
Alternative data (AltData) serves as a critical driver for competitive advantage in quantitative trading, requiring firms to move beyond traditional market data to incorporate real-world insights. Eric Mannes, a veteran at Jane Street, details the transition from ad-hoc data management to a centralized engineering effort that transforms messy, external information into structured, actionable inputs. This process demands deep domain expertise to navigate complex contract conventions, physical supply constraints, and metadata inconsistencies. While AI and large language models significantly accelerate feature extraction and productivity, they introduce risks like future data leakage, making rigorous human verification and high-quality data pipelines essential. Ultimately, the ability to clean, interpret, and integrate diverse, non-traditional datasets allows traders to identify unique market signals and maintain an edge in an increasingly efficient, adversarial environment.
Sign in to continue reading, translating and more.
Open full episode in Podwise