Newest papers

Where Backtests and Live Trading Part Ways: Six Defects from One Intraday System

Alex Vidovich, Zivana Zerjal · 2026 · arXiv (Cornell University)

DOI: 10.48550/arxiv.2610.21846 · cited by 0 works (OpenAlex) · open-access PDF

Spanish

Los backtests de las estrategias intraday pueden desplazarse de lo que realmente hace el sistema en vivo; en nuestro caso, el drift nunca se mostró en estadísticas resumidas. Informamos de seis defectos encontrados durante la ejecución de un sistema de acciones estadounidense en vivo y comparando sus días en vivo con su replay backtest. Tres provienen de datos: un feed predeterminado que llevaba una pequeña fracción del volumen de apertura real, precios históricos ajustados para dividendos que las órdenes en vivo nunca vieron, y datos que el sistema en vivo descargó pero el backtest nunca hizo. Dos provienen de código: un paso de dimensionamiento que superó un límite que la cuenta en vivo impone, y las marcas de tiempo UTC leídas como el tiempo de Nueva York. Uno proviene de la configuración: configuraciones que se activaron

Machine translation (Meta M2M100); the English original follows.

Abstract

Backtests of intraday strategies can drift away from what the live system actually does; in our case, the drift never showed in summary statistics. We report six defects found while running one intraday US equity system live and comparing its live days with their backtest replay. Three came from data: a default feed that carried a small fraction of real opening volume, historical prices adjusted for dividends that live orders never see, and data the live system downloaded but the backtest never did. Two came from code: a sizing step that exceeded a limit the live account enforces, and UTC timestamps read as New York time. One came from configuration: settings that were switched on but never reached the code that places orders. For each we give the symptom, the size of the error, the cause and the test or control that now guards against it. All six sat where two paths that should agree had drifted apart. We close with a ten-point checklist. We make no claim about whether the system is profitable.

Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese

Topics: Equity (law), Computer science, Economics, Timestamp, Limit (mathematics), Sizing

Is this paper real? This page was built from OpenAlex and arXiv records, not generated. Software can verify any reference in one call: GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.21846, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txt

Sources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange