Summary
Niels Kaastrup-Larsen and Rob Carver discuss current challenges in trend following, including recent performance drivers, the rise of perpetual futures, the risk of overfitting with AI-generated strategies, and lessons from research on asset class attribution and drawdowns. They emphasize the importance of robust processes and true diversification, highlighting commodities as a consistently valuable source of crisis alpha.
- Commodities have been the dominant source of trend following returns in the last seven years and are crucial for crisis-period performance.
- Fixed income trend returns were strong in the 2010s, while equities contributed little to trend returns historically.
- Overfitting remains a major risk, now amplified by easy AI-generated backtests that lack robust process.
- A robust research process is essential for evaluating strategies, regardless of whether the idea came from AI or human insight.
- Perpetual futures could simplify trading but may disrupt traditional futures markets and hurt hedging and price discovery.
- True diversification across uncorrelated markets like commodities is more valuable than chasing past strong sectors.
- The new Fed Chair’s plan to revamp economic data collection could impair economic-data-dependent strategies.
- Draw down analysis across asset classes offers limited statistical power but suggests diversification benefits.