Time Series Momentum, made practical: what the evidence really says

Trend-following isn’t a campfire story told by quant; it’s a repeatable pattern that shows up across equity indices, bonds, commodities, and currencies. This executive summary unpacks what the landmark “Time Series Momentum” study actually found, why it matters for portfolio construction, and how to put a simple, risk-aware version to work.

Listen to a conversational review of this article:

What the paper set out to test

• Question: Do assets that rose over the past year tend to keep rising (and vice versa) over the next month(s)?
• Scope: 58 liquid futures/forwards spanning equity indices, government bonds, commodities, and major FX pairs, with long histories and robust liquidity screens.
• Method in one line: Look at each asset’s own past return (not its rank versus peers), then go long if positive and short if negative—size positions by volatility so no single market dominates.

What they found (in plain English)

• Trends persist, then fade: Positive (or negative) 12-month moves tend to continue for 1–12 months and partially reverse thereafter.
• It’s everywhere: The effect appears across all 58 contracts examined.
• Real excess returns: A diversified time-series momentum (TS-MOM) portfolio earns large, statistically strong alphas with little exposure to standard equity style factors.
• Shock absorber: TS-MOM shines in extreme markets—big up and big down months—making it a useful crisis diversifier.
• Who’s on each side: Speculators typically lean with the trend and profit; hedgers take the other side and, on average, pay for the protection/liquidity.
• What drives it under the bonnet: The return premium is mainly linked to simple autocovariance (an asset’s own return persistence), not cross-asset lead–lag quirks, and shows up in both spot moves (information flow) and roll yield (hedging pressure).

How they built the core strategy

1. Signal: Use the past 12-month excess return for each instrument (no peer ranking).
2. Position: Long if the look-back is positive, short if negative.
3. Sizing: Scale each position by inverse ex-ante volatility (e.g., an EWMA estimate) so each market contributes similar risk; a representative setup targeted ~10% annualised portfolio volatility.
4. Holding: Rebalance monthly; be aware that trends often start to mean-revert after ~12 months.
5. Diversify: Combine many liquid markets; that breadth is a big part of the edge.

Why this matters for allocators

• Different beast to cross-sectional momentum: This is not “buy recent winners, sell losers” across a stock universe; it’s an each-asset trend that travels well across asset classes.
• Lower correlation to core bets: Helps smooth the ride versus traditional long-only equity/bond risk.
• Behavioural and structural roots: A mix of under-reaction/over-reaction (information diffusion) and persistent hedging pressure (roll yield) offers intuitive, enduring foundations.

A simple TS-MOM playbook you can actually implement

1. Define your universe: Start with major futures: SPX 500, bond futures, G10 FX, and the large commodity complexes.
2. Compute signals monthly: 12-month excess return per instrument; skip fancy extras until your data pipeline is rock-solid.
3. Risk first: Use an exponentially weighted volatility estimate; cap any single asset’s risk to avoid concentration.
4. Execution basics: Rebalance once a month; avoid trading at illiquid times; net down offsetting exposures.
5. Govern the reversal: Monitor ageing trends; the effect tends to cool after ~12 months—consider fading or de-weighting over-ripe legs.
6. Stress test: Check behaviour in crises; TS-MOM historically cushioned extremes—treat it as a diversifier, not a silver bullet.

See the trend concept against real markets (illustrative)

Use the tabs to scan different macro markets; TS-MOM looks for sustained direction, not one-day pops.

Risks, frictions, and good habits

• Costs & slippage: Monthly turnover is modest, but trading illiquid contracts or poor timing can erode edge.
• Crowding: Trend-following is popular; stick to robust sizing and broad diversification.
• Parameter humility: 12-month look-back is a sturdy default, but avoid over-tuning; keep the engine simple and reliable.
• Know the cycle: Expect partial reversals after a year; don’t treat trends as permanent.

Putting theory into practice

• Start small, scale sensibly: Implement with a low target volatility, then grow as process confidence rises.
• Measure what matters: Track risk-adjusted returns, crisis-period behaviour, and correlation to your core book.
• Stay systematic: The edge comes from consistency across markets and time—avoid discretionary overrides.

Educational content only—this is not investment advice.