CoChat
Sign Up
Cited by 241Financial Markets and Investment StrategiesMarket Dynamics and VolatilityEconomics of Agriculture and Food Markets

Effects of Index-fund Investing on Commodity Futures Prices

Hamilton, James D., Wu, Jing Cynthia·2015·International Economic Review
DOI 10.1111/iere.12099
Open this paper in CoChat — free account

Abstract

We develop a simple model of futures arbitrage that implies that if purchases by commodity index funds influence futures prices, then the notional positions of the index investors should help predict excess returns in these contracts. We find no evidence that the positions of index traders in agricultural contracts as identified by the Commodity Futures Trading Commission can help predict returns on the near futures contracts. Although there is some support that these positions might help predict changes in oil futures prices over 2006–2009, the relation breaks down out of sample.

AI snapshot

Written from the paper’s full text.

TL;DR

Hamilton and Wu propose a risk-premium framework and test whether index-fund positions predict commodity futures returns. They find essentially no evidence for the 12 agricultural commodities, and while imputed oil index positions predict oil returns in-sample over 2006–2009, the relation breaks down out of sample.

Key findings

  • For the 12 agricultural commodities, SCOT index-trader positions show no statistically significant predictive power: level and weekly-change coefficients are not significant for any of the 12, adjusted \(R^2\) is usually negative, and for 13-week changes none of the 24 estimated slope coefficients is statistically distinguishable from zero. Weekly changes are significant at 5% for cotton and almost for sugar, but with opposite signs.
  • Roll-window daily regressions find a statistically significant \(\hat\pi_4\) for only 1 of 12 agricultural commodities (soybeans), and it has the opposite sign predicted by the price-impact hypothesis. For the next contract, only soybeans and hogs are significant; for the spread, soybean oil and cotton are significantly negative and hogs significantly positive. With a 10-day window, none of the \(S_\tau\) coefficients is significant for individual agricultural contracts, and only cotton shows a significant difference between the first two contracts.
  • For crude oil, Masters-imputed index holdings show insignificant level and weekly-difference coefficients with negative \(R^2\), but the 13-week change is significant in sample (Masters: coefficient 0.0440*, \(R^2=0.0438\); regression-based all-commodity estimate: 0.0448*, \(R^2=0.0499\)). Out of sample from January 17, 2010 to January 3, 2012, the oil regression MSE is 24.01 versus 21.97 for a random walk; Bai-Perron finds breaks at September 30, 2008 and January 13, 2009, with the later subsample coefficient negative (\(t=-0.75\)).
  • The same Masters variable also predicts S&P 500 returns over Singleton’s sample (0.0125*, \(R^2=0.0130\)) but has out-of-sample MSE 6.5520 versus 6.3688 for a random walk. Overall, the paper concludes there is little evidence that index-fund investing measurably affects commodity futures prices.

Methods

  • The paper sketches a simple futures-arbitrage/risk-premium model implying that if index-fund buying influences prices by changing the risk premium, the notional positions of index investors should help predict excess returns; it then estimates reduced-form predictive regressions of futures returns on lagged index-trader positions.
  • Data are CFTC Supplemental Commitments of Traders weekly positions for 12 agricultural commodities from April 11, 2006 to January 3, 2012, daily futures prices from Norma’s Historical Data, and crude oil index positions imputed via Masters’ method and a generalized regression using all 12 agricultural commodities; out-of-sample evaluation covers January 17, 2010 to January 3, 2012.
  • Main analysis uses OLS regressions for levels, 1-week differences, and 13-week differences; daily roll-window regressions; out-of-sample MSE comparisons; and Bai-Perron structural break tests.

Limitations

  • Stated: CFTC index-trader designations have problems but are described as the best high-frequency publicly available data; the CFTC Index Investment Data report is only quarterly and shorter-period, so it is not usable for these regressions; the authors do not infer structural risk-pricing parameters because the key reduced-form coefficients are statistically indistinguishable from zero.
  • Stated: even if index-fund buying affected futures prices, explaining how that would change the equilibrium spot price is a separate challenge.
  • Evident from the text: the sample is limited to post-2006 and specific commodities; imputed crude oil holdings are sensitive to the choice of agricultural commodity; and the in-sample oil relation appears driven by the 2007–2009 recession and breaks down out of sample.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1111/iere.12099

Cites

45 of 45 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 28 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § 5 Predicting returns for crude oil contracts. · p. 27

    Coefficient on x_{t-1}-x_{t-14} is positive over 2006-2008 subsample, as in Singleton and full sample.

