Extracting Benefits From Anomalies (EP.08)
Summary
In Episode 8 of ReSolve’s Masterclass, hosts Adam Butler, Mike Philbrick, and Rodrigo Gordillo challenge the assumption that linear relationships best capture market anomalies. They argue that traditional factor investing often leaves excess information on the table by ignoring non-linear shapes, conditionality, and market-specific nuances. The discussion highlights how machine learning helps identify optimal complexity and how ensemble strategies can generate sustainable alpha by diversifying error terms across many variables. The episode also examines David Swensen’s approach to hiring in…
Topics discussed
Linear vs. Non-Linear Relationships in Markets
Welcome to Resolve's Long Horizon Investing Masterclass
Sponsor: Horizon Resolve Adaptive Asset Allocation ETF
Episode 8 Intro: Challenging Linear Factor Assumptions
Limitations of Traditional Factor Literature
Value Factor Variability Across Markets (e.g., Japan)
Statistical Challenges in Measuring Value Factors
Non-Linear Trends and Inverted Momentum Signals
Market Specificity and Participant Differences
Why Traditional Factor Investing Still Works
The Complexity Gap Between Beta and Alpha
Conditionality and Interacting Market Variables
Seasonality and Mechanical Intuition in Assets
Machine Learning and the Bias-Variance Tradeoff
Applying ML to Financial Time Series Correctly
Ensemble Models and Error Cancellation
Expanding Sample Size Cross-Sectionally
Defining Robust Benchmarks for Factors
Sustainable Alpha and Continuous Innovation
Investor Comfort and Risk Tolerance in Alpha
Lessons from David Swenson and Yale Endowment
Preview: Tail Risk and Liquidity Shocks
Outro and Resources
Listen ad-free on Castria