Chia-Hung Liu  Yu-Lun Chen  Min-Hsi Chung  Ya-Kai Chang 

Traders’ Concentration, Hedging Pressure, and Risk

Premium in Futures Markets
(Received Jul 04, 2022; First Revision Dec 26, 2022; Second Revision Feb 03, 2023;
Accepted Mar 20, 2023)

Purpose – This study explores the relationship between market concentration and risk premiums in the
U.S. futures markets from 2010 to 2018. When studying the determinants of risk premiums in
futures markets, existing literature focuses mainly on the risk-averse behavior of hedgers.

Design/methodology/approach – This study provides a new measure of market concentration, which

measures the amount of information content held by traders in futures markets.

Findings – 1. Market concentration and hedging pressure coexist in explaining the determinants of risk
premiums in various futures markets. 2. Market concentration and hedging pressures have no
short-term impact on futures risk premiums. As the holding period increases, the impact of market
concentration and hedging pressures on futures risk premiums will increase.

Research limitations/implications – Market concentration does not distinguish between commercial

and non-commercial traders, so the largest traders may be classified as either commercial or non-
commercial traders.

Practical implications/Social implications – Market concentration provides an alternative measure of

risk premiums in the futures markets. For some futures products that cannot be explained by
hedging pressure, changes in risk premiums can be predicted through market concentration.

Originality/value – This study will provide a lot of practical help for market investors who set up
investment strategies and policymakers who establish trading rules to prevent market manipulation,
thereby helping establish a more efficient and fair trading environment in futures markets.

Keywords – Futures, Concentration, Risk premium, Hedging pressure


Chia-Hung Liu is a Master of Department of Finance, College of Business, Chung Yuan Christian University,
Taiwan.

Yu-Lun Chen is a professor of Department of Finance, College of Business, Chung Yuan Christian
University, Taiwan.

Min-Hsi Chung is a Ph.D. candidate of Department of Finance, College of Commerce, National Chengchi
University, Taiwan.

Ya-Kai Chang is an associate professor of Department of Finance, College of Business, Chung Yuan
Christian University, Taiwan. (Corresponding Author) Email:ykchang@cycu.edu.tw
DOI:10.6656/MR.202307_42(3).ENG002

Management Review
Vol.42 (Jul 2023), 115-138
1. Research Motivation and Purpose
In the financial market, issues such as international market instability and supply chain changes will frequently
cause changes to stock market prices, which can impact both weighted stock price indexes and individual stocks. To
respond to these unpredictable occurrences in the market, derivative financial products such as forward contracts, futures,
and options have emerged. As to the prices of these derivative financial products, regarding the futures market, the cost
of carry model proposed by Cornell and French (1983a, b) takes into account warehousing cost, as well as stock,
transportation, insurance, interest, and tax costs of products, then uses these factors to calculate the theoretical prices of
futures.
However, reality often differs from theory. The theoretical prices calculated by the cost of carry model are not
actually consistent with actual prices, implying that there are other reasons for such a discrepancy between futures and
spot prices. Past research findings have indicated that the difference between futures and spot prices in the futures market
is due to varying degrees of risk premium (De Roon et al., 2000). De Roon et al. (2000) pointed out that the measurement
of risk premium originates from the difference between future spot prices and current futures prices. Because in theory
future spot prices and current futures prices should converge, the return gained from futures products is a good
measurement index for futures risk premium. Consequently, this study used futures return as a proxy variable for risk
premium to explore the impact of hedging pressure and trader concentration on risk premium.
Discussion on futures risk premium can be traced as far back as the backwardation theory of Keynes (1930) and
Hicks (1939), where through researching the agricultural production market, it was believed that hedgers basically
assume a net short position in the futures market. This, as a result, will drive down futures prices, cause prices to fall
below expected prices, and lead to mispricing in the futures market. Because a large stock of agricultural products will
be harvested in the future, this will also result in immense spot pressure. On the other hand, the research of Houthakker
(1957) and Cootner (1960) believed that the backwardation theory can also be applied to the net long position, in which
case futures prices are driven above expected prices. In such futures contracts, speculators play the role opposite to that
of hedgers through taking risks and expecting to get positive returns. Therefore, speculators can cause deviated futures
prices to revert to reasonable, expected prices, an effect that forms the basis for the hedging pressure theory.
In continuation of the hedging pressure theory, the research of Dusak (1973), Black (1976), and Jagannathan (1985)
found that the net positions of hedgers in the futures market can be used to measure hedging pressure. Hedging pressure
arises from the fact that hedgers are reluctant to take on time and transaction risks. To secure future interests, they will
forgo the opportunity to obtain higher returns, and will either use the method of selling futures to guarantee personal
interest, or purchasing futures to avoid rises in raw material costs by way of fixing production costs. In addition, when
speculators expect that future spot prices will diverge from known futures prices, they will invest in the futures market
for arbitrage.
When constructing a model for impact on futures prices, Stoll (1979) and Hirshleifer (1988) combined the two
factors of hedging pressure and systematic risk to provide better model explanatory power. When considering systematic
risk, Carter et al. (1983) and Bessembinder (1992) used the indicator of hedging pressure to conduct empirical research
on futures prices, finding that hedging pressure will indeed influence futures prices and determining hedging pressure
to be a factor that can measure changes in futures prices. De Roon et al. (2000) further conducted cross-market research,
separating markets into four categories (the financial, agricultural, mineral, and currency markets), finding that the
prices of futures are not only influenced by the hedging pressure in their respective markets, but also by cross-hedging
pressure from different markets. In view of the above literature, this study carried out both same-market and cross-
market research, assuming that hedging pressure has a significant positive correlation with futures risk premium. In
addition, De Roon et al. (2000) studied the relationship between hedging pressure and futures prices in the period from
January 1986 to December 1994, 25 years before this study, making it necessary to demonstrate the applicability of past
research findings to the current situation of the futures market. As a result, this study first determined the validity of
past empirical results in a different research period.
In regard to literature on futures prices, in addition to the two factors of systematic risk and hedging pressure, other
scholars have used the amount of information owned by traders as a basis for measuring risk premium. Researchers in
this field believe that traders can more accurately predict product prices due to having a better understanding of market
conditions. Consequently, when unreasonable prices arise in the futures market, traders tend to choose to invest in the
market to acquire profits. For example, Stulz (1996) and Knill et al. (2006) believe that the routine business activities
of traders give them an advantage over other market participants by allowing them to more accurately predict price
fluctuations and therefore receive positive returns in the futures market.
Information asymmetry and product price formation make up the core of market microstructure research.
Regarding the measurement of information asymmetry in the futures market, Knill et al. (2006) used analysts’ revenue
forecasts for energy companies to measure the degree of information asymmetry in the futures market, finding that
degree of information asymmetry influences futures market prices. Concentration has been a factor seldom discussed
in research on futures markets in the past. When the open interest of futures contracts is concentrated in the hands of a
small number of buyers (sellers), this indicates that a small number of buyers (sellers) hold relatively greater market

116 Management Review, Jul 2023

power and that traders may be able to manipulate futures prices to obtain profit and may also imply the forming of
information asymmetry. This paper constitutes the first study to explore the influence of futures contract concentration
on risk premium.
This study used market concentration as a basis to discuss whether information asymmetry exists in the futures
market and aims to determine whether trader concentration influences futures market prices. Changes in trader
concentration may suggest that traders have obtained private information, which allows them to increase their net long
(net short) futures positions and subsequently obtain positive returns in the futures market. This study used the
concentration data of the Commitments of Traders (COT) report of the Commodity Futures Trading Commission (CFTC)
to find out whether information asymmetry exists in the US futures market. Concentration of traders was measured by
the difference between the concentration of net long positions and net short positions, When the concentration of traders
is positive (negative) in value, this means that futures prices will increase (decrease). Based on the above literature, we
can therefore expect that trader concentration has a significant positive correlation with futures risk premium.
Lastly, because the hedging pressure theory established by the research of Keynes (1930), Hicks (1939),
Houthakker (1957), and Cootner (1960) did not consider information asymmetry, the explanatory powers for risk
premium of hedging pressure and information asymmetry may differ under different futures markets. For example,
while spot pressure is unlikely to occur in financial markets, the expected harvesting times for crops in agricultural
markets will, in comparison, produce much greater hedging pressure. Therefore, this study asserts that discussion on
the factors that influence risk premium should simultaneously consider information asymmetry and hedging pressure.
In summary, there has yet to be any scholarship that has used market concentration to measure information asymmetry
or to simultaneously include information asymmetry and hedging pressure in models when discussing the influence of
information asymmetry and hedging pressure regarding risk premium. Whether changes in the market concentration of
informed traders can be used as a factor to explain futures risk premium is thus the primary objective of this study.
Moreover, whether the market concentration used in this study has better explanatory power for futures risk premium
than the hedging pressure focused on in past studies is also another point of emphasis of this study.
After the testing of the subjects selected for this study (the US’s financial futures market, agricultural futures market,
mineral futures market, and currency futures markets for the period from 2010 to 2018), we found that the market
concentration of the top four traders influences the prices of futures. Because traders who hold more futures contract
positions usually have more market information, they have a greater ability to predict price fluctuations. Additionally,
after the market concentration indicator and the hedging pressure indicator used by past scholars were put into the same
regression model, the predictiveness for futures prices showed a significant increase. The market concentration used in
this study thus provides another indicator for past literature to use in the explanation of futures return, whereby the
changes in futures return for some futures products that were unable to be explained using hedging pressure can be
predicted through market concentration. Lastly, this study found that for both hedging pressure and market concentration,
long-term predictions are always advantageous to short-term predictions, confirming that hedgers will invest in the
futures market to hedge in order to avoid risk and that information response does not occur in a short period of time.
The above results are expected to provide a significant amount of practical assistance to the devising of futures
investment strategies by market investors, formulation of trade regulations by policymakers, and prevention of market
manipulation, and to help to establish a more efficient and fair-trading environment for the futures market. The
subsequent framework for this study is as follows: the second chapter deals with sources; the third chapter is an
introduction to research methods; the fourth chapter presents empirical results and analysis; and the fifth chapter is a
conclusion based on the empirical results collected by this study.

2. Sources
2.1 Samples and Information Sources
The research period for this study was from January 1, 2010, to December 31, 2018, averting the 2008 financial
crisis to prevent the systematic influence of this period from affecting the research results. CFTC is the competent
authority for futures transactions, while the COT is a report published by the CFTC every Friday that aggregates the
open interest distributions of traders from when the market closes on Tuesday of the same week. Because the market
concentration data from the CFTC’s COT reports are provided once a week, the data frequency for this study is weekly.
The hedging pressure and concentration data used in this study were obtained from the CFTC’s COT reports, while the
recent month price data for all futures was obtained from Datastream. The futures markets in this study mainly included
the following: the financial futures market (including S&P index, T-bonds, and Eurodollars), the agricultural futures
market (including wheat, corn, soybean, and live cattle), the mineral futures market (including gold, silver, platinum,
and crude oil), and the currency futures market (including the Swiss franc, British pound, Japanese yen, and Canadian
dollar). The above research targets were chosen based on the research of De Roon et al. (2000) and constitute active and
important representatives for the futures of each market. COT reports have also been used by much past research on
hedging pressure for the carrying out of analysis (Chang et al., 2013; Chen & Yang, 2021; De Roon et al., 2000; Smales,
2016; Wang, 2003).

Market Concentration and Risk Premium 117

2.2 CFTC’s COT Reports
2.2.1 Commercial and non-commercial positions
When traders buy and sell futures products, traders that use the futures contracts for some products to engage in
hedging behaviors as defined under CFTC Regulation 1.3, 17 CFR 1.3(z) have all their reported futures positions
classified as commercial positions, whereas traders not categorized as commercial traders are classified in the report as
non-commercial traders. The COT example provided in Figure 1 can be referenced as an example.

Figure 1
COT sample (Adapted from CFTC website)

2.2.2 The net position concentrations of the four top traders

The CFTC’s COT reports point out that the percentage of open interest held by the four top traders for each futures
product, regardless of whether the nature of transactions is commercial or non-commercial, is divided into gross long
positions and gross short positions. Concentration data is obtained through the net long positions and net short positions,
whereby net position concentration is the value obtained after gross long position and gross short position have offset
each other for the futures product of a given trader. Because trader concentration in regard to the 4 top traders does not
differentiate between commercial and non-commercial traders, the top traders may fall under either commercial or non-
commercial classifications.
2.3 Definition and Description of Variables
2.3.1 Hedging Pressure Indicator
In the formula for the measurement of hedging pressure proposed by De Roon et al. (2000), commercial short
position is subtracted from long position and then divided by the total number of positions to obtain the hedging pressure
indicator, shown in Formula 1 below.
(𝐿𝑜𝑛𝑔 𝑃𝑜𝑠𝑖𝑡𝑖𝑜𝑛𝑠 −𝑆ℎ𝑜𝑟𝑡 𝑃𝑜𝑠𝑖𝑡𝑖𝑜𝑛𝑠 )
𝑞𝑖,𝑡 = (𝐿𝑜𝑛𝑔 𝑃𝑜𝑠𝑖𝑡𝑖𝑜𝑛𝑠𝑖,𝑡 +𝑆ℎ𝑜𝑟𝑡 𝑃𝑜𝑠𝑖𝑡𝑖𝑜𝑛𝑠𝑖,𝑡 ) (1)
𝑖,𝑡 𝑖,𝑡

In Formula 1, 𝐿𝑜𝑛𝑔 𝑃𝑜𝑠𝑖𝑡𝑖𝑜𝑛𝑠𝑖,𝑡 represents the commercial long positions held by futures contract 𝑖 at time t, and
𝑆ℎ𝑜𝑟𝑡 𝑃𝑜𝑠𝑖𝑡𝑖𝑜𝑛𝑠𝑖,𝑡 represents the commercial short positions held by futures contract 𝑖 at time t.
2.3.2 Futures market concentration
When new market information is released, all investors will accordingly determine their response strategies. Using
the positive (negative) information in the stock market as an example, investors may collectively enter the long (short)
positions of the futures market. In such circumstances, although the total number of net long positions (short positions)
will increase, the market concentration of the top four traders is unlikely to change because the proportion of net long
(net short) positions held by traders remained unaffected. On the other hand, when new market information contains
uncertainties, risk-seeking investors may respectively enter long positions or short positions in the futures market, while
risk-averse investors may choose to liquidate their open positions. In a situation such as this, the trader concentration of
net long positions and net short positions may simultaneously increase. However, traders who hold more positions often
tend to have more market information. When highly exclusive information is acquired, they are likely to change the
quantity of their net long or short positions but other traders will, as a rule, not follow suit, causing changes to market

118 Management Review, Jul 2023

concentration. As a result, this paper referenced the method of De Roon et al. (2000) to obtain an indicator for the
measurement of trader concentration, for which the concentration of net short positions for the previous top four traders
in the CFTC’s COT report was subtracted from the concentration of net long positions of the previous top four traders
and the resulting figure standardized by dividing it by the concentration for all positions. The formula for the
measurement of market concentration is shown below:
(𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛_𝑁𝑒𝑡 𝐿𝑜𝑛𝑔 −𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛_𝑁𝑒𝑡 𝑆ℎ𝑜𝑟𝑡 )
𝐶𝑜𝑛𝑖,𝑡 = (𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛_𝑁𝑒𝑡 𝐿𝑜𝑛𝑔𝑖,𝑡 +𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛_𝑁𝑒𝑡 𝑆ℎ𝑜𝑟𝑡𝑖,𝑡 ) (2)
𝑖,𝑡 𝑖,𝑡

In formula 2, 𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛_𝑁𝑒𝑡 𝐿𝑜𝑛𝑔𝑖,𝑡 represents the concentration of net long positions held by futures
contract 𝑖 at time t, while 𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛_𝑁𝑒𝑡 𝑆ℎ𝑜𝑟𝑡𝑖,𝑡 represents the concentration of net short positions held by
futures contract 𝑖 at time t.

3. Research Methods and Model

This paper primarily used the model of De Roon et al. (2000) as its basis. De Roon et al. (2000) believed that
futures risk premium is influenced by market risk and hedging pressure from the same market. In addition, De Roon et
al. (2000) further proposed that, besides own-market hedging pressure, futures risk premium is influenced by cross-
market hedging pressure. De Roon et al. (2000) pointed out that the measurement of risk premium originates from the
difference between future spot prices and current futures prices. Because, in theory, future spot prices and current futures
prices should converge, the return gained from futures products is a good measurement index for futures risk premium.
In addition, futures do not have transaction costs during the initial period. According to the investment portfolio theory,
in the exclusion of risk-free assets, excess returns are equivalent to futures returns and risk premium refers to past
average excess returns. Consequently, this study used futures return as a proxy variable for risk premium to explore the
impact of hedging pressure and concentration on risk premium.
To validate the hedging pressure hypothesis as well as consider short, mid, and long-term changes (including the
changes after one week, two weeks, and four weeks) to subsequent futures returns, we referenced De Roon et al. (2000)
to establish the following empirical model:
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗)
𝑟𝑖,𝑡+𝑘 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + ∑𝑛𝑆=1 𝜃𝑠,𝑖 𝑞𝑠,𝑡 + 𝜀𝑖,𝑡+𝑘 (3)
In formula (3), (𝑗)=1,2,3,4 represents the 4 regression models generated for the financial, agricultural, mineral, and
currency futures markets, respectively. For example, (𝑗)=1 represents the financial futures market, where 𝑠 stands for
S&P index, T-bonds, and Eurodollars; (𝑗)=2 stands for the agricultural futures market, where 𝑠 stands for wheat, corn,
soybean, and live cattle; (𝑗)=3 represents the mineral futures market, where 𝑠 stands for gold, silver, platinum, and crude
oil; and (𝑗)=4 represents the currency futures market, where 𝑠 stands for Swissfrancs, British pounds, Japanese yen, and
Canadian dollars. 𝑟𝑖,𝑡+𝑘 refers to the recent month futures return rate for futures contract 𝑖 at time t + k where k=1, 2,
𝑠&𝑝500 (𝑗)
and 4 weeks; 𝛼0 is the intercept term; 𝑟𝑚𝑡 is the return rate for the S&P 500 index at time t; 𝑞𝑠,𝑡 is the hedging
(𝑗)
pressure of futures contract 𝑠 at time t; and 𝜀𝑖,𝑡+𝑘 is the error term for futures contract 𝑖. This study expects to accept
Hypothesis 1, that hedging pressure has a positive correlation with futures risk premium (futures return rate).
To validate the hypothesis of information asymmetry, we referenced De Roon et al. (2000) to establish the
following empirical model:
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗)
𝑟𝑖,𝑡+𝑘 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + ∑𝑛𝑆=1 𝛾𝑠,𝑖 𝐶𝑜𝑛𝑠,𝑡 + 𝜀𝑖,𝑡+𝑘 (4)
(𝑗)
In formula (4), 𝐶𝑜𝑛𝑠,𝑡 is the market concentration for futures contract 𝑠 at time t, while the definitions for the other
variables are the same as above. This study assumes that Hypothesis 2 is correct, and that market concentration has a
positive correlation with futures risk premium (futures return rate). In line with past research, hedging pressure (Model
3) took into consideration the intrinsic cross-market variables for cross-hedging, while this paper was the first to propose
information asymmetry (Model 4).
To simultaneously validate the hedging pressure hypothesis and the information asymmetry hypothesis, we
adjusted the model of De Roon et al. (2000) as follows:
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗) (𝑗)
𝑟𝑖,𝑡+𝑘 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + 𝛾𝑖 𝐶𝑜𝑛𝑖,𝑡 + ∑𝑛𝑆=1 𝜃𝑠,𝑖 𝑞𝑠,𝑡 + 𝜀𝑖,𝑡+𝑘 (5)
In formula (5), 𝐶𝑜𝑛𝑖,𝑡 refers to the market concentration for futures contract 𝑖 at time t, while the definitions of the
other variables are the same as above. This study assumes Hypotheses 3 is correct, and that market concentration and
hedging pressure are positively correlated with futures risk premium (futures return rate). Model 5 is a combination of
Model 3 and Model 4, in which, to verify both the hedging pressure hypothesis and information asymmetry hypothesis,
hedging pressure (Model 3) took into consideration the intrinsic cross-market variables for cross-hedging proposed by
past research, while this paper was the first to propose information asymmetry (Model 4). Because currently no research

Market Concentration and Risk Premium 119

exists for cross-market information asymmetry, Model 5 was simplified to only include the market concentrations for
each futures market.

4. Empirical Results and Analysis

4.1 Data Analysis
Table 1, Descriptive Statistics - Futures Risk Premiums, shows the futures risk premiums for the period from 2010
to 2018. In terms of the financial futures market, the average futures return for the S&P 500 was 0.0069 and the average
futures return for T-bond was 0.0018, indicating that during this period the average futures risk premiums for S&P 500
and T-bond was positive; on the other hand, the average futures risk premium for Eurodollars was slightly negative. In
terms of the agricultural industry, the average risk premiums for corn and soybean futures were negative, while
compared to corn and soybean, the average futures risk premium for live cattle exhibited a large positive value. In terms
of the mineral futures market, apart from gold, which showed a positive futures risk premium, the remaining futures
products all showed negative average risk premiums, indicating that the risk premiums for futures in the minerals market
are over the long-term generally negative. In terms of the currency futures market, aside from Swiss francs, which
showed a positive futures risk premium, the remaining futures products all showed negative average risk premiums,
indicating that the risk premiums for futures in the foreign currency market are also over the long-term generally
negative. Looking at the standard deviation column, it can be seen that the standard deviations for the agricultural and
mineral futures markets were generally higher than those of the financial and currency markets, indicating that, for the
agricultural and mineral markets, the distribution of risk premium is relatively dispersed and has high volatility.
To verify whether sequences were stationary, this paper conducted the unit root test for all risk premium sequences.
Table 1, Descriptive Statistics - Futures Risk Premiums, shows that no sequence had a single root problem, indicating
that all risk premium sequences were stationary. In addition, to inspect whether the problem of collinearity existed for
the variables in each market, collinearity testing was conducted on all of the variables in each market. A correlation
coefficient between variables exceeding 0.8 indicates that problems of collinearity may exist (Gujarati, 2003). From
Table 1, it can be seen that the absolute values of the average correlation coefficients for all variables in different markets
was less than 0.8, indicating that no issues of collinearity existed between variables.
From Table 2, Descriptive Statistics - Hedging Pressure, the following is known. In terms of the financial futures
market, S&P 500, T-bonds, and Eurodollars all had positive average hedging pressure values, indicating that commercial
traders in the financial futures market primarily held long positions. In terms of the agricultural futures market, except
for wheat futures, which had a positive average value, the average values for corn, soybean, and live cattle were all
negative, indicating that commercial traders in the agricultural futures market held more short positions than long
positions. This may be because producers of crop futures have relatively high expected spot pressure, and as a result,
hold short positions in the futures market in hopes that buyers will purchase products once they are harvested. In terms
of the mineral futures market, the average hedging pressure values for all four mineral futures products were negative,
and the coefficients were also higher than those of the other three futures markets. In terms of the currency futures
market, besides Canadian dollar futures having a negative average hedging pressure value, the average hedging pressure
values for the remaining three currency futures products were all positive. Looking at the standard deviation column,
the standard deviation for the currency futures market was higher than those of the other three markets, indicating that
more changes occur in the positions of traders in the currency market.
From Table 3, Descriptive Statistics - Market Concentration, the following is known. In terms of the financial
futures market, the average market concentration value for S&P 500 was negative, indicating that the previous top four
short traders held more open interest positions than those of the previous top 4 long traders. On the other hand, the
average market concentration values for T-bonds and Eurodollars were both positive, indicating that the market
concentration for these two futures products are inclined towards long positions. In terms of the agricultural futures
market, only live cattle showed a negative average market concentration value. The average market concentrations for
the other three crops were positive, indicating that the long positions in the agricultural futures market were largely
concentrated in the hands of previous top four traders. In terms of the mineral futures market, the absolute values for
average market concentration were clearly greater than those of the other three markets and all had negative values
except for crude oil, indicating that there was an extremely high concentration of short positions. This may represent
the fact that during the research period, the short positions in the mineral futures market were held by an extremely
small number of traders, or that traders had large amounts of unfavorable information in regard to mineral futures,
leading to the outcome where a small number of traders held a large amount of short positions. In terms of the currency
futures market, only the Canadian dollar had a negative average market concentration value; the average values for the
other three currency futures products were around 0.1; and the positions held by the previous top four long traders were
slightly greater than those of the previous top four short traders.

120 Management Review, Jul 2023

Table 1
Descriptive Statistics and Analysis of Futures Risk Premium
Average Correlation Coefficients
Standard Financial Agricultural Mineral Currency Number of
Variable Average Deviation Skewness Unit Root Test Futures Futures Futures Futures Samples
Financial S&P500 0.0069 0.0345 –0.8392 0.0000*** –0.1132 0.1142 0.1773 0.1018 466
Futures T-bond 0.0018 0.0268 0.4044 0.0000*** –0.1676 –0.0859 –0.0124 –0.0072 466
Eurodollars –0.0002 0.0006 –1.2622 0.0006*** 0.1516 0.0926 0.0994 0.1509 466
Agricultural Wheat 0.0000 0.0801 0.5407 0.0000*** 0.0766 0.4163 0.0580 0.1348 466
Futures Corn –0.0001 0.0723 –0.1280 0.0000*** 0.0297 0.4479 0.1451 0.1361 466
Soybean –0.0008 0.0591 –0.4572 0.0000*** 0.0473 0.3673 0.1493 0.1381 466
Live Cattle 0.0030 0.0470 –0.6314 0.0000*** 0.0076 –0.0041 0.0270 –0.0449 466
Mineral Gold 0.0010 0.0426 –0.1488 0.0000*** 0.0649 0.0782 0.5271 0.2813 466
Futures Silver –0.0016 0.0796 –0.1213 0.0001*** 0.0937 0.1125 0.5730 0.2610 466
Platinum –0.0059 0.0557 –0.1992 0.0000*** 0.1264 0.1104 0.5385 0.2902 466
Crude Oil –0.0046 0.0903 –0.3932 0.0000*** 0.0673 0.0784 0.1572 0.2205 466
Currency Swiss Francs 0.0004 0.0304 –0.1578 0.0001*** 0.0800 0.0886 0.2777 0.3498 466
Futures British Pound –0.0021 0.0238 –0.7615 0.0000*** 0.0562 0.1417 0.1710 0.3071 466
Japanese Yen –0.0017 0.0254 –0.3049 0.0008*** 0.0181 –0.0013 0.2113 0.1605 466
Canadian Dollars –0.0022 0.0214 –0.0106 0.0000*** 0.1730 0.1350 0.3930 0.2624 466
Notes: 1. *Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level. 2. The risk premium is defined as R_(i,t)=ln(F_(i,t)⁄F_(i,t-4) ). 3. The augmented Dickey–
Fuller (ADF) statistics test the null hypothesis that an examined series has a unit root, and the corresponding lag for this test is 17. 4. The average correlation coefficient is the average of the
correlation coefficients of the risk premium of the commodity in the same type of futures market (e.g., financial, agricultural, mineral, or currency).

Market Concentration and Risk Premium 121

Table 2
Descriptive Statistics and Analysis of Hedging Pressure
Standard Average Correlation Number of
Variable Average Deviation Coefficients Samples
Financial S&P500 0.0002 0.1052 0.1334 470
Futures 470
T-bond 0.0107 0.0598 –0.0334
Eurodollars 0.0368 0.0705 0.0855 470
Agricultural Wheat 0.1059 0.1040 0.3198 470
Futures 470
Corn –0.0260 0.0884 0.2917
Soybean –0.0508 0.1201 0.1559 470
Live Cattle –0.1615 0.1024 –0.1407 470
Mineral Gold –0.2589 0.1194 0.3947 470
Futures 470
Silver –0.2592 0.1422 0.3612
Platinum –0.5377 0.1547 0.3015 470
Crude Oil –0.1460 0.0552 –0.3364 470
Currency Swiss Francs 0.2164 0.4701 0.2996 470
Futures 470
British Pound 0.1107 0.2314 0.1381
Japanese Yen 0.2258 0.3290 0.1025 470

Canadian Dollars –0.0215 0.3771 0.2732 470

(𝐿𝑜𝑛𝑔 𝑃𝑜𝑠𝑖𝑡𝑖𝑜𝑛𝑠𝑖,𝑡 −𝑆ℎ𝑜𝑟𝑡 𝑃𝑜𝑠𝑖𝑡𝑖𝑜𝑛𝑠𝑖,𝑡 )
Notes: 1. The hedging pressure is defined as 𝑞𝑖,𝑡 = . 2. The average correlation coefficient is
(𝐿𝑜𝑛𝑔 𝑃𝑜𝑠𝑖𝑡𝑖𝑜𝑛𝑠𝑖,𝑡 +𝑆ℎ𝑜𝑟𝑡 𝑃𝑜𝑠𝑖𝑡𝑖𝑜𝑛𝑠𝑖,𝑡 )
the average of the correlation coefficients of the hedging pressure of the commodity in the same type of futures market (e.g.,
financial, agricultural, mineral, or currency).

Table 3
Descriptive Statistics and Analysis of Market Concentration
Standard Average Correlation Number of
Variable Average Deviation Coefficients Samples
Financial S&P500 –0.0553 0.1668 –0.1273 470
Futures 470
T-bond 0.0580 0.1348 –0.1287
Eurodollars 0.2161 0.1946 –0.0951 470
Agricultural Wheat 0.1478 0.0986 0.1759 470
Futures 470
Corn 0.0825 0.1120 0.1821
Soybean 0.0108 0.1746 –0.0377 470

Live Cattle –0.0242 0.1650 0.1219 470

Mineral Gold –0.1902 0.1141 0.1950 470
Futures 470
Silver –0.3132 0.1383 0.0501
Platinum –0.2461 0.0883 0.1701 470

Crude Oil 0.0396 0.1459 –0.2668 470

Currency Swiss Francs 0.1172 0.2291 0.3760 470
Futures 470
British Pound 0.1377 0.1686 0.3706
Japanese Yen 0.1055 0.1658 0.1813 470

Canadian Dollars –0.0782 0.2109 0.3666 470

(𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛_𝑁𝑒𝑡 𝐿𝑜𝑛𝑔𝑖,𝑡 −𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛_𝑁𝑒𝑡 𝑆ℎ𝑜𝑟𝑡𝑖,𝑡 )
Notes: The market concentration is defined as 𝐶𝑜𝑛𝑖,𝑡 = . 2.The average
(𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛_𝑁𝑒𝑡 𝐿𝑜𝑛𝑔𝑖,𝑡 +𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛_𝑁𝑒𝑡 𝑆ℎ𝑜𝑟𝑡𝑖,𝑡 )
correlation coefficient is the average of the correlation coefficients of the market concentration of the commodity in the
same type of futures market (e.g., financial, agricultural, mineral, or currency).

122 Management Review, Jul 2023

4.2 Influence of Hedging Pressure on Futures Risk Premium
To validate the hedging pressure hypothesis, we first considered the long-term changes in the future return of
futures (including the change in futures return of up to four weeks, 𝑟𝑖,𝑡+4 ). The empirical results of the relationship
between Model 3’s cross-hedging pressure and futures risk premium are shown in Table 4. From Table 4, we can see
that in the financial futures market, hedging pressure held a considerable amount of explanatory power in regard to the
changes in futures risk premium in one’s own market, and could explain the changes in own-market futures risk premium
for all three variables of S&P 500, T-bonds, and Eurodollars. However, the same did not hold true for other markets, in
which only the hedging pressures of wheat in the agricultural market, gold and platinum in the mineral futures market,
and the British pound in the currency market could explain the changes in futures risk premium for one’s own market.
Only the Eurodollar in the financial futures market had a negative coefficient. For other futures products with significant
coefficients, hedging pressure was positively correlated with one’s own risk premium, indicating that when commercial
traders who are regarded as hedgers occupy short positions, futures risk premium will decrease as predicted by the
hedging pressure theory; additionally, when commercial traders occupy long positions, futures risk premium will
increase, which conforms to the argument forwarded by Stulz (1996) stating that commercial traders with advantages
regarding information in their industry can obtain positive risk premiums when they occupy long positions in the futures
market. In addition, in terms of cross-market hedging pressure, only the hedging pressures of the former variables for
T-bonds to S&P 500 and Euro-dollar; wheat to soybean, live cattle to corn; gold to crude oil; silver to gold; crude oil to
gold, silver, and platinum, Swiss francs to British pounds; and Japanese Yen to Canadian dollars could explain the
changes of the latter variables, indicating that cross-market hedging’s explanatory power for futures risk premium is
relatively weak.