    “The coefficient on ˜ x t -1 -˜ x t -14 is positive over the 2006-2008 subsample, as it was found to be in Singleton's regressions and in our regressions using the full sample of observations from 2006 to 2012 as well as the 2006-2010 regression in row 1 of Table 5.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.
  • Findingown data
    § ABSTRACT · p. 2

    The relation between index-fund positions and oil futures returns breaks down out of sample.

    “We review evidence that these positions might help predict changes in oil futures prices, and find that while there is some support for this in the earlier data, this appears to be driven by some of the dramatic features of the 2007-2009 recession, with the relation breaking down out of sample.”
    • Also stated in: § 1 Introduction · p. 4 · plain
      “We find that while such measures still fit in sample, their out-of-sample performance is poor.”
    Quote in the paperSubstantial claim
  • Findingown data
    § 5 Predicting returns for crude oil contracts. · p. 24

    Figure 1 suggests a strong connection between Masters crude oil index holdings and crude oil price, particularly during 2008 and 2009.

    “The figure suggests a strong connection between these two series, particularly during 2008 and 2009.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.
  • Findingown data
    § 6 Conclusion. · p. 30

    Both relations broke down when trying to describe data since 2009.

    “We found, however, that both relations broke down when trying to describe the data since 2009.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.
  • Findingown data
    § 6 Conclusion. · p. 30

    Imputed crude oil positions could help predict crude oil futures returns.

    “We reviewed evidence that positions in crude oil contracts imputed from the reported agricultural holdings could help predict crude oil futures returns”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.
  • May combine two claims. 42
  • Classified as probably not a substantial claim. 27
  • The quote only partly states this claim. 6

  • May combine two claims.

    Findingown data
    § 4 Predicting futures returns for agricultural commodities. · p. 17

    For 12 agricultural commodities, coefficients on change in index notional positions are statistically significant at 5% for cotton and almost for sugar, but opposite signs, and not significant for other 10.

    “The coefficients on the change in index notional positions turn out to be statistically significant at the 5% level for cotton and almost statistically significant for sugar. However, the coefficients are of opposite signs, and neither ˆ φ nor ˆ π are statistically significantly different from zero for any of the other 10 commodities.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 4 Predicting futures returns for agricultural commodities. · p. 19

    For daily returns on near contract, π4 is statistically significant for only 1 of 12 agricultural commodities (soybeans), and of opposite sign predicted by price impact hypothesis.

    “We find a statistically significantly value for ˆ π 4 for only 1 of the 12 agricultural commodities (soybeans), and this is of the opposite sign predicted by the simple price impact hypothesis- if index-fund selling is depressing the price on roll days, one would expect the price of the near contract to fall rather than rise on those days.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Each of these indexes has a defined calendar schedule at which a near contract is sold and the next contract is purchased.
    • Rests on an assumption: The Goldman strategy begins the roll into the next contract on the 5th business day of the month and is completed on the 9th, whereas the Dow Jones strategy begins the roll on the 6th and ends on the 10th.

  • May combine two claims.

    Findingown data
    § 5 Predicting returns for crude oil contracts. · p. 26

    Using all-commodity regression-based estimates, results similar: levels/first-difference not significant, negative R2; 13-week change significant. And same for every bivariate estimate.

    “The results turn out to be quite similar to those based on Masters' original series. Nothing is significant and R 2 are negative for the levels or first-difference specifications, but there is significant predictability from the 13-week change in ˜ X [all] crude oil, t . Interestingly, we found the same is true for every one of the bivariate estimates plotted in Figure 3: the 13-week change in”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Findingnone given
    § 5 Predicting returns for crude oil contracts. · p. 26

    Agricultural positions cannot predict agricultural prices but do predict crude oil prices; direct measure of crude oil index holding would do poorer job at predicting oil prices than index imputed from agricultural holdings.

    “However, it is puzzling that agricultural positions cannot predict agricultural prices but do predict crude oil prices, and that a direct measure of crude oil index holding would do a poorer job at predicting oil prices than does an index imputed from agricultural holdings.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • May combine two claims.

    Findingown data
    § 5 Predicting returns for crude oil contracts. · p. 24

    For crude oil, levels and weekly-difference regressions using Masters index have negative R2 and statistically insignificant coefficients.

    “Both the levels and weekly-difference regression results are similar to those for agricultural commodities, with negative values for R 2 and statistically insignificant coefficients.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • May combine two claims.

    Findingown data
    § 5 Predicting returns for crude oil contracts. · p. 28

    Masters indicator collapses as recession worsened in 2008 but rebounds before recovery, movements precede oil price moves.

    “the striking feature of the Masters indicator is that it collapses as the recession worsened in 2008 but began to rebound sharply before the recovery began, key movements that precede equally dramatic parallel moves in oil prices.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • May combine two claims.

