The Market Creates Its Own Trends — Richard Brennan | Systematic Investor | Ep.418

Episode Overview

• This episode explores the enduring validity and structural drivers of trend-following investment strategies, debunking the common post-GFC narrative that the strategy's "edge" has been permanently lost or arbitraged away.
• It reframes financial markets as Complex Adaptive Systems (CAS) rather than static, machine-like environments, drawing heavily on the pioneering work of W. Brian Arthur and the Santa Fe Institute to explain how participant interactions shape market realities.
• The discussion covers key economic and systemic phenomena—such as reflexivity, path dependence, increasing returns, and Goodhart's law—showing how these concepts drive the massive, non-linear market trends (outliers) that trend followers exploit.
• It provides a deep conceptual foundation for why responsive, price-based trading strategies are logically superior to predictive models in an inherently uncertain, co-created, and ever-evolving market ecology.

Key Concepts

• The Migration of the Trend-Following "Edge": Empirical research spanning four decades shows the trend-following edge did not disappear after the Global Financial Crisis; rather, it migrated. While shorter-term horizons (e.g., 20 days) experienced a severe decline in risk-adjusted performance, the edge relocated to medium- and longer-term horizons (e.g., 200 days).
• Outliers as Structural Market Features: Massive market trends are not temporary, exploitable anomalies that can be easily arbitraged away. Instead, they are permanent, structural features of markets driven by human behavior, leverage, feedback loops, forced liquidations, and changing capital flows.
• Complex Adaptive Systems (CAS): Financial markets are environments where independent participants interact, learn, and adapt to one another. Because investors' expectations drive actions that directly change market prices—which then feed back to reshape those very expectations—markets cannot be studied or modeled in isolation.
• The El Farol Bar Problem: This thought experiment models a collective decision-making system where individual predictions can become self-defeating (e.g., if everyone predicts a bar will be empty, everyone goes, making it crowded). It illustrates that in reflexive systems, there is no single "correct" rational expectation or mathematical equilibrium because an actor's optimal choice depends entirely on what everyone else does.
• Reflexivity and Positive Feedback Loops: Popularized by George Soros, reflexivity describes the circular loop between beliefs, actions, and reality. While negative feedback forces push a system back toward equilibrium (mean reversion), positive feedback loops amplify change away from equilibrium, driving the explosive momentum and trends that power trend-following returns.
• Increasing Returns vs. Diminishing Returns: Traditional economics relies on diminishing returns (which restore equilibrium). In contrast, knowledge, network, and technology-driven systems exhibit increasing returns, where early adoption or success makes further expansion more likely, creating "winner-take-all" dynamics and dominant market outliers.
• Path Dependence: This principle states that the route a market takes to reach its current state directly determines its future outcomes. The eventual state of a system is not pre-determined by starting conditions but is co-created step-by-step along the path.
• Goodhart's Law & System Gaming: Goodhart's Law states that when a measure becomes a target, it ceases to be a good measure. Because participants naturally find ways to "game" systemic boundaries, risk limits, and regulations to maximize rewards, their adaptive behaviors inevitably shift the stability of the entire market.

