papermomentumctafuturesjfe

Moskowitz, Ooi & Pedersen (2012) — "Time Series Momentum"

The foundational paper establishing that an asset's own past returns positively predict its future returns across asset classes. Journal of Financial Economics, 104(2), 228–250.

Citation

Moskowitz, T.J., Ooi, Y.H., & Pedersen, L.H. (2012). "Time Series Momentum." Journal of Financial Economics, 104(2), 228–250.

Core contribution

Documents that time-series momentum (TSMOM) — the strategy of going long instruments with positive recent past returns and short those with negative past returns — produces statistically and economically significant risk-adjusted returns across:

  • 58 liquid instruments: equity index, currency, commodity, and bond futures.
  • Lookback windows from 1 to 60 months.
  • The strategy is especially profitable in extreme market environments (equity-market crashes), giving it tail-hedging properties when added to a traditional 60/40 stock-bond portfolio.

Why it matters

  • Triggered the modern academic and practitioner literature on managed futures / CTA strategies.
  • Established the binary rule as the baseline: f(s_t) = \text{sign}(\sum_{\tau=1}^{12} r_{t-\tau}) with volatility-scaled position.
  • The basis for the entire line of research that Moskowitz Sabbatucci Tamoni Uhl 2025 Nonlinear TSMOM extends.

Key findings

  • TSMOM with a 1-year lookback has been profitable since at least the 1980s across asset classes.
  • Profits come largely from the tails — particularly equity downturns (1987, 2008, 2020).
  • Reversal patterns at very long horizons (3–5 years) — momentum is mean-reverting on a long horizon.
  • The strategy has negative correlation with traditional stock-bond portfolios, making it a useful diversifier.

See also