Trend Following: A Multi-Century Edge Hidden in Plain Sight
Why a simple, century-old idea — buy what is going up, sell what is going down — remains one of the most academically validated and behaviourally durable sources of return in modern markets.
Few ideas in investing are simultaneously this old, this simple, and this stubbornly profitable. Trend following — the systematic act of taking long positions in markets that are rising and short positions in those that are falling — has been documented across decades of futures data, across asset classes, and across regimes that include world wars, currency collapses and a global pandemic.
Academic work from Hurst, Ooi and Pedersen famously extended the evidence back more than a century, finding consistent, risk-adjusted returns across equities, bonds, commodities and currencies. The interesting question is not whether the effect exists — the data settle that — but why it persists in liquid, sophisticated markets.
Why the trend premium does not get arbitraged away
Markets discount information slowly when humans are involved. Three behavioural frictions explain most of it: anchoring on prior beliefs, the disposition effect (selling winners too early), and herding once a story becomes consensus. Institutional frictions — mandate constraints, peer-risk aversion, drawdown intolerance — reinforce the slow diffusion of new information into price. Trend systems profit from this latency.
What disciplined trend following actually looks like
A robust trend programme is not a single moving-average crossover. It is a portfolio of signals across timeframes (short, medium, long) and instruments, sized by volatility targeting rather than fixed notional, and rebalanced systematically. Risk is managed by ex-ante volatility scaling — not stop losses applied after the fact.
The role in an institutional portfolio
Trend’s most valuable contribution is not its standalone Sharpe; it is its convexity in equity drawdowns. Because trend systems can go short, they tend to perform best when traditional risk assets perform worst. That is not insurance — it is positive expected return that happens to be conditionally negatively correlated to equities in stress periods. For a 60/40 investor, even a 5–10% allocation has historically improved risk-adjusted outcomes.
FAQ
Is trend following the same as momentum?
They are cousins. Cross-sectional momentum ranks assets relative to each other within an asset class. Time-series momentum (trend following) judges each asset against its own history. The two capture overlapping but distinct premia.
Does trend following stop working in choppy markets?
Sideways regimes are the natural cost of doing business. A well-diversified, multi-timeframe trend portfolio amortises this cost across hundreds of independent positions.
August Quants Research
The August Quants research desk publishes educational essays on systematic investing, market structure, ML in finance and portfolio construction. We write for institutional readers who value rigour over noise.