Skip to content

Time Series Momentum: summary and key takeaways

The authors test whether an instrument's own past 12-month excess return predicts its next return across 58 futures and forwards in equities, bonds, currencies and commodities. Every contract shows positive predictability, and a volatility-scaled diversified trend portfolio earns large alphas against standard factors. The effect lasts about a year and then partly reverses.

Moskowitz, T.J.; Ooi, Y.H.; Pedersen, L.H. · 2012 · Journal of Financial EconomicsSummary updated 20 November 2026
Published2012Journal article
Asset focusMulti-assetTrend following
TypeStrategyTrend & Momentum
Reading time5 minthis summary

Time Series Momentum at a glance

The question
Does a security's own past return predict its future return across asset classes, and who earns or pays for that pattern?
Dataset & period
58 liquid futures and forwards (commodities, currency pairs, developed equity indices, government bond futures), January 1965 to December 2009. Monthly returns; strategy evaluation mainly from 1985 when instrument coverage is broad. CFTC weekly positions of speculators and hedgers cover only part of the contracts.
Method
Each month the paper regresses a volatility-scaled return on the same instrument's volatility-scaled return lagged h months, using a pooled panel with standard errors clustered by time. It then builds trading strategies that go long after a positive past return and short after a negative one, sizing each position inversely to ex-ante volatility. The main strategy, TSMOM, uses a 12-month look-back and 1-month holding period, and is averaged across instruments. Alphas are measured against market, bond, commodity, Fama-French and value/momentum factors.

Key findings of Time Series Momentum

  • All 58 contracts show positive returns to the 12-month trend strategy, and 52 are statistically different from zero at the 5% level.
  • The diversified TSMOM portfolio has an alpha of about 1.58% per month, or 4.75% per quarter, after controlling for the market, SMB, HML and UMD.
  • Against the value and momentum everywhere factors, the TSMOM alpha is still 1.09% per month with a t-stat of 5.40.
  • Regressed on cross-sectional momentum across the same assets, TSMOM has a beta of 0.66 (t-stat 15.17, R-square 44%) and a remaining alpha of 76 basis points per month (t-stat 5.90).
  • Within-class correlations of TSMOM returns run from 0.37 to 0.38 for equities and fixed income down to 0.10 and 0.07 for commodities and currencies, and cross-asset-class TSMOM correlations are all positive.
  • Trends persist for about a year and then partly reverse. Speculators' net positions follow the trend and mean-revert at about the same horizon, which the paper reads as speculators profiting at hedgers' expense.

Robustness, as reported in the paper

  • In the earlier 1966 to 1985 sample, which has fewer instruments, TSMOM still earns a significant return with an annualized Sharpe ratio of 1.1.
  • Results hold across look-back and holding periods, asset classes, and the far futures contract; the paper reports they are similar with simple OLS and with cash equity indices.
  • TSMOM returns show no significant link to the TED spread, the VIX, or Baker-Wurgler sentiment measures, and the correlation between contract illiquidity and TSMOM Sharpe ratio is -0.16.
  • The paper states its results are robust to more sophisticated volatility models than the simple exponentially weighted one it uses.

Caveats and limits

  • Returns are gross; the paper does not report transaction costs or slippage in its main results (inferred).
  • The sample ends in December 2009 and the paper is a September 2011 working-paper version, so later data is not covered.
  • TSMOM gives up money in sharp trend reversals, such as March to May 2009 after the crisis, and in the third quarter of 2008.
  • Hedger/speculator labels come from CFTC commercial and non-commercial classes, which the paper says add noise, and the direction of causality between returns and positions is indeterminate.

Items marked (inferred) are our reading of the design, not statements by the authors.

Why it matters

The paper gives a simple, volatility-scaled definition of trend following that can be replicated across asset classes. It shows the premium is distinct from stock-level cross-sectional momentum and has option-straddle-like payoffs in extreme markets. Practitioners can use it as a benchmark for trend and managed-futures exposure.

Questions about this paper

What is time series momentum?

It is the predictive link between an instrument's own past excess return and its future return. The strategy goes long when the past return is positive and short when it is negative, using volatility-based position sizes.

How is it different from cross-sectional momentum?

Cross-sectional momentum ranks securities against each other, while time series momentum looks only at each security's own past return. The paper finds the two are related but TSMOM keeps a significant alpha after controlling for cross-sectional momentum.

What does the paper say drives the profits?

Positive auto-covariance between lagged 1-year returns and next-month returns is the main driver. The paper also finds speculators gain from trends while hedgers take the other side.

Where can I read the original paper?

The original is at its publisher: https://doi.org/10.1016/j.jfineco.2011.11.003. We link to the paper and do not host the PDF.

Follow new quant research

Download MacroFinance market data as CSV or Excel, or pull it by API, and test the ideas in these papers yourself. New summaries are added as we review them.

Create a free account
Summaries are written by MacroFinance from the published papers, in our own words, for information only. They are not investment advice and may omit detail. Papers are not hosted here: each page links to the original at its publisher.