Alternative Data Themes and Alpha Decay in Quantitative Trading

Summary

This conference trip report summarizes a talk about seeking trading signals in alternative data. Examples include satellite and drone imagery, purchase receipts, social media, industrial sensor data, agriculture, energy supply and demand, weather, and geographic monitoring. The central idea is that wider access to information and faster sharing of strategies can erode an edge, encouraging researchers to investigate less commonly used data sources.

The report suggests that machine learning, including deep learning, may help analyze large and varied datasets, and mentions exploratory ideas involving oil supply and demand and weather derivatives. It offers themes and examples rather than a tested research process: it provides no data specifications, signal construction, backtest, or performance evidence. The ideas therefore serve as a research agenda, with practical value dependent on data quality, analysis, and validation.

Key ideas

• Alternative data can include imagery, transaction receipts, social feeds, sensor readings, weather, and energy information.
• The report frames alpha decay as a reason to investigate less widely used data sources.
• Machine learning is presented as one possible way to process large and varied datasets.
• Oil supply and demand and weather derivatives are mentioned as exploratory strategy areas.
• The report supplies research themes but no tested signals or performance evidence.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.