Table 4
Influence of Hedging Pressure on Long-Term Futures Risk Premium
Panel A: Financial Futures Market (𝑗)= 1
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Systematic
Risk S&P500 T-bond Eurodollars
S&P500 –0.0900*** 0.0347** –0.0834*** –0.0356 0.039
(0.0473) (0.0155) (0.0268) (0.0233)
T-bond –0.0249 –0.0012 0.0478** –0.0140 0.014
(0.0375) (0.0123) (0.0213) (0.0184)
Eurodollars –0.0014* 0.0002 –0.0009* –0.0031*** 0.114
(0.0009) (0.0003) (0.0005) (0.0004)
Panel B: Agricultural Futures Market (𝑗)= 2
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Systematic
Risk Wheat Corn Soybean Live Cattle
Wheat –0.1880* 0.1139** 0.0630 0.0102 –0.0172 0.050
(0.1108) (0.0462) (0.0569) (0.0346) (0.0379)
Corn –0.1015 0.0606 –0.0079 0.0002 –0.0784** 0.019
(0.1016) (0.0424) (0.0522) (0.0318) (0.0347)
Soybean –0.0487 0.0789** –0.0035 –0.0063 –0.0151 0.018
(0.0830) (0.0346) (0.0427) (0.0260) (0.0284)
Live Cattle –0.0434 –0.0304 0.0486 –0.0022 0.0018 0.005
(0.0664) (0.0277) (0.0341) (0.0208) (0.0227)
Panel C: Mineral Futures Market (𝑗)= 3
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Systematic
Risk Gold Silver Platinum Crude Oil
Gold –0.0385 0.0771** –0.0455* –0.0109 0.0843** 0.024
(0.0596) (0.0327) (0.0249) (0.0179) (0.0390)
Silver 0.0374 0.0199 –0.0060 –0.0159 0.1715** 0.016
(0.1110) (0.0610) (0.0463) (0.0334) (0.0727)

Market Concentration and Risk Premium 123

 Platinum 0.0702 –0.0008 0.0025 0.0421* 0.1147** 0.017
(0.0782) (0.0430) (0.0326) (0.0235) (0.0512)
Crude Oil –0.0687 –0.2139*** 0.0136 0.0429 0.0498 0.053
(0.1245) (0.0684) (0.0519) (0.0374) (0.0815)
Panel D: Currency Futures Market (𝑗)= 4
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Systematic Canadian
Swiss Francs British Pound Japanese Yen
Risk Dollars
Swiss Francs 0.0917** –0.0015 0.0091 0.0005 –0.0038 0.017
(0.0424) (0.0035) (0.0069) (0.0046) (0.0042)
British Pound 0.0373 –0.0070** 0.0144*** 0.0071** 0.0000 0.029
(0.0329) (0.0027) (0.0054) (0.0036) (0.0033)
Japanese Yen –0.0092 0.0015 0.0083 –0.0021 –0.0056 0.013
(0.0358) (0.0029) (0.0059) (0.0039) (0.0035)
Canadian Dollars –0.0384 0.0007 0.0024 –0.0142*** 0.0021 0.053
(0.0295) (0.0024) (0.0048) (0.0032) (0.0029)
Notes: 1. *Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level. 2. This table reports
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗)
the regression results of the following model: 𝑟𝑖,𝑡+4 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + ∑𝑛𝑆=1 𝜃𝑠,𝑖 𝑞𝑠,𝑡 + 𝜀𝑖,𝑡+4 , where (𝑗)=1,2,3,4,
denoting financial, agricultural, mineral, or currency futures market, respectively. For example, when (𝑗)=1 refers to
financial futures market, s is S&P 500, T–bond, and Eurodollar. When (𝑗)=2 refers to agricultural futures market, s is wheat,
corn, soybeans, and live cattle. When (𝑗)=3 refers to mineral futures market, s is gold, silver, platinum, and crude oil. When
(𝑗)=4 refers to currency futures market, s is Swiss franc, British pound, Japanese yen, and Canadian dollar; 𝑟𝑖,𝑡+4 represents
𝑠&𝑝500
the futures contract i’s return at time t+4; 𝛼0 is the intercept term; 𝑟𝑚𝑡 represents the return of the S&P 500 index at
(𝑗) (𝑗)
time t; 𝑞𝑠,𝑡 is the hedging pressure of futures contract s at time t; 𝜀𝑖,𝑡+4 is the futures contract i’s error term. 3. The standard
errors are reported in parentheses; The number of observations is 462.

4.3 The Influence of Market Concentration on Futures Risk Premium

To validate the information asymmetry hypothesis, we considered the long-term changes in the future return of
futures (change in futures return after four weeks, 𝑟𝑖,𝑡+4 ) based on the relationship between Model 4’s market
concentration and futures risk premium, the empirical results, which are shown in Table 5. From Table 5, we see that
for T-bonds and Eurodollars in the financial futures market, wheat and corn in the agricultural futures market, and
platinum and crude oil in the mineral futures market, market concentration was able to explain the futures risk premium
for one’s own market. On the other hand, market concentration could not explain futures risk premium for any futures
products in the currency futures market. From the significant correlations between market concentration and futures risk
premium illustrated above, only the Eurodollar had a negative coefficient, while the coefficients for the other variables
were significant and positive, verifying that a higher market concentration brings about an increase in futures risk
premium. In addition, when open interest in the futures market is concentrated in a small number of traders, these traders
will indeed reap the benefits of such futures contracts; when concentrated in open positions, futures risk premium will
increase, and when concentrated in short positions, futures risk premium will decrease. This is consistent with this
study’s predicted result, that changes in trader concentration are a result of traders changing their futures holding
strategies due to benefiting from more information.

Table 5
Influence of Market Concentration on Long-Term Futures Risk Premium
Panel A: Financial Futures Market (𝑗)= 1
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk S&P500 T-bond Eurodollars
S&P500 –0.0928* 0.0076 –0.0002 –0.0067 0.012
(0.0475) (0.0098) (0.0121) (0.0084)
T-bond –0.0336 0.0089 0.0382*** 0.0228*** 0.056
(0.0363) (0.0075) (0.0093) (0.0064)

124 Management Review, Jul 2023

 Eurodollars –0.0015* 0.0001 0.0012*** –0.0005*** 0.101
(0.0009) (0.0002) (0.0002) (0.0002)
Panel B: Agricultural Futures Market (𝑗)= 2
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk Wheat Corn Soybean Live Cattle
Wheat –0.1318 0.0922** 0.0797** 0.0083 –0.0049 0.039
(0.1123) (0.0395) (0.0366) (0.0213) (0.0236)
Corn –0.0097 –0.0096 0.1202*** 0.0069 –0.0227 0.031
(0.1017) (0.0357) (0.0332) (0.0193) (0.0214)
Soybean 0.0036 0.0341 0.0524* 0.0060 0.0170 0.023
(0.0835) (0.0293) (0.0272) (0.0159) (0.0175)
Live Cattle –0.0050 –0.0482** 0.0731*** –0.0020 0.0033 0.031
(0.0660) (0.0232) (0.0215) (0.0125) (0.0139)
Panel C: Mineral Futures Market (𝑗)= 3
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk Gold Silver Platinum Crude Oil
Gold –0.0261 0.0232 –0.0111 0.0256 0.0457*** 0.010
(0.0591) (0.0224) (0.0180) (0.0247) (0.0154)
Silver 0.0607 0.0132 –0.0209 0.0732 0.0809*** 0.032
(0.1095) (0.0415) (0.0334) (0.0458) (0.0285)
Platinum 0.0779 –0.0166 0.0182 0.0905*** 0.0540*** 0.036
(0.0770) (0.0292) (0.0235) (0.0322) (0.0201)
Crude Oil –0.0577 –0.0075 –0.0061 –0.1972*** 0.0624* 0.053
(0.1238) (0.0469) (0.0378) (0.0518) (0.0323)
Panel D: Currency Futures Market (𝑗)=4
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk Swiss Francs British Pound Japanese Yen Canadian Dollars
Swiss Francs 0.0822* –0.0072 0.0271*** –0.0056 –0.0187** 0.034
(0.0421) (0.0072) (0.0100) (0.0087) (0.0082)
British Pound 0.0509 –0.0088 0.0012 0.0049 0.0047 0.011
(0.0333) (0.0057) (0.0079) (0.0069) (0.0065)
Japanese Yen –0.0146 0.0031 0.0035 –0.0049 –0.0128* 0.010
(0.0359) (0.0061) (0.0085) (0.0074) (0.0070)
Canadian Dollars –0.0420 –0.0001 0.0076 –0.0248*** –0.0007 0.043
(0.0297) (0.0051) (0.0070) (0.0061) (0.0058)
Notes: 1. *Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level. 2. This table reports the
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗)
regression results of the following model: 𝑟𝑖,𝑡+4 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + ∑𝑛𝑆=1 𝛾𝑠,𝑖 𝐶𝑜𝑛𝑠,𝑡 + 𝜀𝑖,𝑡+4 , where (𝑗) =1,2,3,4,
denoting financial, agricultural, mineral, or currency futures market, respectively. For example, when (𝑗)=1 refers to financial
futures market, s is S&P 500, T–bond, and Eurodollar. When (𝑗)=2 refers to agricultural futures market, s is wheat, corn,
soybeans, and live cattle. When (𝑗)=3 refers to mineral futures market, s is gold, silver, platinum, and crude oil. When (𝑗)=4
refers to currency futures market, s is Swiss franc, British pound, Japanese yen, and Canadian dollar; 𝑟𝑖,𝑡+4 represents the
𝑠&𝑝500
futures contract i’s return at time t+4; 𝛼0 is the intercept term; 𝑟𝑚𝑡 represents the return of the S&P 500 index at time t;
(𝑗) (𝑗)
𝐶𝑜𝑛𝑠,𝑡 is the market concentration of futures contract s at time t; 𝜀𝑖,𝑡+4 is the futures contract i’s error term. 3. The standard
errors are reported in parentheses; The number of observations is 462.

In addition, in terms of cross-market concentration, only the former variables of the following could explain the
changes in futures risk of the latter variables: this applies to T-bonds to Eurodollars; Eurodollars to T-bonds; wheat to
live cattle; corn to wheat, soybean, and live-cattle; platinum to crude oil; crude oil to gold, silver, and platinum; British

Market Concentration and Risk Premium 125

pounds to Swiss francs; Japanese yen to Canadian dollars; and Canadian dollars to Swiss francs and Japanese yen.
Among them, the market concentrations of corn and crude oil had the greatest explanatory powers and could explain
the risk premiums of other futures products in the same market.
4.4 Influence of Trader Concentration and Hedging Pressure on Futures Risk Premium
We considered the long-term changes in the future return of futures (including the change in futures return of up
to four weeks, 𝑟𝑖,𝑡+4 ) based on the relationship between the combination of market concentration and hedging pressure
and futures risk premium, the empirical results of which are shown in Table 6. Table 6 shows that in the financial futures
market, the futures risk premiums regarding T-bonds and Eurodollars were simultaneously affected by their own market
concentration and hedging pressure, while the futures risk premium of S&P 500 was only impacted by its own hedging
pressure. In the agricultural futures market, the futures risk premiums of wheat and live cattle were simultaneously
influenced by their own market concentration and hedging pressure, but the futures risk premium of corn was only
influenced by its own market concentration. In the mineral futures market, the futures risk premium of platinum was
simultaneously influenced by its own market concentration and hedging pressure, but the futures risk premium of gold
was only influenced by its own hedging pressure. In terms of the currency futures market, the Swiss franc was influenced
by its own market concentration, while the British pound was influenced by its own hedging pressure. In summary,
some futures products will be simultaneously influenced by market concentration and hedging pressure, while others
will only be influenced by a single indicator. Only a small number of futures products were not influenced by either
indicator, demonstrating that in many cases, market concentration and hedging pressure are mutually reinforcing when
explaining futures risk premium and that they can also coexist. The new indicator proposed by this study of market
concentration serves to provide new reference values for past research, while the role played by information asymmetry
in the futures market is likewise crucial, providing varying degrees of explanatory power for different futures products.
The Eurodollar is the most actively traded futures product on the CME. Its short-term interest rate not only affects
T-bond rates, but also exerts a huge impact on other financial markets, making its result one that is especially significant.
Furthermore, due to data limitations, the market concentration indicator is unable to be differentiated based on
transaction type (the top traders may fall under commercial or non-commercial classifications); therefore, hedging
pressure and market concentration may have a certain degree of correlation that caused a portion of the results of Table
6 to differ from Table 5.