    Findingown data
    § 5 Predicting returns for crude oil contracts. · p. 25

    Biggest adjusted R2 for oil regression is obtained with n=12, close to Singleton's n=13.

    “The biggest R 2 turns out to be obtained by setting n = 12 , close to the value n = 13 proposed by Singleton.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • May combine two claims.

    Findingown data
    § 4 Predicting futures returns for agricultural commodities. · p. 20

    With 1st-10th window, none of coefficients on Sτ are statistically significant for any individual agricultural contracts, and only one significant difference between first two contracts (cotton).

    “For this specification, none of the coefficients on S τ are statistically significant for any of the individual agricultural contracts (columns 2 and 4 of Table 3), and there is only one statistically significant difference between the coefficients on the first two contracts (cotton in column 6).”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Each of these indexes has a defined calendar schedule at which a near contract is sold and the next contract is purchased.
    • Rests on an assumption: The Goldman strategy begins the roll into the next contract on the 5th business day of the month and is completed on the 9th, whereas the Dow Jones strategy begins the roll on the 6th and ends on the 10th.

  • May combine two claims.

    Findingcitation
    § 5 Predicting returns for crude oil contracts. · p. 24

    13-week change in Masters crude oil index appears helpful for predicting changes in crude oil futures prices; this predictive power is for oil but not other 12 commodities.

    “the last row of the first block of Table 4 reproduces Singleton's finding that the 13-week change in ˜ X [Masters] crude oil ,t appears to be quite helpful for predicting changes in crude oil futures prices. Since the use of a 13-week window appears to increase the variable's predictive power for oil returns but not for any of the other 12 commodities, we were curious to look at the performance of”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • May combine two claims.

    Findingcitation
    § 5 Predicting returns for crude oil contracts. · p. 27

    Bai and Perron test leads to conclusion of two structural breaks in oil return regression dated Sept 30, 2008 and Jan 13, 2009, both inside Singleton sample.

    “the Bai and Perron (1998) test leads to the conclusion that there are two structural breaks 10 in the oil return regression (14) dated at September 30, 2008 and January 13, 2009- both inside the original Singleton sample of data.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Bai and Perron (1998)

      Not checked in the free run.

    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • May combine two claims.

    Findingown data
    § 5 Predicting returns for crude oil contracts. · p. 27

    Out-of-sample MSE for oil return prediction using Singleton's regression is 24.01, compared to 21.97 for no-change forecast.

    “If we use those coefficients to predict oil returns over January 17, 2010 through January 3, 2012, the out-of-sample mean squared error is 24.01. That compares with an out-of-sample MSE of 21.97 if we had instead simply always forecast r t = 0 (see the first row of Table 5).”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • May combine two claims.

    Findingown data
    § 4 Predicting futures returns for agricultural commodities. · p. 20

    For return on next minus near, π4 is significantly negative for soybean oil and cotton, and significantly positive for hogs.

    “For soybean oil and cotton we find a statistically significant negative value for π 4 and for hogs we find a statistically significant positive value.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Each of these indexes has a defined calendar schedule at which a near contract is sold and the next contract is purchased.
    • Rests on an assumption: The Goldman strategy begins the roll into the next contract on the 5th business day of the month and is completed on the 9th, whereas the Dow Jones strategy begins the roll on the 6th and ends on the 10th.

  • May combine two claims.

    Findingown data
    § 4 Predicting futures returns for agricultural commodities. · p. 16

    The coefficient estimates for index trader positions are not statistically significantly different from zero for any of the 12 commodities for which index trader positions are reported, and adjusted R2 for these regressions are usually negative.

    “The coefficient estimates ˆ φ 1 and ˆ π 1 are not statistically significantly different from zero for any of the 12 commodities for which index trader positions are reported, and adjusted R 2 for these regressions are usually negative.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § 5 Predicting returns for crude oil contracts. · p. 27

    When estimated with break points, coefficient is negative over second and third subsamples, with t-stat -0.75 for Jan 30, 2009 through Jan 3, 2012.

    “However, when estimated with the two indicated break points, the coefficient turns out to be negative over both the second and third subsamples, with a t statistic of -0 . 75 for data from January 30, 2009 through January 3, 2012.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • May combine two claims.

    Findingown data
    § References · p. 40

    In the in-sample predictability regression for crude oil, the coefficient on Xt-1 is 0.0553 and is significant at the 5% level.

    “| Oil | -0.4858 | -0.1507* | 0.0553* | 0.0834 | 24.0113 | 21.9747 |”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § References · p. 37

    In the regression of average returns during index-fund roll versus normal times, the coefficient on Sτ for the spread of cotton is -0.1920 and is significant at the 5% level.

