Skip to content

Systematic Risk, Hedging Pressure, and Risk Premiums in Futures Markets

Authors: Hendrik Bessembinder | Year: 1992 | Journal: Review of Financial Studies, 5(4), 637-667

Thesis

Futures risk premiums are determined by two forces: systematic risk (covariance with the market portfolio) and hedging pressure (net positioning of hedgers). When commercial hedgers are net short (producers hedging output), speculators earn a positive risk premium for providing liquidity -- this is the Keynes-Hicks "normal backwardation" hypothesis, given empirical teeth. Bessembinder shows that hedging pressure, measured from CFTC Commitments of Traders (COT) data, is a significant predictor of futures returns even after controlling for systematic risk. Futures with greater net short hedging pressure earn higher expected returns.

Key Math

The two-factor model for expected futures excess returns:

\[E[r_{f,t}] = \beta_f \cdot \lambda_m + \gamma \cdot HP_{f,t}\]

Where \(\beta_f\) is the futures contract's market beta, \(\lambda_m\) is the equity market risk premium, and \(HP_{f,t}\) is hedging pressure:

\[HP_{f,t} = \frac{N_{\text{short}}^{\text{commercial}} - N_{\text{long}}^{\text{commercial}}}{N_{\text{short}}^{\text{commercial}} + N_{\text{long}}^{\text{commercial}}}\]

When \(HP > 0\) (hedgers net short), speculators earn a premium for going long. The empirical test regresses:

\[r_{f,t+1} = \alpha + \beta_1 \cdot r_{m,t+1} + \beta_2 \cdot HP_{f,t} + \epsilon_{f,t+1}\]

with \(\hat{\beta}_2 > 0\) and significant.

Data & Method

  • Weekly returns on 32 futures contracts (agricultural, metals including gold and silver, financial) from 1967-1989.
  • CFTC Commitments of Traders data for commercial vs. non-commercial positioning (released weekly, reporting Tuesday positions).
  • Two-pass Fama-MacBeth cross-sectional regressions.
  • Controls: market beta (CRSP value-weighted), lagged basis, open interest.
  • Robustness: subsamples, exclusion of agricultural contracts, different market proxies.

Our Replication Verdict

PARTIALLY CONFIRMED -- (1) The hedging pressure effect is real and economically meaningful for agricultural and energy commodities where producers systematically hedge. (2) Gold is a problem case: gold's commercial positioning is dominated by dealers and fabricators, not traditional producers. Central bank activity (selling pre-2010, buying post-2010) is not captured in COT data. The hedging pressure signal for gold is noisy and has low predictive power in our tests (t-stat ~1.2 vs. >2.5 for grains). (3) Silver is slightly better because industrial users (electronics, solar) create more consistent hedging pressure patterns. (4) The COT data has a reporting lag (Tuesday data released Friday, used Monday) that degrades the signal. (5) Post-2004, the growth of commodity index investing blurred the distinction between hedgers and speculators in COT classifications.

Signal Mapping

  • COT positioning signal (SS5.3) -- used as a secondary signal with reduced weight for gold (0.1 weight) and moderate weight for silver (0.2 weight).
  • We compute a z-score of net commercial positioning relative to its 52-week rolling distribution. Extreme net short positioning by commercials is a weak bullish signal.
  • Implementation note: We lag the signal by 3 business days to account for reporting delay.
  • Combined with managed money positioning (added to COT in 2006) for a richer picture of the speculative/hedging balance.

References

  • Bessembinder, H. (1992). "Systematic Risk, Hedging Pressure, and Risk Premiums in Futures Markets." Review of Financial Studies, 5(4), 637-667. DOI: 10.1093/rfs/5.4.637
  • De Roon, F.A., Nijman, T.E. & Veld, C. (2000). "Hedging Pressure Effects in Futures Markets." Journal of Finance, 55(3), 1437-1456.
  • Acharya, V.V., Lochstoer, L.A. & Ramadorai, T. (2013). "Limits to Arbitrage and Hedging." Journal of Financial Economics, 109(2), 441-465.
  • Hong, H. & Yogo, M. (2012). "What Does Futures Market Interest Tell Us about the Macroeconomy and Asset Prices?" Journal of Financial Economics, 105(3), 473-490.