Table 6
Influence of Market Concentration and Hedging Pressure on Long-term Futures Risk
Premium
Panel A: Financial Futures Market (𝑗)= 1
Dependent
Variable Independent Variable Adj. R2
Hedging Pressure
Systematic Market Eurodollar
S&P500 T-bond
Risk Concentration s
S&P500 –0.0871* –0.0126 0.0488** –0.0858*** –0.0392* 0.042
(0.0474) (0.0130) (0.0213) (0.0269) (0.0236)
T-bond –0.0272 0.0307*** 0.0069 0.0355* –0.0059 0.035
(0.0371) (0.0097) (0.0124) (0.0214) (0.0184)
Eurodollars – 0.122
–0.0014 0.0004** 0.0003 –0.0009*
0.0040***
(0.0009) (0.0002) (0.0003) (0.0005)
(0.0006)
Panel B: Agricultural Futures Market (𝑗)= 2
Dependent
Variable Independent Variable Adj. R2
Hedging Pressure
Systematic Market
Risk Concentration Wheat Corn Soybean Live Cattle
Wheat –0.1877* 0.0916** 0.0999** 0.0647 –0.0041 –0.0285 0.062
(0.1102) (0.0383) (0.0464) (0.0566) (0.0350) (0.0380)
Corn – 0.058
–0.0358 0.1330*** 0.0966** –0.0214 –0.0186
0.1041***
(0.1009) (0.0311) (0.0424) (0.0513) (0.0315)
(0.0346)

126 Management Review, Jul 2023

 Soybean –0.0485 –0.0289 0.0876** –0.0183 0.0306 –0.0127 0.021
(0.0830) (0.0285) (0.0357) (0.0451) (0.0447) (0.0285)
Live Cattle –0.0383 0.0537** –0.0372 0.0657* 0.0044 –0.0621* 0.018
(0.0661) (0.0227) (0.0277) (0.0347) (0.0209) (0.0351)
Panel C: Mineral Futures Market (𝑗)= 3
Dependent
Variable Independent Variable Adj. R2
Hedging Pressure
Systematic Market
Risk Concentration Gold Silver Platinum Crude Oil
Gold –0.0384 –0.0012 0.0780** –0.0455* –0.0111 0.0840** 0.025
(0.0597) (0.0226) (0.0364) (0.0249) (0.0184) (0.0394)
Silver 0.0379 0.0108 0.0164 –0.0060 –0.0183 0.1844** 0.016
(0.1111) (0.0469) (0.0629) (0.0463) (0.0349) (0.0920)
Platinum 0.0656 0.0751** –0.0093 –0.0193 0.0415* 0.0592 0.026
(0.0780) (0.0376) (0.0431) (0.0343) (0.0234) (0.0581)
Crude Oil –0.0571 0.0289 –0.2229*** 0.0298 0.0461 0.0134 0.054
(0.1257) (0.0423) (0.0697) (0.0571) (0.0377) (0.0974)
Panel D: Currency Futures Market (𝑗)= 4
Dependent
Variable Independent Variable Adj. R2
Hedging Pressure
Systematic Market British Japanese Canadian
Swiss Francs
Risk Concentration Pound Yen Dollars
Swiss Francs 0.1091** –0.0284** 0.0099 0.0105 0.0013 –0.0024 0.027
(0.0431) (0.0134) (0.0064) (0.0070) (0.0046) (0.0042)
British 0.0414 –0.0143 –0.0067** 0.0213*** 0.0078** 0.0009 0.034
Pound (0.0330) (0.0094) (0.0027) (0.0070) (0.0036) (0.0033)
Japanese –0.0087 –0.0017 0.0015 0.0084 –0.0016 –0.0056 0.013
Yen (0.0359) (0.0090) (0.0030) (0.0059) (0.0047) (0.0035)
Canadian –0.0383 0.0006 0.0006 0.0023 –0.0142*** 0.0018 0.053
Dollars (0.0296) (0.0079) (0.0024) (0.0051) (0.0032) (0.0042)
Notes: 1. *Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level. 2. This table reports the
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗) (𝑗)
regression results of the following model: 𝑟𝑖,𝑡+4 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + 𝛾𝑖 𝐶𝑜𝑛𝑖,𝑡 + ∑𝑛𝑆=1 𝜃𝑠,𝑖 𝑞𝑠,𝑡 + 𝜀𝑖,𝑡+4 , where
(𝑗)=1,2,3,4, denoting financial, agricultural, mineral, or currency futures market, respectively. For example, when (𝑗)=1 refers
to financial futures market, s is S&P 500, T–bond, and Eurodollar. When (𝑗)=2 refers to agricultural futures market, s is wheat,
corn, soybeans, and live cattle. When (𝑗)=3 refers to mineral futures market, s is gold, silver, platinum, and crude oil. When
(𝑗)=4 refers to currency futures market, s is Swiss franc, British pound, Japanese yen, and Canadian dollar; 𝑟𝑖,𝑡+4 represents
𝑠&𝑝500
the futures contract i’s return at time t+4; 𝛼0 is the intercept term; 𝑟𝑚𝑡 represents the return of the S&P 500 index at time
(𝑗) (𝑗)
t; 𝐶𝑜𝑛𝑠,𝑡 is the market concentration of futures contract s at time t; 𝑞𝑠,𝑡 is the hedging pressure of futures contract s at time t;
(𝑗)
𝜀𝑖,𝑡+4 is the futures contract i’s error term. 3. The standard errors are reported in parentheses; The number of observations is
462.

4.5 Robustness Check

Table 4 and Table 6 of this paper were constructed after conducting analysis using futures risk premium after four
weeks as the dependent variable. The reason for adopting this methodology was that changes in trader position were not
expected to be the result of closing positions in a short period of time; in addition, hedgers invest in the futures market
with the aim of eliminating future spot pressure. It was therefore expected that traders would hold futures products for
a relatively long period of time, or even until maturity. Consequently, this chapter conducted robustness checks by using
different calculation methods for futures risk premium. Appendix Table A1 to A3 used weekly premium (𝑟𝑖,𝑡+1 ), and
Appendix Table A4 to A6 used bi-weekly risk premium (𝑟𝑖,𝑡+2 ) and replicated the methods of Table 4 to Table 6 to verify
whether the indicators of hedging pressure and market concentration were also valid with respect to short-term futures
price changes.
From Appendix Table A1, Relationship between Cross-Hedging Pressure and Futures Risk Premium (after One

Market Concentration and Risk Premium 127

Week), it can be seen that only the hedging pressures of S&P 500, Eurodollars, and wheat had a significant relationship
with futures risk premium after one week, while such significance was also not strong, indicating that the hedging
pressure of most futures products will not influence futures risk premium after one week. As expected, short-term
changes in futures prices do not have a significant relationship with hedging pressure. This was further verified by cross-
hedging pressure, where it was observed that only the hedging pressure of gold had a significant relationship with the
futures risk premium of crude oil, while no significance was observed for the cross-hedging pressures of all other
variables.
From Appendix Table A2 Relationship between Cross-Market Concentration and Futures Risk Premium (after One
Week), it can be observed that the market concentrations of Eurodollars, corn, and platinum had a significant relationship
with their own futures risk premium after one week. Compared with Table 5, the quantity of significant results decreased
by half, indicating that changes in market concentration have relatively weak explanatory power for short-term futures
risk premium. In terms of cross-market concentration, only the market concentrations of T-bonds to Eurodollars and
Japanese yen to Canadian dollars showed significant relationships. The market concentrations for the remaining
variables did not exhibit any significant results, indicating that besides the market concentration of one’s own market
being unable to adequately explain futures risk premium after one week, the market concentration of related futures
markets also lacks such explanatory power.
From Appendix Table A3, Relationship between Cross-Market Concentration and Cross-Hedging Pressure and
Futures Risk Premium, it can be seen that, in line with the results of Appendix Table A1 and Table A2, regarding market
concentration, only the market concentrations of corn and platinum had a significant relationship with own-market
futures risk premium after one week, while for hedging pressure, only the hedging pressures regarding Eurodollars and
wheat could explain own-market futures risk premium after one week. The results show that even when market
concentration and hedging pressure were included in the same regression model, the results were not affected. In
summary, the results of Appendix Table A1 to A3 show that hedging pressure and market concentration are not a good
predictor of futures risk premium after one week, demonstrating that these two indicators are not suited to explaining
the changes in short-term futures risk premium.
From Appendix Table A4, Relationship between Cross-Hedging Pressure and Futures Risk Premium, it can be
observed that only the hedging pressures of S&P 500, Eurodollars, wheat, gold, and platinum had a significant
relationship with same-market futures risk premium after two weeks. In terms of cross-hedging pressure, only for live
cattle to corn, gold to crude oil, silver to gold, and Japanese yen to Canadian dollars did the former variables have
explanatory power regarding the futures risk premiums after two weeks of the latter variables.
Compared with Appendix Table A1’s futures risk premium after one week, the futures products for which hedging
pressure had a significant effect increased to include gold and platinum but compared to Table 4’s futures risk premium
after four weeks, the quantity of products was still less and exhibited lower significance. This indicates that the influence
of hedging pressure on futures risk premium will indeed increase as more time elapses. It can also be seen that the effect
on futures risk premium after two weeks is between that for futures risk premium after one week and four weeks,
respectively. Compared with Appendix Table A1, futures products influenced by cross-hedging pressure also increased
to include live cattle to corn, silver to gold, and Japanese yen to Canadian dollars. Such results verify that the hedging
pressures of some futures products can provide explanatory power to other futures products, and that the relationship
between hedging pressure and futures risk premium increases the more that time elapses.
From Appendix Table A5, Relationship between Cross-Market Concentration and Futures Risk Premium (after
Two Weeks), the market concentrations of T-bonds, Eurodollars, wheat, corn, and platinum had significant relationships
with own-market futures risk premium after two weeks. In terms of cross-market concentration, the former variables
for T-bonds to Eurodollars, Eurodollars to T-bonds, corn to live cattle, platinum to crude oil, crude oil to gold and silver,
British pounds to Swiss francs, Japanese yen to Canadian dollars, and Canadian dollars to Swiss francs provided
significant explanatory power for the futures risk premiums of the latter variables.
In line with the results of Appendix Table A4, the influence of market concentration on futures risk premium after
two weeks was greater than that for futures risk premium after one week but less than that for futures risk premium after
four weeks. Compared with Appendix Table A2, the variables whose market concentration could explain futures risk
premium increased to include T-bonds and wheat. In terms of cross-market concentration, the variables also included
Eurodollars to T-bonds, corn to live cattle, platinum to crude oil, crude oil to gold and silver, British pounds to Swiss
francs, and Canadian dollars to Swiss francs, verifying that the market concentrations of some futures products can
explain the futures risk premiums after two weeks of other futures products.
Appendix Table A6, Relationship between Market Concentration and Cross-Hedging Pressure on Futures Risk
Premium, continued the pattern of the results of Appendix Table A4 and Table A5, showing that, in terms of market
concentration, T-bonds, corn, live cattle, and platinum could explain the changes in own-market futures risk premium
after one week, and in terms of hedging pressure, those of S&P 500, Eurodollars, wheat, platinum, and British pounds
could explain the changes in own-market futures risk premium after two weeks. The results for both market
concentration and hedging pressure were like Appendix Table A4 and Table A5. Although it is worthy of note that in

128 Management Review, Jul 2023

Appendix Table A4 and Table A5, both own-market hedging pressure and cross-market concentration had a relatively
significant effect on the Eurodollar, when both variables were used in the same regression model, the effect of market
concentration became weaker. Such a result was also seen in Table 6, Relationship between Cross-Market Concentration
and Cross-Hedging Pressure and Futures Risk Premium (after Four weeks), indicating that for the Eurodollar, hedging
pressure is a better explanation of risk premium than market concentration. In summary of Appendixes Table A1 to
Table A6, hedging pressure and market concentration both displayed better effectiveness in explaining changes in long-
term futures risk premium than short-term, in that sense aligning with this paper’s expectations.

