Why Trend Following Never Looks the Same Twice | Systematic Investor | Ep.365

Episode Overview

• The episode explores the concept of "return dispersion" among trend-following managers, questioning why seemingly similar strategies produce vastly different results.
• It breaks down the key "design choices"—such as model speed, asset allocation, and market selection—that act as a unique "fingerprint" for each manager's performance.
• The conversation addresses the common discrepancy between the theoretical performance of backtested models and the actual performance of real-world CTA indices.
• It analyzes how trend-following strategies perform under different types of market stress, distinguishing between sudden "magnitude shocks" and more gradual "correlation shocks."

Key Concepts

• Return Dispersion: The central theme, referring to the wide range of returns among trend-following managers. This dispersion is not random but a result of deliberate strategy design choices.
• Strategy Design Choices: Key factors that create performance differences include:
  • Speed: How quickly a model reacts to new trends (fast vs. slow).
  • Asset Allocation: Tilts toward specific asset classes and the inclusion of alternative or non-traditional markets.
  • Carry: The decision to incorporate carry strategies alongside trend following.
  • Risk Weighting: Methodologies like equal risk versus market-capacity weighting.
  • Signal Type: The choice between different signals, such as moving averages versus channel breakouts.
• Crisis Alpha: The ability of a strategy to perform well during major market sell-offs. The discussion highlights that faster trend-following models are generally more effective at capturing crisis alpha.
• Index Replication Challenges: It is extremely difficult to accurately replicate a composite index like the SG Trend Index in a simple backtest because it comprises diverse managers with varying strategies, making modeled Sharpe ratios often appear higher than reality.
• Magnitude vs. Correlation Surprises: A framework for understanding market events:
  • Magnitude Surprise: A large, sudden shock where all assets move together (e.g., SVB collapse). CTAs tend to perform poorly on these days.
  • Correlation Surprise: A shift in the underlying relationships between assets, which often signals a new trend and is a more favorable environment for trend followers.

Quotes

• At 0:12:53 - "Trend following has a reputation of being kind of the simplest strategy...you buy strength, you sell weakness, you cut losses, you let winners run. So how different can managers really be?" - Niels Kaastrup-Larsen posing the central question that the Man Group paper on dispersion seeks to answer.
• At 0:13:24 - "Dispersion isn't random. It is the fingerprint of each system...it is a design choice." - Niels Kaastrup-Larsen explaining the main thesis of the Man Group paper.
• At 0:16:40 - "Being faster is actually really helpful for really bad environments." - Katy Kaminski sharing a key takeaway from the Man Group paper regarding strategy speed and crisis performance.
• At 23:07 - "'My only quibble is that the model numbers seem off well by a lot... Over 25 years the average Sharpe ratio seems to be around .75. The cold reality is that the actual Sharpe ratio of the SG Trend index is .36.'" - Niels Kaastrup-Larsen reads a critique from Andrew Beer about the difference between a model's Sharpe ratio and the actual SG Trend Index's performance.
• At 38:50 - "There's a history of them and we plot them, there's a few of them that are really big and purely magnitude... you can guess which ones they are: Brexit, you know, SVB, Black Friday." - Katy Kaminski lists examples of "magnitude surprise" days—sudden, large market shocks—and notes that trend-following strategies historically do not perform well during these specific events.

Takeaways

• Differences in trend-following performance are not due to luck but are the direct result of intentional design choices in speed, market selection, and risk management.
• While faster models may provide better protection during market crises, no single strategy design is optimal for all market conditions, reinforcing the need for diversification.
• Trend followers are vulnerable to sudden, sharp market shocks ("magnitude surprises") but are well-positioned to profit from environments where asset correlations break down and new trends emerge.
• Instead of trying to find one "perfect" manager, investors should understand the different drivers of return and consider diversifying across 2-3 managers with different strategic approaches to smooth out performance.