    “| Cotton | -0.0035 | 0.0731 | 0.0227 | -0.0987 | 0.0249 | -0.1920* |”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § References · p. 38

    In the regression of average returns during the 10-day window around index roll versus normal times, the constant for the spread of cattle is -0.0396 and is significant at the 5% level.

    “| Cattle | 0.0335 | -0.0660 | -0.0061 | -0.0477 | -0.0396* | 0.0182 |”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data
    § References · p. 36

    In the test for predictability of agricultural commodity returns, the coefficient on the 1-week difference of Xt-1 for cotton is 0.2530 and is significant at the 5% level.

    “1-wk diff 0.1601 -0.2249 0.2530* 0.0176”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § References · p. 37

    In the regression of average returns during index-fund roll versus normal times, the coefficient on Sτ for the spread of bean oil is -0.0415 and is significant at the 5% level.

    “| BeanOil | 0.0131 | 0.2267 | 0.0060 | 0.1851 | -0.0071 | -0.0415* |”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § References · p. 39

    In the test for predictability of crude oil returns using the regression method, the coefficient on the 13-week difference of Xt-1 is 0.0448 and is significant at the 5% level.

    “| 13-wk diff | -0.1337 | -0.1240* | 0.0448* | 0.0499 |”
    Quote in the paperNot a substantial claim

  • May combine two claims. The quote only partly states this claim.

    Findingown data
    § References · p. 38

    In the regression of average returns during the 10-day window around index roll versus normal times, the coefficient on Sτ for the second contract of oil is -0.2693 and is significant at the 5% level.

    “| Oil | 0.1334 | -0.3333* | 0.1000 | -0.2693* | -0.0334 | 0.0641 |”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

    Findingown data
    § References · p. 38

    In the regression of average returns during the 10-day window around index roll versus normal times, the constant for the spread of coffee is -0.0167 and is significant at the 5% level.

    “| | 0.0093 | 0.0839 | -0.0074 | 0.0981 | -0.0167* | 0.0142 |”
    Quote in the paperNot a substantial claim

  • May combine two claims. The quote only partly states this claim.

    Findingown data
    § References · p. 37

    In the regression of average returns during index-fund roll versus normal times, the coefficient on Sτ for the second contract of hogs is 0.2428 and is significant at the 5% level.

    “| Hogs | -0.0542 | 0.0896 | -0.0752 | 0.2428* | -0.0210 | 0.1532* |”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § References · p. 37

    In the regression of average returns during index-fund roll versus normal times, the coefficient on Sτ for the second contract of beans is 0.3720 and is significant at the 5% level.

    “| Beans | 0.0195 | 0.3710* | -0.0015 | 0.3720* | -0.0211 | 0.0010 |”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data
    § References · p. 39

    In the test for predictability of crude oil returns using Masters' method, the coefficient on the 13-week difference of Xt-1 is 0.0440 and is significant at the 5% level.

    “| 13-wk diff | -0.2298 | -0.1171 | 0.0440* | 0.0438 |”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims. The quote only partly states this claim.

    Findingown data
    § References · p. 37

    In the regression of average returns during index-fund roll versus normal times, the constant for the spread of coffee is -0.0155 and is significant at the 5% level.

    “| Coffee | 0.0339 | -0.0663 | 0.0184 | -0.0529 | -0.0155* | 0.0133 |”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § References · p. 37

    In the regression of average returns during index-fund roll versus normal times, the coefficient on Sτ for the first contract of beans is 0.3710 and is significant at the 5% level.

    “| Beans | 0.0195 | 0.3710* | -0.0015 | 0.3720* | -0.0211 | 0.0010 |”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § References · p. 39

    In the test for predictability of crude oil returns using the regression method, the coefficient on the 13-week difference of rt-1 is -0.1240 and is significant at the 5% level.

    “| 13-wk diff | -0.1337 | -0.1240* | 0.0448* | 0.0499 |”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data
    § References · p. 40

    In the in-sample predictability regression for the S&P500, the coefficient on Xt-1 is 0.0125 and is significant at the 5% level.

    “| S&P500 | -0.1332 | -0.0637 | 0.0125* | 0.0130 | 6.5520 | 6.3688 |”
    Quote in the paperNot a substantial claim

  • May combine two claims. The quote only partly states this claim.

    Findingown data
    § References · p. 38

    In the regression of average returns during the 10-day window around index roll versus normal times, the coefficient on Sτ for the first contract of oil is -0.3333 and is significant at the 5% level.