5. Conclusion
This paper discusses the relationship between risk premium and market concentration in the United States’ futures
market. Past research primarily used the hedging behaviors of hedgers as the measurement for futures risk premium,
while this study confirmed that market concentration influences the prices of futures because traders who hold more
futures contract positions usually have more market information and thus have a better ability to predict price
fluctuations. When the market concentration for the previous top four traders increase, a rise in net open positions has
a significant positive correlation with recent month futures risk premium, whereas an increase in net short positions has
a significant negative correlation with recent month futures risk premium.
Market concentration and hedging pressure measure different aspects. Market concentration measures the degree
of information asymmetry in the futures market and hedging pressure measures the hedging behaviors engaged in by
investors who hold or will hold spot stocks. Under the condition that market concentration and hedging pressure are
simultaneously considered, market concentration retains its significance in the prediction of futures prices. As a result,
market concentration provides another indicator for past literature to use in forming explanations regarding futures risk
premium, whereby the changes in futures risk premium for some futures products that were unable to be explained using
hedging pressure can be predicted by means of market concentration.
In summary, this study discovered that, besides the currency futures market where it had relatively weaker
explanatory power, market concentration had significant explanatory power in every market, verifying that market
concentration can serve as an important indicator for predicting futures prices. It also does not conflict with the hedging
pressure indicator, with the two providing different dimensions of explanation for futures risk premium. This finding
thus marks the main contribution of this study. The above results enable the prediction that they will provide
considerable practical assistance to the devising of futures investment strategies by market investors, the formulation of
trade regulations by policymakers, and the prevention of market manipulation, and can also help to establish a more
efficient and fair-trading environment for the futures market.

Market Concentration and Risk Premium 129

 References
Bessembinder, H. (1992). Systematic risk, hedging pressure, and risk premiums in futures markets. Review of Financial
Studies, 5(4), 637–667. https://doi.org/10.1093/rfs/5.4.637
Black, F. (1976). The pricing of commodity contracts. Journal of Financial Economics, 3(1–2), 167–179.
https://doi.org/10.1016/0304-405x(76)90024-6
Carter, C. A., Rausser, G. C., & Schmitz, A. (1983). Efficient asset portfolios and the theory of normal
backwardation. Journal of Political Economy, 91(2), 319–331. http://dx.doi.org/10.1086/261148
Chang, Y. K., Chen, Y. L., Chou, R. K., & Gau, Y. F. (2013). The effectiveness of position limits: Evidence from the
foreign exchange futures markets. Journal of Banking and Finance, 37(11), 4501–4509.
https://doi.org/10.1016/j.jbankfin.2013.02.033
Chen, Y. L., & Yang, J. J. (2021). Trader positions in VIX futures. Journal of Empirical Finance, 61, 1–17.
https://doi.org/10.1016/j.jempfin.2020.12.003
Cootner, P. H. (1960). Returns to speculators: Telser versus Keynes. Journal of Political Economy, 68(4), 396–404.
http://dx.doi.org/10.1086/258347
Cornell, B., & French, K. R. (1983a). The pricing of stock index futures. Journal of Futures Markets, 3(1), 1–14.
https://doi.org/10.1002/fut.3990030102
Cornell, B., & French, K. R. (1983b). Taxes and the pricing of stock index futures. Journal of Finance, 38(3), 675–694.
De Roon, F. A., Nijman, T. E., & Veld, C. (2000). Hedging pressure effects in futures markets. Journal of Finance, 55(3),
1437–1456. https://doi.org/10.1111/0022-1082.00253
Dusak, K. (1973). Futures trading and investor returns: An investigation of commodity market risk premiums. Journal
of Political Economy, 81(6), 1387–1406. https://doi.org/10.1086/260133
Gujarati, D. (2003). Multicollinearity: What happens if the regressors are correlated? In Basic econometrics (4th ed.,
pp. 341−386). McGraw-Hill.
Hicks, J. R. (1939). Value and capital. Oxford University Press.
Hirshleifer, D. (1988). Residual risk, trading costs, and commodity futures risk premia. Review of Financial
Studies, 1(2), 173–193. https://doi.org/10.1093/rfs/1.2.173
Houthakker, H. S. (1957). Can speculators forecast prices? Review of Economics and Statistics, 39(2), 143–151.
https://doi.org/10.2307/1928531
Jagannathan, R. (1985). An investigation of commodity futures prices using the consumption‐based intertemporal
capital asset pricing model. Journal of Finance, 40(1), 175–191. https://doi.org/10.1111/j.1540-
6261.1985.tb04943.x
Keynes, J. M. (1930). A treatise on money (Vol. 2). London: Macmillan.
Knill, A., Minnick, K., & Nejadmalayeri, A. (2006). Selective hedging, information asymmetry, and futures
prices. Journal of Business, 79(3), 1475–1501. https://doi.org/10.1086/500682
Smales, L. (2016). Trading behavior in S&P 500 index futures. Review of Financial Economics, 28, 46–55.
https://doi.org/10.1016/j.rfe.2015.11.001
Stoll, H. R. (1979). Commodity futures and spot price determination and hedging in capital market equilibrium. Journal
of Financial and Quantitative Analysis, 14(4), 873–894. https://doi.org/10.2307/2330460
Stulz, R. M. (1996). Rethinking risk management. Journal of Applied Corporate Finance, 9(3), 8–25.
https://doi.org/10.1111/j.1745-6622.1996.tb00295.x
Wang, C. (2003). The behavior and performance of major types of futures traders. Journal of Futures Markets, 23(1),
1–31. https://doi.org/10.1002/fut.10056

130 Management Review, Jul 2023

 Appendix

Table A1
Influence of Hedging Pressure on Futures Risk Premium After One Week
Panel A: Financial Futures Market (𝑗)= 1
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Systematic
Risk S&P500 T-bond Eurodollars
S&P500 -0.0620 0.0165* -0.0122 -0.0106 0.012
(0.0469) (0.0090) (0.0154) (0.0134)
T-bond 0.0075 -0.0029 0.0045 -0.0044 0.002
(0.0328) (0.0063) (0.0108) (0.0094)
Eurodollars 0.0008 0.0001 -0.0002 -0.0008*** 0.036
(0.0007) (0.0001) (0.0002) (0.0002)
Panel B: Agricultural Futures Market (𝑗)= 2
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Systematic
Risk Wheat Corn Soybean Live Cattle
Wheat -0.0920 0.0485* 0.0100 -0.0028 -0.0083 0.018
(0.1046) (0.0258) (0.0317) (0.0191) (0.0209)
Corn -0.0655 0.0136 0.0079 -0.0065 -0.0206 0.006
(0.0891) (0.0220) (0.0270) (0.0163) (0.0178)
Soybean -0.0024 0.0093 0.0005 -0.0001 -0.0026 0.001
(0.0732) (0.0181) (0.0222) (0.0134) (0.0146)
Live Cattle 0.0589 -0.0131 0.0181 0.0034 0.0046 0.005
(0.0602) (0.0149) (0.0183) (0.0110) (0.0120)
Panel C: Mineral Futures Market (𝑗)= 3
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Systematic
Risk Gold Silver Platinum Crude Oil
Gold -0.1017** 0.0203 -0.0134 -0.0014 0.0200 0.015
(0.0505) (0.0163) (0.0124) (0.0089) (0.0193)
Silver -0.2277** -0.0012 0.0008 0.0016 0.0322 0.015
(0.0942) (0.0304) (0.0231) (0.0165) (0.0360)
Platinum -0.0425 -0.0055 -0.0029 0.0154 0.0196 0.006
(0.0670) (0.0216) (0.0164) (0.0118) (0.0256)
Crude Oil -0.1048 -0.0610* 0.0122 0.0169 0.0153 0.013
(0.1082) (0.0349) (0.0266) (0.0190) (0.0413)
Panel D: Currency Futures Market (𝑗)= 4
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Systematic
Risk Swiss Francs British Pound Japanese Yen Canadian Dollars
Swiss Francs 0.0562 0.0009 0.0001 0.0010 -0.0017 0.006
(0.0387) (0.0019) (0.0036) (0.0025) (0.0022)
British Pound 0.0152 -0.0018 0.0030 0.0026 -0.0002 0.008
(0.0299) (0.0014) (0.0028) (0.0019) (0.0017)
Japanese Yen -0.0149 0.0010 0.0012 0.0003 -0.0021 0.005
(0.0304) (0.0015) (0.0029) (0.0019) (0.0018)
Canadian Dollars -0.0388 0.0002 0.0015 -0.0027 0.0004 0.011
(0.0274) (0.0013) (0.0026) (0.0017) (0.0016)

Market Concentration and Risk Premium 131

Notes: 1. *Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level. 2. This table reports
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗)
the regression results of the following model: 𝑟𝑖,𝑡+1 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + ∑𝑛𝑆=1 𝜃𝑠,𝑖 𝑞𝑠,𝑡 + 𝜀𝑖,𝑡+1 , where (𝑗)=1,2,3,4,
denoting financial, agricultural, mineral, or currency futures market, respectively. For example, when (𝑗)=1 refers to
financial futures market, s is S&P 500, T-bond, and Eurodollar. When (𝑗)=2 refers to agricultural futures market, s is wheat,
corn, soybeans, and live cattle. When (𝑗)=3 refers to mineral futures market, s is gold, silver, platinum, and crude oil. When
(𝑗)=4 refers to currency futures market, s is Swiss franc, British pound, Japanese yen, and Canadian dollar; 𝑟𝑖,𝑡+1 represents
𝑠&𝑝500
the futures contract i’s return at time t+1; 𝛼0 is the intercept term; 𝑟𝑚𝑡 represents the return of the S&P 500 index at
(𝑗) (𝑗)
time t; 𝑞𝑠,𝑡 is the hedging pressure of futures contract s at time t; 𝜀𝑖,𝑡+1 is the futures contract i’s error term. 3. The standard
errors are reported in parentheses; The number of observations is 468.