Quotes

• At 0:02:33 - "The received wisdom from many allocators and investors was that the edge of trend following had weakened post-GFC... but many were saying that the industry had become crowded and the opportunity had been arbitraged away." - Explaining the skepticism and dominant market narratives surrounding trend following after its decade-long "winter."
• At 0:04:07 - "We concluded that the edge had not disappeared, but rather, it had moved." - Revealing the core finding of their research into long-term market trends.
• At 0:04:47 - "Since 2020, the strongest results have concentrated around the medium to long-term 200-day horizon, whereas the risk-adjusted performance over the shorter term, such as the 20-day horizon, had fallen to roughly one-quarter of its earlier level." - Highlighting the specific structural migration of trend-following returns from short to long-term lookbacks.
• At 0:06:53 - "Outliers, or these big trends, are actually structural features of markets. They are not temporary anomalies waiting to be arbitraged away." - Arguing against the academic view that markets eventually eliminate all inefficiencies.
• At 0:07:22 - "Markets contain interacting participants, leverage, feedback, forced activity, and changing flows of capital... those ingredients create movements that feed upon themselves and travel much farther than anyone expects." - Outlining the behavioral, mechanical, and structural drivers of massive market trends.
• At 0:17:39 - "How clever your choice is depends entirely on what everyone chooses, not what you choose... We believe we are using our models to forecast an outcome by ourselves, but we're in fact [subject to] everyone's models which are actually deciding the outcome." - Explaining the reflexivity of markets through the lens of the El Farol Bar problem.
• At 0:25:20 - "The future price is not waiting to be discovered like the answer to a puzzle; it emerges from all the participant decisions, not just your own." - Refuting the traditional "efficient market" view that prices simply reflect a static intrinsic value.
• At 0:27:24 - "No rule became permanently superior because its use of that rule helped change the environment in which it operated." - Explaining why static trading strategies or economic models eventually degrade as market participants adapt to them.
• At 0:31:07 - "The future is not sitting there fully formed waiting to be uncovered; participants are helping write it step by step." - Explaining why trend followers focus on responding to current price action rather than predicting targets.
• At 0:33:32 - "Diminishing return... eventually the land becomes crowded and each new worker contributes less... but Brian Arthur studied situations in which the opposite can occur... with increasing returns, expansion makes further expansion more attractive." - Distinguishing agricultural/industrial economics from tech and network-driven market trends.
• At 0:36:24 - "Instead of assuming that the market had found this correct balance, they asked: what happens if different participants interact, learn, and adapt, and we observe what emerges?" - Describing the paradigm shift pioneered by the Santa Fe Institute toward evolutionary economic models.
• At 0:46:55 - "Increasing returns is where adoption helps create the winner... A small early advantage created that environment that continued to favor the same technology." - Explaining how market dominance is often driven by feedback loops rather than pure product superiority.
• At 0:47:11 - "The eventual outcome is not contained in the starting conditions alone, it's actually created step-by-step along the path." - Defining path dependence and why historical context is vital to understanding current market states.
• At 0:48:54 - "Price-based strategies emerged... learning did not guide everyone smoothly toward one correct answer, but the search for better rules helped create the instability everyone was trying to understand." - Showing how trend-following naturally emerges as a dominant survival strategy in adaptive agent-based simulations.
• At 0:52:26 - "Gaming doesn't necessarily mean breaking the rules; it can mean following them in a way that earns a reward while defeating their purpose." - Explaining how systematic rules are inevitably gamed by participants, changing the system's overall behavior.
• At 0:54:04 - "An isolated break becomes a transition into a different market environment... and a boundary can look strongest immediately before the behavior built around it helps produce that break." - Describing how stability encourages leverage and risk-taking, which ultimately triggers sudden market transitions.
• At 0:54:35 - "A backtest provides evidence, but it's certainly not a promise for the future... relationships change over time." - Highlighting the reality of market non-stationarity and the limits of historical simulations.

Takeaways

• Trade Reactively, Not Predictively: Abandon predictive forecasting models that target specific future prices. Instead, implement a responsive strategy that reacts to established price trends using rules-based trailing stops to capture moves as they unfold.
• Maintain a Long-Term Trend Horizon: Avoid over-concentrating in short-term lookbacks (e.g., 20 days) where edge has eroded, and instead design and deploy trend-following systems focused on medium- to long-term (e.g., 200-day) horizons.
• Diversify Extensively Across Markets: Since future market states cannot be mapped with certainty, maintain highly diversified exposure across a wide array of asset classes to maximize the mathematical probability of catching rare, unpredictable outlier trends.
• Preserve the Skew by Avoiding Early Profit-Taking: Do not compromise long-term returns by attempting to "smooth" equity curves through taking partial profits too early. The majority of trend-following returns are driven by letting rare, massive outliers run.
• Keep Individual Losses Small: Expect and accept frequent small losses during non-trending or choppy periods as a necessary business cost of waiting for the massive, trend-enforcing feedback loops that generate primary returns.
• Anticipate That Stability Breeds Instability: Guard against systemic fragility during prolonged periods of market calm. Recognize that as volatility declines, participants naturally increase leverage and drop hedges, setting the stage for sudden, reflexively driven market crashes (Minsky moments).
• Design Systems for Non-Stationarity: Do not rely on backtests as static promises of future performance. Continuously build robust risk controls and expect market relationships to shift over time as participants learn, adapt, and game existing rules.