    “| Oil | 0.1334 | -0.3333* | 0.1000 | -0.2693* | -0.0334 | 0.0641 |”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § References · p. 37

    In the regression of average returns during index-fund roll versus normal times, the constant for the spread of cattle is -0.0371 and is significant at the 5% level.

    “| Cattle | 0.0260 | -0.0574 | -0.0111 | -0.0445 | -0.0371* | 0.0129 |”
    Quote in the paperNot a substantial claim

  • May combine two claims. The quote only partly states this claim.

    Findingown data
    § References · p. 38

    In the regression of average returns during the 10-day window around index roll versus normal times, the coefficient on Sτ for the spread of cotton is -0.1504 and is significant at the 5% level.

    “| Cotton | -0.0121 | 0.0997 | 0.0194 | -0.0388 | 0.0303 | -0.1504* |”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § References · p. 40

    In the in-sample predictability regression for crude oil, the coefficient on rt-1 is -0.1507 and is significant at the 5% level.

    “| Oil | -0.4858 | -0.1507* | 0.0553* | 0.0834 | 24.0113 | 21.9747 |”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § References · p. 37

    In the regression of average returns during index-fund roll versus normal times, the coefficient on Sτ for the spread of hogs is 0.1532 and is significant at the 5% level.

    “| Hogs | -0.0542 | 0.0896 | -0.0752 | 0.2428* | -0.0210 | 0.1532* |”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationnone given
    § 6 Conclusion. · p. 30

    Methodology for such imputation could be generalized to make use of all available data.

    “and noted that the methodology for such imputation could be generalized to make use of all the available data.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationnone given
    § 6 Conclusion. · p. 30

    This is closely related to dynamics of index investing during Great Recession.

    “though this is closely related to the dynamics of index investing during the Great Recession”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • May combine two claims.

    Interpretationreasoning
    § 4 Predicting futures returns for agricultural commodities. · p. 15

    Swap dealers are implicitly short in futures agreements arranged over-the-counter and hedge with an offsetting long position on organized exchanges that get reported to the CFTC.

    “The swap dealers are thus implicitly short in futures agreements arranged over-the-counter, and hedge with an offsetting long position on organized exchanges that get reported to the CFTC.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § 5 Predicting returns for crude oil contracts. · p. 27

    The correlation identified by Singleton has no success at describing data since his paper was written, and seems not to have captured a stable predictive relation even within his sample.

    “The correlation identified by Singleton thus has no success at describing data since his paper was written, and indeed seems not to have captured a stable predictive relation even within the sample that he analyzed.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § 1 Introduction · p. 3

    Stoll and Whaley (2010) found that in recent years up to half of the open interest in outstanding agricultural commodity futures contracts was held by institutions characterized by the CFTC as commodity index traders.

    “Stoll and Whaley (2010) found that in recent years up to half of the open interest in outstanding agricultural commodity futures contracts was held by institutions characterized by the Commodity Futures Trading Commission (CFTC) as commodity index traders.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Stoll and Whaley (2010)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning
    § 4 Predicting futures returns for agricultural commodities. · p. 16

    The true measure of notional exposure would add a constant to x_t and have no effect on slope coefficients in regressions.

    “Note that X t is measured in number of contracts, so that the true measure of notional exposure would further multiply X t by the number of barrels of oil in a single contract. This would simply add a constant to ˜ x t and would have no effect on the slope coefficients in any of the regressions reported in this section, so we have always used the simpler expression given in the text.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationnone given
    § 3 Index-fund investors and the price of risk. · p. 9

    If index-fund buyers want to take the long side of a futures contract, somebody else must be persuaded to take the short side.

    “If index-fund buyers want to take the long side of the contract, somebody else must be persuaded to take the short side.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationown data
    § 5 Predicting returns for crude oil contracts. · p. 28

    If used only recession data, would conclude index-trader positions could predict stock prices.

    “In other words, if we used only data from the recession, we would conclude that index-trader positions in soybean oil, Kansas City wheat, and feeder cattle could also be used to predict stock prices.”
    Quote in the paperNot a substantial claim
    • The authors state this as an assumption: Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.
    • Rests on an assumption: The CFTC does not report a weekly estimate of index-fund positions in crude oil.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § 2 Previous literature on the effects of index-fund investing on commodity futures prices. · p. 7

    In order to persuade a counterparty to take the other side, the equilibrium price of a futures contract might be pushed below the expected future spot price, producing normal backwardation.

    “In order to persuade a counterparty to take the other side, the equilibrium price of a futures contract might be pushed below the expected future spot price to produce a situation sometimes described as 'normal backwardation.'”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Keynes (1930)

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationreasoning
    § 3 Index-fund investors and the price of risk. · p. 14

    If index-fund positions exert a significant effect on the pricing of futures contracts, a regression using observable notional positions should find a nonzero coefficient on the index-fund position measure.

