The Lie of the Efficient Market: Why Trillion-Dollar Trends Keep Proving Finance Textbooks Wrong

Traditional financial education presents a reassuring narrative: asset markets are efficient, price movements are random, and the optimal long-term strategy is to buy and hold a diversified equity portfolio. This approach depends on mean reversion, assuming prices will return to their intrinsic value over time.

This is the world of convergent risk.

An alternative approach to global markets disregards intrinsic value, ignores price-earnings ratios, and rejects the randomness of price movements. This is the domain of divergent risk and the basis of trend following. Rather than expecting markets to revert to the mean, trend following assumes that once a market moves, it is more likely to continue in that direction.

Recognising this distinction is essential. Trend following is not merely a trading technique; it represents a fundamentally different paradigm that challenges traditional modern portfolio theory.

Deconstructing the Efficient Market Hypothesis

The bedrock of traditional finance is the Efficient Market Hypothesis, or EMH. Developed in its modern form by Eugene Fama in the 1960s, EMH states that asset prices fully reflect all available information. Because markets are efficient, price changes must only occur in response to new, unpredictable information. Therefore, past price movements cannot be used to predict future returns. Prices follow a random walk, meaning the next price move is entirely independent of the last.

If EMH were entirely accurate, trend following would be mathematically impossible. Trend followers analyse past price data to identify sustained directional moves and take positions expecting these trends to persist. Without any predictive power in past prices, systematic trend following would be eroded by transaction costs.

In reality, this theory faces significant challenges. Financial history features prolonged market moves that, according to random walk models, should occur only once in several billion years.

On Black Monday, October 19, 1987, the S&P 500 fell over 20% in one day. No significant news or macroeconomic data justified this revaluation. The crash resulted from market mechanics, investor panic, and early computer-driven portfolio insurance models that amplified the decline.

A similar dislocation occurred during the 2000 Dot-Com Bust. Technology stocks traded at extreme valuations without earnings, fueled by speculation. When the bubble burst, the Nasdaq lost nearly 80% of its value over several years, resulting in a prolonged downward trend.

The 2008 Great Financial Crisis followed a similar pattern. The collapse of subprime mortgages triggered a prolonged decline in global equities that persisted for over 18 months, well after the initial data became public.

More recently, the 2022 Inflationary Shock demonstrated similar dynamics. Rapid monetary tightening after years of low interest rates led to historic trends in bond markets and significant upward movements in agricultural and energy commodities.

These events illustrate that markets often fail to adjust instantly or rationally to new information. Instead, they experience prolonged periods of under-reaction or over-reaction, resulting in trends.

The Structural Drivers of Momentum

If markets are not perfectly efficient, why do trends persist? Why are these anomalies not immediately arbitraged away? The answer lies in human psychology and market structure, as described by behavioural finance. Trends result from persistent, well-documented flaws in how people process information and manage risk.

Anchoring Bias and Under-Reaction

When new information enters the market, investors seldom price it in fully or immediately. They anchor to previous prices. For example, after a company announces significant earnings or a commodity faces a supply shortage, investors often assume the change is temporary and underreact. As the new reality becomes clear over time, capital gradually flows into the asset, creating a sustained upward trend.

Herding and Over-Reaction

Human social behaviour extends to financial markets. As an asset rises, it draws media attention and public interest, leading to Fear of Missing Out (FOMO). Institutional managers, seeking to match peers’ performance, also buy in. This herd behaviour creates a self-fulfilling cycle, driving prices higher and attracting more buyers, often pushing the asset beyond its fundamental value.

Risk Management Liquidation

Downward trends are often intensified by structural risk management rules. When markets decline, leveraged investors face margin calls and must provide additional collateral. If unable, their positions are automatically liquidated. Institutional funds, governed by strict Value at Risk (VaR) limits, are also forced to sell as volatility rises. This forced selling amplifies institutional pressure and extends downward trends.

