CTA Strategy (商品交易顾问策略 / 管理期货)
A managed-futures strategy that systematically trades futures across asset classes using systematic rules. The dominant real-world implementation of Time-Series Momentum.
Definition
CTA = Commodity Trading Advisor. In modern usage, "CTA" and "managed futures" are nearly synonymous — a hedge fund or trading firm that uses systematic, often trend-following, futures strategies across equities, rates, FX, and commodities. The strategy is:
- Systematic (rule-based, not discretionary).
- Diversified across asset classes (typically 30–100+ futures markets).
- Futures-based (margin-efficient, easy to go long or short).
- Volatility-targeted (see Volatility Targeting).
Connection to Time-Series Momentum
Most CTAs are essentially productionized TSMOM with risk overlays:
- Signal: linear or binary momentum over 1M–12M lookbacks.
- Sizing: Volatility Targeting to a target annualized vol (often 10–15%).
- Universe: 30–100+ liquid futures, often equally weighted or risk-parity-weighted within sector.
The Moskowitz, Sabbatucci, Tamoni & Uhl (2025) paper explicitly notes that "popular trend-following and managed futures (CTA) strategies of many hedge funds are based on this research, commanding many hundreds of billions of dollars in assets under management."
Typical CTA construction
| Layer | Decision |
|---|---|
| Universe selection | Liquid futures, multi-asset-class |
| Signal | Time-Series Momentum (1M, 3M, 12M blend), some add Cross-Sectional Momentum |
| Position weighting | f(s_t) \cdot \sigma_{target}/\hat\sigma_t — see Nonlinear Signal Mapping and Volatility Targeting |
| Risk overlay | Maximum leverage cap, sector exposure caps, drawdown stops |
| Execution | Intraday or end-of-day; minimize slippage; see Bordigoni et al. (2016) |
Why the convex payoff matters
The main selling point of CTAs to traditional stock-bond investors is the convex / tail-hedge payoff documented by Moskowitz et al. (2012) and confirmed by Moskowitz Sabbatucci Tamoni Uhl 2025 Nonlinear TSMOM:
- CTA returns are positive on average (positive alpha unconditionally).
- CTA returns are most positive when stock/bond markets crash (2008, 2020).
- Combined with a 60/40 stock-bond portfolio, a 10–25% CTA allocation has historically improved Sharpe ratio and reduced drawdown.
Drawbacks
- Momentum crashes: rapid rebounds from bear markets can produce large CTA drawdowns (Daniel & Moskowitz, 2016).
- Capacity: many contracts have limited capacity; the largest CTAs are approaching capacity in some markets.
- Crowding: correlated strategies can amplify drawdowns (Bollen et al., 2021).
- Fee load: typical CTA fee is 2/20, similar to hedge funds, which is steep given the rule-based nature.
Open questions
- Does Ferson-Siegel Optimal Weighting apply uniformly across CTA sub-strategies, or do different lookbacks need different f?
- What is the right re-estimation cadence for the empirical NL function f?
- How should CTA exposure be sized dynamically given momentum crash risk?
See also
- Time-Series Momentum — primary signal.
- Volatility Targeting — sizing layer.
- Nonlinear Signal Mapping — modern weighting choice.
- Cross-Sectional Momentum — secondary signal some CTAs add.
- Moskowitz Ooi Pedersen 2012 Time Series Momentum — foundational paper.
- Moskowitz Sabbatucci Tamoni Uhl 2025 Nonlinear TSMOM — modern extension.
- Lookback Straddle Portfolio (回望跨式期权组合) — CTA 收益的期权等价物，Fung-Hsieh 七因子模型 趋势因子的来源。
- Fung & Hsieh 2004 Hedge Fund Benchmarks — 把"CTA 赚什么钱"翻译成可交易 ABS 因子的标准方法。
References
- Moskowitz, T.J., Ooi, Y.H., & Pedersen, L.H. (2012). "Time Series Momentum." JFE, 104(2).
- Hurst, B., Ooi, Y.H., & Pedersen, L.H. (2017). "A Century of Evidence on Trend-Following Investing." JPM, 44(1).
- Daniel, K. & Moskowitz, T.J. (2016). "Momentum Crashes." JFE, 122.
- Moskowitz, T.J., Sabbatucci, R., Tamoni, A., & Uhl, B. (2025). "Nonlinear Time Series Momentum." Working paper.