keeping the search history visible to subsequent evaluation.
A genetic-algorithm research project exploring model combinations and the distinction between search fitness and out-of-sample evidence.
Research connecting a forecast to position construction and execution—the steps that sit between a prediction and an implementable system.
A foreign-exchange study combining cross-currency strength and trend information, with turnover and sizing as first-class concerns.
Try a broader term or clear the domain filter.
Editorial review · A conceptual reading sequence, separate from the studies and archived systems above.
The research profiles describe the questions. The engineering page explains how data, experiments and review fit together.
Inside the stack