    “If the latter is exerting a significant effect on the pricing of futures contracts, we should find a nonzero coefficient on π n -1 in (10) when x t is based on observable measures of the notional positions taken by index-fund traders.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § 2 Previous literature on the effects of index-fund investing on commodity futures prices. · p. 7

    Keynes (1930) proposed that risk premia in commodity futures prices could arise from the desire of producers of the physical commodity to hedge their price risk by selling futures contracts.

    “Keynes (1930) proposed that risk premia in commodity futures prices could arise from the desire of producers of the physical commodity to hedge their price risk by selling futures contracts.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Keynes (1930)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable). May combine two claims.

    Study design and conductcitation
    § 4 Predicting futures returns for agricultural commodities. · p. 15

    While the CFTC designations are not without some problems, they appear to be the best high-frequency data that are publicly available for the purposes of this study.

    “While the CFTC designations are not without some problems (Cheng, Kirilenko, and Xiong (2012), Stoll and Whaley (2010)), these appear to be the best high-frequency data that are publicly available for our purposes.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Cheng, Kirilenko, and Xiong (2012)

      Not checked in the free run.

    • cites Stoll and Whaley (2010)

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation
    § 2 Previous literature on the effects of index-fund investing on commodity futures prices. · p. 8

    Limited working capital on the part of potential arbitrageurs is the key factor determining how much the futures price might deviate from the expected future spot price.

    “Etula (forthcoming), Acharya, Lochstoer, and Ramadorai (forthcoming), Danielsson, Shin, and Zigrand (2011), and Cheng, Kirilenko, and Xiong (2012) have stressed the role of limited working capital on the part of potential arbitrageurs as the key factor determining how much the futures price might deviate from the expected future spot price.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Etula (forthcoming)

      Not checked in the free run.

    • cites Acharya, Lochstoer, and Ramadorai (forthcoming)

      Not checked in the free run.

    • cites Danielsson, Shin, and Zigrand (2011)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S1 (asserted), S3 (load-bearing).

    Premisecitation
    § 2 Previous literature on the effects of index-fund investing on commodity futures prices. · p. 8

    The growing volume of commodity index investors could produce hedging price pressure on the buy side.

    “Brunetti and Reiffen (2011), Acharya, Lochstoer, and Ramadorai (forthcoming), Cheng, Kirilenko, and Xiong (2012), and Hamilton and Wu (forthcoming) proposed that the growing volume of commodity index investors could produce hedging price pressure on the buy side”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Brunetti and Reiffen (2011)

      Not checked in the free run.

    • cites Acharya, Lochstoer, and Ramadorai (forthcoming)

      Not checked in the free run.

    • cites Cheng, Kirilenko, and Xiong (2012)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § 1 Introduction · p. 3

    Barclays Capital estimated that exchange traded financial products following index strategies grew from negligible amounts in 2003 to a quarter trillion dollars by 2008.

    “Barclays Capital estimated that exchange traded financial products following such strategies grew from negligible amounts in 2003 to a quarter trillion dollars by 2008 (Irwin and Sanders (2011)).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Irwin and Sanders (2011)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § 2 Previous literature on the effects of index-fund investing on commodity futures prices. · p. 7

    Evidence on whether normal backwardation holds as a general characterization of commodity futures risk premia is mixed.

    “Evidence on backwardation is mixed. Carter, Rausser, and Schmitz (1983), Chang (1985), Bessembinder (1992), and De Roon, Nijman, and Veld (2000) provided empirical support for such an interpretation of the risk premium in commodity futures, while Marcus (1984), Hartzmark (1987), and Kolb (1992) concluded that it does not hold as a general characterization.”
    Quote in the paperNot a substantial claim0 of 3 sources checked
    • cites Carter, Rausser, and Schmitz (1983)

      Not checked in the free run.

    • cites Chang (1985)

      Not checked in the free run.

    • cites Bessembinder (1992)

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation
    § 2 Previous literature on the effects of index-fund investing on commodity futures prices. · p. 8

    Cootner (1960) argued that hedging pressure could lead to expected returns for positions in futures contracts as a necessary inducement to potential counterparties to take the other side from hedgers.

    “Cootner (1960) argued more generally that hedging pressure (of which the desire by Keynes's producers to sell forward is one example) could lead to expected returns for positions in futures contracts as a necessary inducement to potential counterparties to take the other side of the contract from the hedgers.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Cootner (1960)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationcitation
    § 4 Predicting futures returns for agricultural commodities. · p. 16

    This result is consistent with the large number of previous studies that have found limited predictability of commodity futures returns using related regressions.