Table A2
Influence of Market Concentration on Futures Risk Premium After One Week
Panel A: Financial Futures Market (𝑗)=1
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk S&P500 T-bond Eurodollars
S&P500 -0.0577 0.0055 0.0012 -0.0008 0.005
(0.0469) (0.0056) (0.0069) (0.0048)
T-bond 0.0071 -0.0005 0.0069 0.0049 0.009
(0.0326) (0.0039) (0.0048) (0.0033)
Eurodollars 0.0009 0.0000 0.0002** -0.0001** 0.028
(0.0007) (0.0001) (0.0001) (0.0001)
Panel B: Agricultural Futures Market (𝑗)=2
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk Wheat Corn Soybean Live Cattle
Wheat -0.0753 0.0264 0.0202 -0.0016 -0.0037 0.009
(0.1055) (0.0218) (0.0199) (0.0118) (0.0131)
Corn -0.0395 -0.0177 0.0417** -0.0004 -0.0073 0.014
(0.0891) (0.0185) (0.0168) (0.0100) (0.0111)
Soybean 0.0087 0.0039 0.0141 0.0011 0.0039 0.004
(0.0734) (0.0152) (0.0139) (0.0082) (0.0091)
Live Cattle 0.0643 -0.0118 0.0148 0.0012 0.0021 0.007
(0.0604) (0.0125) (0.0114) (0.0068) (0.0075)
Panel C: Mineral Futures Market (𝑗)=3
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk Gold Silver Platinum Crude Oil
Gold -0.1031** 0.0015 -0.0015 0.0132 0.0102 0.017
(0.0503) (0.0110) (0.0088) (0.0123) (0.0077)
Silver -0.2309** -0.0069 0.0000 0.0349 0.0148 0.021
(0.0937) (0.0204) (0.0164) (0.0229) (0.0142)
Platinum -0.0480 -0.0070 0.0001 0.0333** 0.0075 0.012
(0.0666) (0.0145) (0.0117) (0.0163) (0.0101)
Crude Oil -0.0987 -0.0097 0.0056 -0.0276 0.0172 0.009
(0.1082) (0.0236) (0.0189) (0.0264) (0.0164)
Panel D: Currency Futures Market (𝑗)=4
Dependent Variable Independent Variable Adj. R2
Systematic Risk Market Concentration

132 Management Review, Jul 2023

 Canadian
Swiss Francs British Pound Japanese Yen
Dollars
Swiss Francs 0.0499 0.0000 0.0056 0.0005 -0.0049 0.008
(0.0388) (0.0039) (0.0054) (0.0047) (0.0044)
British Pound 0.0177 -0.0017 -0.0004 0.0018 0.0001 0.002
(0.0301) (0.0030) (0.0042) (0.0036) (0.0034)
Japanese Yen -0.0171 0.0002 -0.0003 0.0003 -0.0020 0.002
(0.0306) (0.0031) (0.0043) (0.0037) (0.0035)
Canadian Dollars -0.0407 -0.0003 0.0032 -0.0068** -0.0013 0.015
(0.0275) (0.0027) (0.0038) (0.0033) (0.0031)
Notes: 1. *Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level. 2. This table reports the
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗)
regression results of the following model: 𝑟𝑖,𝑡+1 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + ∑𝑛𝑆=1 𝛾𝑠,𝑖 𝐶𝑜𝑛𝑠,𝑡 + 𝜀𝑖,𝑡+1 , where (𝑗) =1,2,3,4,
denoting financial, agricultural, mineral, or currency futures market, respectively. For example, when (𝑗)=1 refers to financial
futures market, s is S&P 500, T-bond, and Eurodollar. When (𝑗)=2 refers to agricultural futures market, s is wheat, corn,
soybeans, and live cattle. When (𝑗)=3 refers to mineral futures market, s is gold, silver, platinum, and crude oil. When (𝑗)=4
refers to currency futures market, s is Swiss franc, British pound, Japanese yen, and Canadian dollar; 𝑟𝑖,𝑡+1 represents the
𝑠&𝑝500
futures contract i’s return at time t+1; 𝛼0 is the intercept term; 𝑟𝑚𝑡 represents the return of the S&P 500 index at time t;
(𝑗) (𝑗)
𝐶𝑜𝑛𝑠,𝑡 is the market concentration of futures contract s at time t; 𝜀𝑖,𝑡+1 is the futures contract i’s error term. 3. The standard
errors are reported in parentheses; The number of observations is 468.

Table A3
Influence of Market Concentration and Hedging Pressure on Futures Risk Premium
After One Week
Panel A: Financial Futures Market (𝑗)=1
Dependent Independent Variable Adj. R2
Variable
Hedging Pressure
Systematic Market
Risk Concentration S&P500 T-bond Eurodollars
S&P500 -0.0619 -0.0029 0.0197 -0.0129 -0.0113 0.012
(0.04370) (0.0075) (0.0123) (0.0155) (0.0136)
T-bond 0.0066 0.0056 -0.0014 0.0026 -0.0031 0.005
(0.0328) (0.0050) (0.0064) (0.0109) (0.0095)
Eurodollars 0.0008 0.0001 0.0001 -0.0002 -0.0010*** 0.038
(0.0007) (0.0001) (0.0001) (0.0002) (0.0003)
Panel B: Agricultural Futures Market (𝑗)=2
Dependent Independent Variable Adj. R2
Variable
Hedging Pressure
Systematic Market
Risk Concentration Wheat Corn Soybean Live Cattle
Wheat -0.0906 0.0229 0.0450* 0.0102 -0.0062 -0.0112 0.020
(0.1046) (0.0215) (0.0260) (0.0317) (0.0194) (0.0211)
Corn -0.0450 0.0423*** 0.0254 0.0038 -0.0136 -0.0294 0.021
(0.0889) (0.0161) (0.0223) (0.0269) (0.0164) (0.0180)
Soybean -0.0027 -0.0050 0.0108 -0.0021 0.0062 -0.0023 0.001
(0.0733) (0.0149) (0.0186) (0.0235) (0.0232) (0.0147)
Live Cattle 0.0607 0.0136 -0.0147 0.0224 0.0049 -0.0116 0.007
(0.0602) (0.0122) (0.0149) (0.0187) (0.0111) (0.0189)
Panel C: Mineral Futures Market (𝑗)=3
Dependent Independent Variable Adj. R2
Variable
Hedging Pressure

Market Concentration and Risk Premium 133

 Systematic Market Gold Silver Platinum Crude Oil
Risk Concentration
Gold -0.1015** -0.0014 0.0213 -0.0134 -0.0016 0.0197 0.015
(0.0506) (0.0111) (0.0181) (0.0124) (0.0091) (0.0195)
Silver -0.2276** 0.0019 -0.0018 0.0008 0.0012 0.0344 0.015
(0.0943) (0.0232) (0.0314) (0.0232) (0.0173) (0.0456)
Platinum -0.0509 0.0411** -0.0103 -0.0146 0.0155 -0.0106 0.016
(0.0668) (0.0187) (0.0216) (0.0172) (0.0117) (0.0289)
Crude Oil -0.1031 0.0155 -0.0659* 0.0208 0.0187 -0.0046 0.014
(0.1083) (0.0214) (0.0356) (0.0291) (0.0192) (0.0496)
Panel D: Currency Futures Market (𝑗)=4
Dependent Independent Variable Adj. R2
Variable
Hedging Pressure
Systematic Market Swiss British Japanese Canadian
Risk Concentration Francs Pound Yen Dollars
Swiss Francs 0.0585 -0.0045 0.0027 0.0003 0.0011 -0.0015 0.007
(0.0389) (0.0071) (0.0034) (0.0037) (0.0025) (0.0023)
British 0.0163 -0.0042 -0.0017 0.0050 0.0028 0.0001 0.009
Pound (0.0299) (0.0049) (0.0014) (0.0036) (0.0019) (0.0018)
Japanese -0.0149 0.0002 0.0010 0.0012 0.0002 -0.0021 0.005
Yen (0.0305) (0.0045) (0.0015) (0.0029) (0.0023) (0.0018)
Canadian -0.0410 -0.0032 0.0003 0.0022 -0.0026 0.0016 0.013
Dollars (0.0276) (0.0043) (0.0013) (0.0027) (0.0017) (0.0023)
Notes: 1. *Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level. 2. This table reports the
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗) (𝑗)
regression results of the following model: 𝑟𝑖,𝑡+1 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + 𝛾𝑖 𝐶𝑜𝑛𝑖,𝑡 + ∑𝑛𝑆=1 𝜃𝑠,𝑖 𝑞𝑠,𝑡 + 𝜀𝑖,𝑡+1 , where
(𝑗)=1,2,3,4, denoting financial, agricultural, mineral, or currency futures market, respectively. For example, when (𝑗)=1 refers
to financial futures market, s is S&P 500, T-bond, and Eurodollar. When (𝑗)=2 refers to agricultural futures market, s is wheat,
corn, soybeans, and live cattle. When (𝑗)=3 refers to mineral futures market, s is gold, silver, platinum, and crude oil. When
(𝑗)=4 refers to currency futures market, s is Swiss franc, British pound, Japanese yen, and Canadian dollar; 𝑟𝑖,𝑡+1 represents
𝑠&𝑝500
the futures contract i’s return at time t+1; 𝛼0 is the intercept term; 𝑟𝑚𝑡 represents the return of the S&P 500 index at time
(𝑗) (𝑗)
t; 𝐶𝑜𝑛𝑠,𝑡 is the market concentration of futures contract s at time t; 𝑞𝑠,𝑡 is the hedging pressure of futures contract s at time t;
(𝑗)
𝜀𝑖,𝑡+1 is the futures contract i’s error term. 3. The standard errors are reported in parentheses; The number of observations is
468.

Table A4
Influence of Hedging Pressure on Futures Risk Premium After Two Weeks
Panel A: Financial Futures Market (𝑗)= 1
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Systematic Risk S&P500 T-bond Eurodollars
S&P500 -0.1694*** 0.0291** -0.0336 -0.0221 0.044
(0.0467) (0.0120) (0.0208) (0.0181)
T-bond 0.0396 -0.0019 0.0126 -0.0074 0.005
(0.0339) (0.0087) (0.0151) (0.0131)
Eurodollars 0.0003 0.0002 -0.0004 -0.0015*** 0.062
(0.0007) (0.0002) (0.0003) (0.0003)
Panel B: Agricultural Futures Market (𝑗)= 2
Dependent Variable Independent Variable Adj. R2
Systematic Risk Hedging Pressure

134 Management Review, Jul 2023

 Wheat Corn Soybean Live Cattle
Wheat -0.1702 0.0657* 0.0371 -0.0012 -0.0113 0.031
(0.1046) (0.0347) (0.0426) (0.0258) (0.0282)
Corn -0.1514* 0.0251 0.0076 -0.0116 -0.0430* 0.015
(0.0904) (0.0300) (0.0369) (0.0223) (0.0243)
Soybean -0.1563** 0.0254 -0.0004 -0.0048 -0.0099 0.013
(0.0743) (0.0246) (0.0303) (0.0183) (0.0200)
Live Cattle -0.0235 -0.0207 0.0332 0.0010 0.0037 0.004
(0.0622) (0.0206) (0.0254) (0.0154) (0.0168)
Panel C: Mineral Futures Market (𝑗)= 3
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Systematic Risk Gold Silver Platinum Crude Oil
Gold -0.1281** 0.0415* -0.0294* -0.0015 0.0385 0.025
(0.0526) (0.0229) (0.0174) (0.0125) (0.0271)
Silver -0.1974** 0.0121 -0.0074 0.0000 0.0692 0.013
(0.0998) (0.0435) (0.0330) (0.0237) (0.0514)
Platinum -0.0381 -0.0071 -0.0040 0.0276* 0.0410 0.009
(0.0696) (0.0303) (0.0230) (0.0165) (0.0358)
Crude Oil -0.2179* -0.1172** 0.0144 0.0274 0.0267 0.033
(0.1124) (0.0489) (0.0371) (0.0266) (0.0578)
Panel D: Currency Futures Market (𝑗)= 4
Dependent Variable Independent Variable Adj. R2
Hedging Pressure
Canadian
Systematic Risk Swiss Francs British Pound Japanese Yen Dollars
Swiss Francs 0.0475 0.0002 0.0034 0.0012 -0.0027 0.006
(0.0397) (0.0026) (0.0051) (0.0034) (0.0031)
British Pound 0.0046 -0.0031 0.0056 0.0042 -0.0001 0.011
(0.0303) (0.0020) (0.0039) (0.0026) (0.0024)
Japanese Yen -0.0137 0.0016 0.0042 0.0003 -0.0041 0.009
(0.0321) (0.0021) (0.0041) (0.0028) (0.0025)
Canadian Dollars -0.0948*** -0.0004 0.0028 -0.0058** 0.0010 0.042
(0.0277) (0.0018) (0.0036) (0.0024) (0.0022)
Notes: 1. *Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level. 2. This table reports the
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗)
regression results of the following model:𝑟𝑖,𝑡+2 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + ∑𝑛𝑆=1 𝜃𝑠,𝑖 𝑞𝑠,𝑡 + 𝜀𝑖,𝑡+2 , where (𝑗)=1,2,3,4, denoting
financial, agricultural, mineral, or currency futures market, respectively. For example, when (𝑗)=1 refers to financial futures
market, s is S&P 500, T-bond, and Eurodollar. When (𝑗)=2 refers to agricultural futures market, s is wheat, corn, soybeans,
and live cattle. When (𝑗)=3 refers to mineral futures market, s is gold, silver, platinum, and crude oil. When (𝑗)=4 refers to
currency futures market, s is Swiss franc, British pound, Japanese yen, and Canadian dollar; 𝑟𝑖,𝑡+2 represents the futures
𝑠&𝑝500 (𝑗)
contract i’s return at time t+2; 𝛼0 is the intercept term; 𝑟𝑚𝑡 represents the return of the S&P 500 index at time t; 𝑞𝑠,𝑡 is
(𝑗)
the hedging pressure of futures contract s at time t; 𝜀𝑖,𝑡+2 is the futures contract i’s error term. 3. The standard errors are
reported in parentheses; The number of observations is 466.

