Waves | Susquehanna's Technical Blog
Despite the apparent scale of financial data, its effective size is sharply limited by correlation, nonstationarity, and noise. Cross-asset correlations reduce dimensionality, meaning thousands of securities often behave like a much smaller set of factors. Regime changes further compress usable data by making older observations less relevant. Noise compounds the problem by increasing the number of samples required for reliable estimation. This post outlines how these forces interact and explores practical responses, including pooling across assets, denoising techniques, and incorporating structural assumptions. In finance, success depends less on data volume and more on extracting signal from a constrained and evolving dataset. Why massive financial datasets contain far less usable signal than they appear.
Read More
Correlation, Noise, and the True (Statistical) Size of Financial Data
Despite the apparent scale of financial data, its effective size is sharply limited by correlation, nonstationarity, and noise. Cross-asset correlations reduce dimensionality, meaning thousands of securities often behave like a much smaller set of factors. Regime changes further compress usable data by making older observations less relevant. Noise compounds the problem by increasing the number of samples required for reliable estimation. This post outlines how these forces interact and explores practical responses, including pooling across assets, denoising techniques, and incorporating structural assumptions. In finance, success depends less on data volume and more on extracting signal from a constrained and evolving dataset. Why massive financial datasets contain far less usable signal than they appear.
Read More
Categories
Risk management is central to quantitative trading, requiring careful handling of market, model, and execution risks. This post covers key concepts including systematic vs. idiosyncratic risk, hedging with options, and managing exposure across large portfolios. It also explores practical tools like position limits, volatility filters, and flow analysis, along with the impact of slippage and price impact. Effective risk management balances precision with adaptability in uncertain environments. How quant traders identify, measure, and manage risk in complex systems.
Read More
August 27, 2026 by Lyubo
Investing in early quant education requires upfront cost but delivers long-term gains. Structured training slows initial productivity but leads to steeper growth over time, provided employees stay long enough to realize the benefits. At Susquehanna, scale, retention, and collaboration make this investment worthwhile. Education builds shared foundations, accelerates skill development, and strengthens connections across teams, ultimately improving both individual performance and firm-wide outcomes. Why early training drives long-term performance in quant roles.
Read More