    “This result is consistent with the large number of previous studies discussed in Section 2 that have found limited predictability of commodity futures returns using related regressions.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation
    § 2 Previous literature on the effects of index-fund investing on commodity futures prices. · p. 8

    Hamilton and Wu (forthcoming) found that the average compensation to the long position in oil futures contracts has decreased but become substantially more volatile since 2005.

    “with Hamilton and Wu (forthcoming) finding that the average compensation to the long position in oil futures contracts has decreased but become substantially more volatile since 2005.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Hamilton and Wu (forthcoming)

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing), S4 (based).

    Interpretationnone given
    § 6 Conclusion. · p. 30

    Even if one could demonstrate an effect of index-fund buying on commodity futures prices, it would be a separate challenge to explain how this could also end up changing equilibrium spot price.

    “even if one could demonstrate an effect of indexfund buying on commodity futures prices, it would be a separate challenge to explain how this could also end up changing the equilibrium spot price.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Interpretationreasoning
    § ABSTRACT · p. 2

    According to a simple model of futures arbitrage, if index-fund buying influences prices by changing the risk premium, then the notional positions of index investors should help predict excess returns in these contracts.

    “according to a simple model of futures arbitrage, if index-fund buying influences prices by changing the risk premium, then the notional positions of the index investors should help predict excess returns in these contracts.”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: The arbitrageur cares about the mean and variance of their composite portfolio.
    • The authors state this as an assumption: The arbitrageur also invests amounts in assets j=0,...,J, with asset 0 risk free.
    • The authors state this as an assumption: In equilibrium log asset returns and commodity futures prices are affine functions of a vector of factors x_t.
    • The authors state this as an assumption: The factors follow a Gaussian vector autoregression.
    • The authors state this as an assumption: In equilibrium arbitrageurs' positions are affine functions of the underlying factors.

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation
    § 5 Predicting returns for crude oil contracts. · p. 22

    The Dow Jones index holds soybean oil, while the Goldman index does not.

    “Masters (2008) noted that while the Dow Jones index holds soybean oil, the Goldman index does not”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Masters (2008)

      Not checked in the free run.

  • May combine two claims.

    Premisecitation
    § 5 Predicting returns for crude oil contracts. · p. 23

    The Goldman index holds Kansas City wheat whereas Dow Jones does not.

    “Similarly, the Goldman index holds Kansas City wheat whereas Dow Jones does not”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Masters (2008)

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation
    § 5 Predicting returns for crude oil contracts. · p. 22

    Each of these indexes would take positions in a particular set of commodities and roll them over according to a prespecified calendar schedule.

    “Each of these would take positions in a particular set of commodities and roll them over according to a prespecified calendar schedule.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Masters (2008)

      Not checked in the free run.

58 of 58

  • The authors state this as an assumption.

    The arbitrageur cares about the mean and variance of their composite portfolio.

    “We will refer to the index fund's counterparty as an 'arbitrageur', and assume that what the arbitrageur cares about is the mean and variance of their composite portfolio.”

  • The authors state this as an assumption.

    The arbitrageur also invests amounts in assets j=0,...,J, with asset 0 risk free.

    “In addition to potential positions in a variety of futures contracts, we presume that the arbitrageur also invests amounts q jt in assets j = 0 , 1 , ..., J (where asset j = 0 is presumed to be risk free).”

  • The authors state this as an assumption.

    In equilibrium log asset returns and commodity futures prices are affine functions of a vector of factors x_t.

    “We also conjecture that in equilibrium log asset returns and commodity futures prices are affine functions of a vector of factors x t ,”

  • The authors state this as an assumption.

    The factors follow a Gaussian vector autoregression.

    “where the factors themselves can be described using a Gaussian vector autoregression:”

  • The authors state this as an assumption.

    In equilibrium arbitrageurs' positions are affine functions of the underlying factors.

    “If we further conjecture that in equilibrium these positions are also affine functions of the underlying factors,”

  • The authors state this as an assumption.

    Index-fund positions in agricultural commodity i are composed entirely of either traders following the Goldman strategy or traders following the Dow Jones strategy.

    “Suppose we assume that this is composed entirely of either traders following the Goldman strategy, whose notional holdings we denote by ˜ X G it , or traders following the Dow Jones strategy ( ˜ X D it ):”

  • The paper relies on this without support.

    The CFTC does not report a weekly estimate of index-fund positions in crude oil.

    “The CFTC does not report a weekly estimate of index-fund positions in crude oil.”

  • The paper relies on this without support.