Market Concentration and Risk Premium 135

Table A5
Influence of Market Concentration on Futures Risk Premium After Two Weeks
Panel A: Financial Futures Market (𝑗)= 1
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk S&P500 T-bond Eurodollars
S&P500 -0.1663*** 0.0075 0.0010 -0.0036 0.029
(0.0469) (0.0076) (0.0094) (0.0064)
T-bond 0.0367 0.0036 0.0182*** 0.0111** 0.028
(0.0334) (0.0054) (0.0067) (0.0046)
Eurodollars 0.0003 0.0000 0.0006*** -0.0002** 0.048
(0.0007) (0.0001) (0.0001) (0.0001)
Panel B: Agricultural Futures Market (𝑗)= 2
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk Wheat Corn Soybean Live Cattle
Wheat -0.1411 0.0492* 0.0425 0.0010 -0.0067 0.022
(0.1055) (0.0294) (0.0270) (0.0159) (0.0176)
Corn -0.1126 -0.0242 0.0664*** -0.0021 -0.0128 0.023
(0.0904) (0.0252) (0.0231) (0.0136) (0.0151)
Soybean -0.1365* 0.0083 0.0210 0.0011 0.0080 0.016
(0.0746) (0.0208) (0.0191) (0.0113) (0.0125)
Live Cattle -0.0111 -0.0251 0.0347** 0.0003 0.0031 0.014
(0.0622) (0.0173) (0.0159) (0.0094) (0.0104)
Panel C: Mineral Futures Market (𝑗)= 3
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk Gold Silver Platinum Crude Oil
Gold -0.1230** 0.0079 -0.0049 0.0197 0.0211* 0.028
(0.0523) (0.0155) (0.0124) (0.0173) (0.0107)
Silver -0.1919* 0.0055 -0.0070 0.0504 0.0349* 0.024
(0.0988) (0.0293) (0.0235) (0.0326) (0.0203)
Platinum -0.0380 -0.0097 0.0022 0.0609*** 0.0192 0.023
(0.0688) (0.0204) (0.0164) (0.0227) (0.0141)
Crude Oil -0.2127* -0.0035 -0.0033 -0.0786** 0.0339 0.027
(0.1122) (0.0333) (0.0267) (0.0370) (0.0230)
Panel D: Currency Futures Market (𝑗)= 4
Dependent Variable Independent Variable Adj. R2
Market Concentration
Systematic Risk Swiss Francs British Pound Japanese Yen Canadian Dollars
Swiss Francs 0.0388 -0.0012 0.0138* -0.0022 -0.0111* 0.014
(0.0397) (0.0054) (0.0075) (0.0065) (0.0061)
British Pound 0.0100 -0.0024 -0.0012 0.0027 0.0006 0.002
(0.0306) (0.0041) (0.0057) (0.0050) (0.0047)
Japanese Yen -0.0172 0.0017 0.0008 -0.0013 -0.0057 0.004
(0.0323) (0.0044) (0.0061) (0.0053) (0.0049)
Canadian Dollars -0.0965*** -0.0007 0.0051 -0.0118*** -0.0026 0.042
(0.0279) (0.0038) (0.0052) (0.0045) (0.0043)
Notes: 1. *Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level. 2. This table reports the
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗)
regression results of the following model: 𝑟𝑖,𝑡+2 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + ∑𝑛𝑆=1 𝛾𝑠,𝑖 𝐶𝑜𝑛𝑠,𝑡 + 𝜀𝑖,𝑡+2 , where (𝑗) =1,2,3,4,

136 Management Review, Jul 2023

denoting financial, agricultural, mineral, or currency futures market, respectively. For example, when (𝑗)=1 refers to financial
futures market, s is S&P 500, T-bond, and Eurodollar. When (𝑗)=2 refers to agricultural futures market, s is wheat, corn,
soybeans, and live cattle. When (𝑗)=3 refers to mineral futures market, s is gold, silver, platinum, and crude oil. When (𝑗)=4
refers to currency futures market, s is Swiss franc, British pound, Japanese yen, and Canadian dollar; 𝑟𝑖,𝑡+2 represents the
𝑠&𝑝500
futures contract i’s return at time t+2; 𝛼0 is the intercept term; 𝑟𝑚𝑡 represents the return of the S&P 500 index at time t;
(𝑗) (𝑗)
𝐶𝑜𝑛𝑠,𝑡 is the market concentration of futures contract s at time t; 𝜀𝑖,𝑡+2 is the futures contract i’s error term. 3. The standard
errors are reported in parentheses; The number of observations is 466.

Table A6
Influence of Market Concentration and Hedging Pressure on Futures Risk Premium
After Two Weeks
Panel A: Financial Futures Market (𝑗)= 1
Dependent Independent Variable Adj. R2
Variable
Market Hedging Pressure
Systematic Risk
Concentration S&P500 T-bond Eurodollars
S&P500 -0.1685*** -0.0087 0.0389** -0.0355* -0.0245 0.045
(0.0467) (0.0101) (0.0165) (0.0209) (0.0183)
T-bond 0.0374 0.0154** 0.0022 0.0070 -0.0037 0.016
(0.0338) (0.0069) (0.0089) (0.0152) (0.0132)
Eurodollars 0.0003 0.0002 0.0003 -0.0004 -0.0021*** 0.067
(0.0007) (0.0002) (0.0002) (0.0003) (0.0004)
Panel B: Agricultural Futures Market (𝑗)= 2
Dependent Independent Variable Adj. R2
Variable
Market Hedging Pressure
Systematic Risk
Concentration Wheat Corn Soybean Live Cattle
Wheat -0.1661 0.0471 0.0585* 0.0377 -0.0083 -0.0172 0.036
(0.1044) (0.0288) (0.0349) (0.0426) (0.0261) (0.0284)
Corn -0.1220 0.0696*** 0.0445 0.0008 -0.0228 -0.0574** 0.036
(0.0900) (0.0219) (0.0303) (0.0366) (0.0224) (0.0245)
Soybean -0.1567** -0.0095 0.0283 -0.0053 0.0073 -0.0091 0.014
(0.0744) (0.0203) (0.0254) (0.0321) (0.0317) (0.0201)
Live Cattle -0.0186 0.0305* -0.0244 0.0427* 0.0047 -0.0326 0.012
(0.0621) (0.0169) (0.0207) (0.0259) (0.0154) (0.0261)
Panel C: Mineral Futures Market (𝑗)= 3
Dependent Independent Variable Adj. R2
Variable
Market Hedging Pressure
Systematic Risk
Concentration Gold Silver Platinum Crude Oil
Gold -0.1281** -0.0003 0.0417 -0.0294* -0.0015 0.0384 0.025
(0.0527) (0.0156) (0.0255) (0.0174) (0.0128) (0.0274)
Silver -0.1975** -0.0020 0.0127 -0.0074 0.0005 0.0667 0.014
(0.0999) (0.0332) (0.0449) (0.0330) (0.0248) (0.0654)
Platinum -0.0447 0.0728*** -0.0154 -0.0249 0.0276* -0.0124 0.026
(0.0691) (0.0261) (0.0302) (0.0240) (0.0164) (0.0404)
Crude Oil -0.2116* 0.0242 -0.1247** 0.0279 0.0302 -0.0043 0.035
(0.1127) (0.0300) (0.0498) (0.0407) (0.0269) (0.0694)
Panel D: Currency Futures Market (𝑗)= 4

Market Concentration and Risk Premium 137

 Dependent Independent Variable Adj. R2
Variable
Systematic Risk Market Hedging Pressure
Concentration Swiss British Japanese Canadian
Francs Pound Yen Dollars
Swiss Francs 0.0533 -0.0099 0.0042 0.0039 0.0015 -0.0023 0.008
(0.0401) (0.0099) (0.0047) (0.0051) (0.0034) (0.0031)
British Pound 0.0071 -0.0086 -0.0029 0.0097* 0.0046* 0.0005 0.014
(0.0304) (0.0068) (0.0020) (0.0051) (0.0026) (0.0024)
Japanese Yen -0.0134 -0.0014 0.0017 0.0043 0.0007 -0.0041 0.010
(0.0321) (0.0064) (0.0021) (0.0041) (0.0033) (0.0025)
Canadian -0.0974*** -0.0062 -0.0002 0.0041 -0.0056** 0.0033 0.044
Dollars (0.0279) (0.0058) (0.0018) (0.0038) (0.0024) (0.0031)
Notes: 1. *Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level. 2. This table reports the
(𝑗) (𝑗) (𝑗) 𝑠&𝑝500 (𝑗) (𝑗) (𝑗) (𝑗)
regression results of the following model: 𝑟𝑖,𝑡+2 = 𝛼0 + 𝛽𝑖 𝑟𝑚𝑡 + 𝛾𝑖 𝐶𝑜𝑛𝑖,𝑡 + ∑𝑛𝑆=1 𝜃𝑠,𝑖 𝑞𝑠,𝑡 + 𝜀𝑖,𝑡+2 , where
(𝑗)=1,2,3,4, denoting financial, agricultural, mineral, or currency futures market, respectively. For example, when (𝑗)=1 refers
to financial futures market, s is S&P 500, T-bond, and Eurodollar. When (𝑗)=2 refers to agricultural futures market, s is wheat,
corn, soybeans, and live cattle. When (𝑗)=3 refers to mineral futures market, s is gold, silver, platinum, and crude oil. When
(𝑗)=4 refers to currency futures market, s is Swiss franc, British pound, Japanese yen, and Canadian dollar; 𝑟𝑖,𝑡+2 represents
𝑠&𝑝500
the futures contract i’s return at time t+2; 𝛼0 is the intercept term; 𝑟𝑚𝑡 represents the return of the S&P 500 index at time
(𝑗) (𝑗)
t; 𝐶𝑜𝑛𝑠,𝑡 is the market concentration of futures contract s at time t; 𝑞𝑠,𝑡 is the hedging pressure of futures contract s at time t;
(𝑗)
𝜀𝑖,𝑡+2 is the futures contract i’s error term. 3. The standard errors are reported in parentheses; The number of observations is
466.

138 Management Review, Jul 2023