Categories
Author: Chris
Co-Head of Options Quantitative Research
Where he develops models to support high-frequency trading strategies. Before joining Susquehanna, he worked in algorithmic foreign exchange trading. Prior to starting a career in trading, he conducted research and taught mathematics in academia. Chris earned his Ph.D. in Mathematics from the University of Michigan, specializing in several complex variables and differential geometry, and continues to apply rigorous mathematical thinking to quantitative finance and machine learning.
Financial markets generate enormous amounts of data, but the effective signal is far smaller than raw counts suggest. Correlation across assets reduces independent information, regime changes limit how much historical data remains relevant, and noise further increases sample complexity. Together, these effects shrink billions of observations into a much smaller usable dataset. Understanding this constraint is critical for building effective models, emphasizing the importance of structure, denoising, and careful pooling across assets. Why massive financial datasets contain far less usable signal than they appear.
Read More