    Each of these indexes has a defined calendar schedule at which a near contract is sold and the next contract is purchased.

    “Each of these has a defined calendar schedule at which a near contract is sold and the next contract is purchased.”

  • The paper relies on this without support.

    The Goldman strategy begins the roll into the next contract on the 5th business day of the month and is completed on the 9th, whereas the Dow Jones strategy begins the roll on the 6th and ends on the 10th.

    “The Goldman strategy begins the roll into the next contract on the 5th business day of the month and is completed on the 9th, whereas the Dow Jones strategy begins the roll on the 6th and ends on the 10th.”
9 of 9

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The long positions maintained by index funds in near-term futures contracts can create upward pressure on prices, particularly when combined with other market participants' trading strategies.

    “They emphasize that the long positions maintained by index funds in near-term futures contracts can create upward pressure on prices, particularly when combined with other market participants' trading strategies.”

    From Food Financialization: Impact of Derivatives and Index Funds on Agri-Food Market Volatility (2024) · cites this paper as Hamilton and Wu (2015)

    The Introduction (p. 3) describes index funds taking long positions in near-term futures, and Section 3 (pp. 11–12) models how increased buying could raise futures prices through the risk premium. Section 3 (p. 14) also considers commercial hedgers’ positions in the net buying pressure facing arbitrageurs, but it does not establish that combining index-fund positions with other participants’ trading strategies particularly increases upward pressure. The paper’s empirical discussion reports little evidence of a price effect. (LLM, conf=0.82)

    Experimental — accuracy not yet measured

  • Hamilton and Wu (2015) analyze the effects of index-fund investing on commodity prices, and their findings emphasize the complexity of this relationship and the challenge of regulating financialization without negatively impacting market efficiency.

    “Hamilton and Wu (2015) analyze the effects of index-fund investing, which are a crucial part of financialization, on commodity prices. Their findings emphasize the complexity of this relationship and the challenge of regulating financialization without negatively impacting market efficiency”

    From Food Financialization: Impact of Derivatives and Index Funds on Agri-Food Market Volatility (2024) · cites this paper as Hamilton and Wu (2015)

    Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

    Experimental — accuracy not yet measured

  • The findings of Hamilton and Wu (2015) directly support the need for agro-food companies to develop robust strategies for managing price risks that extend beyond traditional methods.

    “the findings of Hamilton and Wu (2015) directly support the need for agro-food companies to develop robust strategies for managing price risks that extend beyond traditional methods.”

    From Food Financialization: Impact of Derivatives and Index Funds on Agri-Food Market Volatility (2024) · cites this paper as Hamilton and Wu (2015)

    Section 4 (pp. 16–17) finds no persuasive evidence that index-trader positions are related to risk premia in agricultural commodities. Our copy does not address whether agro-food companies need price-risk strategies beyond traditional methods; searches for company strategies and hedging recommendations found no such passage. (LLM, conf=0.75)

    Experimental — accuracy not yet measured

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Conflicting definitions of the oil-regression predictor

    Minor

    The text says the oil regressions use 100 times the log of imputed holdings, while the notes to Table 4 define the predictor as the log without the factor of 100. That changes the scale of the reported slope coefficients.

1 of 1

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. Keynes J. M., 1930, A Treatise on Money

Cited by

Sorted by citations

Related

Authors
Hamilton, J. D., & Wu, J. C.
Affiliations
University of California San Diego (US)
Corresponding author
Wu, Jing Cynthia
ORCID
Venue
International Economic Review
Publisher
Wiley
Published
2015-01-23
Volume
56
Issue
1
Pages
187-205
ISSN
  • 0020-6598 (ISSN-L)
  • 1468-2354
DOI
10.1111/iere.12099
OpenAlex
W2333630357
Semantic Scholar
67f761e2c4d86198c77bb3c824163afe1a3e8b68
Cited by
241
References
45
Citations by year
2026: 7 · 2025: 18 · 2024: 12 · 2023: 21 · 2022: 22 · 2021: 23 · 2020: 18 · 2019: 34 · 2018: 31 · 2017: 21 · 2016: 15 · 2015: 9 · 2014: 4 · 2013: 4 · 2012: 2
FWCI
48.39
Type
journal article
Language
en
Open access
closed
Licence
onlinelibrary.wiley.com/termsAndConditions#vor (vor)
Locations
  • International Economic Review (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
Topics
  • Social Sciences › Economics, Econometrics and Finance › Finance › Financial Markets and Investment Strategies
  • Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Market Dynamics and Volatility
  • Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Economics of Agriculture and Food Markets
Keywords
commodity futures prices, agricultural futures, oil futures, abnormal returns, out-of-sample forecasting

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review