Financial economist Andrew Lo introduced the Adaptive Markets Hypothesis (AMH) to provide a cohesive framework. Rather than viewing markets as static and perfectly efficient, AMH sees them as ecosystems where investors compete for survival. Investors do not always act rationally; they use heuristics, learn, and adapt. When conditions change, such as a shift from low to high inflation, old strategies fail and new behaviours emerge. Trends reflect the market ecosystem adapting to new macroeconomic realities.

Defining Convergent vs Divergent Risk

To construct systematic portfolios, investment strategies should be categorised by how they interact with market boundaries. Each strategy aligns with either a convergent or divergent risk profile.

Convergent Risk

A convergent strategy assumes that markets have structure and boundaries. Investors believe they can determine an asset’s fair value and, when prices deviate, enter trades expecting a return to the mean.

Examples include value investing, where stocks with low price-to-earnings ratios are purchased in anticipation of a return to average valuations. Fixed income arbitrage involves exploiting small price discrepancies between similar bonds, expecting the spread to close. Statistical mean reversion strategies buy indices after significant declines, anticipating a rebound.

Convergent strategies typically have a high win rate, generating small, consistent profits. However, they are exposed to negative skewness: rare, extreme events can break assumed boundaries, resulting in catastrophic losses that may erase years of gains.

Divergent Risk

A divergent strategy takes the opposite approach. Divergent investors acknowledge uncertainty about the future, intrinsic value, and market boundaries. Rather than betting on stability, they position for structural change.

Trend following exemplifies a pure divergent strategy. The trend follower waits for a market to break out of historical boundaries and then takes a position in the direction of the breakout, expecting the divergence to continue.

The mathematical profile of a divergent strategy is the inverse of a convergent strategy:

• Win Rate: Convergent strategies maintain a high win rate, typically between 60% and 80%, while divergent strategies operate on a low win rate, usually between 30% and 40%.
• Payoff Structure: Convergent profiles yield small, frequent wins but risk massive, catastrophic losses. Divergent profiles yield small, controlled losses but achieve massive, uncapped wins.
• Statistical Skewness: Convergent returns exhibit negative skewness, representing left-tailed risk, whereas divergent returns exhibit positive skewness, creating right-tailed profit.
• Ideal Environment: Convergent models thrive in stable, range-bound markets, while divergent models excel in high-volatility environments with structural macroeconomic shifts.

Although divergent strategies have a low win rate, their winning trades are disproportionately large. Trend followers do not set profit targets, allowing successful trades to run. A single major trend in markets such as Crude Oil, the US Dollar, or Gold can offset many small losses and drive portfolio returns.

Direction-Agnostic and Asset-Agnostic Trading

Traditional asset selection involves evaluating cash flows, debt-to-equity ratios, and management quality. Trend followers disregard these fundamentals, relying instead on two core principles.

Asset-Agnostic Trading

For systematic trend followers, all liquid instruments are simply ticker symbols with fluctuating prices. There is no need for expertise in supply chains or central bank policy. The system processes price data, extracts trend signals, applies risk management, and executes trades. This asset-agnostic approach enables trading across hundreds of global markets, providing exceptional diversification.

Direction-Agnostic Trading

Traditional investing maintains a structural long bias, exposing investors to risk during prolonged bear markets. Trend following assigns equal weight to long and short positions. The system goes long in upward trends and short in downward trends. By shorting declining assets, trend followers can profit during market downturns. They do not predict or hope for crashes; they respond to price trends as they occur.

Conclusion

Trend following is grounded in the empirical observation that human behaviour and institutional structures consistently generate sustained trends in financial markets. By avoiding predictions and intrinsic value calculations, trend followers embrace divergent risk. They use automated, rule-based systems that accept small, frequent losses to capture large, positively skewed trends across global asset classes.

In the next instalment, we will examine the operational aspects of this strategy, focusing on global futures contracts and the structure of multi-asset diversification.

Clement Ong holds an MBA majoring in finance and a Master of Accountancy. He is currently pursuing an MSc in Financial Engineering.

This article was prepared with the assistance of AI for language and structural refinement. All ideas, frameworks, and interpretations expressed in this article are those of the author.