Odds on Open Former Nomura Managing Director: How the Sell-Side Created Modern Quant Finance
Sep 17, 2025
In a fascinating dialogue, Joe Mezrich, Founder of Metafoura LLC and former Managing Director at Nomura, shares insights from his extensive career in quantitative finance. He discusses how the sell side pioneered modern factor investing and the pitfalls of models like the Barra risk model during market crises. Joe emphasizes the importance of interpretability in machine learning, reflecting on how complexity can overshadow understanding. He also highlights the evolution of market-neutral strategies and the critical role of alternative data in today's finance.
AI Snips
Chapters
Books
Transcript
Episode notes
Early Sell‑Side Innovation At Salomon
- Joe Mezrich describes Salomon Brothers as a creative, cartoon-like place that invented new trading and research techniques.
- He recounts implementing early factor models, robust statistics, and the StockFacts Pro PC tool in the early 1990s.
Applying CART And Early ML In The 1990s
- Mezrich describes early machine learning adoption: reading Leo Breiman and implementing CART models in the 1990s.
- He cites a 1994 application for option hedging and global equity swap structuring at Salomon.
Speed Matters In Risk Modeling
- Risk models like Barra failed in crises partly because they were too slow to adapt to rapidly changing volatility regimes.
- Faster approaches such as GARCH were adopted to handle accelerating market dynamics after 2000.
