Journal of Commodity Markets 3 (2016) 1–15

Contents lists available at ScienceDirect

Journal of Commodity Markets

journal homepage: www.elsevier.com/locate/jcomm

The impact of speculation on commodity futures markets – A

review of the findings of 100 empirical studies$
Marco Haase a, Yvonne Seiler Zimmermann b, Heinz Zimmermann c,n
a
Faculty of Economics (WWZ), University of Basel, Switzerland
b
Institut für Finanzdienstleistungen Zug IFZ, Lucerne University of Applied Sciences and Arts, Switzerland
c
Faculty of Economics (WWZ), University of Basel, Switzerland

a r t i c l e i n f o abstract

Article history: There are numerous empirical studies on the impact of speculation on commodity futures
Received 23 February 2016 markets. The papers strongly differ in terms of the focus variable (e.g. price, volatility,
Received in revised form spill-over effects) of speculative effects, the speculation measure used, and broad quality.
18 July 2016
We review and evaluate the methodology and results of 100 papers which have been
Accepted 21 July 2016
published (or are at least frequently cited) on this subject over the past decade. While the
Available online 4 August 2016
overall picture indicates that the number of studies which support and contradict the
JEL classication: criticized effects of speculation is about the same, the results shift against the criticized
G13 effects if the studies use direct measures of speculation, except for price. Applying dif-
Q02
ferent paper quality standards does not fundamentally change our findings.
C58
& 2016 Elsevier B.V. All rights reserved.
Keywords:
Speculation
Commodity futures prices
Lead–lag relationships

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
2. Methodological approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.1. Selection of papers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2.2. Focus variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.3. Impact direction, strength, and uniformity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.4. Speculation measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2.5. Broad paper quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3. The impact of speculation: all commodities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.1. All papers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.2. Analysis based on focus variable. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.3. Impact of the speculation measure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

☆
We would like to thank the editors Marcel Prokopczuk, Betty Simkins, and Sjur Westgaard for helpful comments which improved the paper sub-
stantially. The study was financially supported by the Federal Commission for Technology and Innovation CTI under Project 168641 PFES_ES. We are
grateful to the members of the expert panel for valuable comments, in particular Alfred Bühler, Marc Engelhard, Martin Hess, and Peter Sigg. Disclosure of
potential conflicts of interest: The third author was a minority shareholder of cyd Research GmbH, a provider of quantitative investment solutions in
commodity derivatives, until 2013.
n
Correspondence to: Faculty of Economics (Wirtschaftswissenschaftliches Zentrum WWZ), Peter Merian-Weg 6, 4002 Basel, Switzerland.
E-mail address: heinz.zimmermann@unibas.ch (H. Zimmermann).

http://dx.doi.org/10.1016/j.jcomm.2016.07.006
2405-8513/& 2016 Elsevier B.V. All rights reserved.
2 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15

3.4. Direct speculation measure: analysis based on focus variable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3.5. Impact of broad paper quality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.6. Analysis based on focus variable. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.7. Analysis based on speculation measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4. The impact of speculation for food commodities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Furhter Reading. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

1. Introduction
(ESMA) to implement position limits for commodity deri-
The number of empirical studies analyzing the re- vatives which, in contrast to the US, will explicitly include
lationship between financial speculation and price for- all commodity contracts including OTC positions. This
mation in commodity futures markets has increased dra- strict regulation is consistent with the European Parlia-
matically over the past decade. This research interest was ment's presumption that “financial speculation in com-
triggered by the speculation debate, which was primarily modity derivatives markets has been seen as one of the
headed by non-governmental organizations (NGOs) and drivers of the peaks in agricultural prices in 2007/2008
has been associated with food product price increases and 2010/11″ (EP 2013).1
during the years 2006–2008 as well as the growing pre- The unanimity of the public concern about the harmful
valence of index-focused investment products. Index in- and therefore criticized effects of speculation (or excess
vestors were blamed for pushing up commodity prices speculation) contrasts the academic literature on this
from their indirect holdings of long positions in com- subject. The sheer number of studies – past and present,
modity futures used by the providers of index products for theoretical and empirical – challenges the existence of a
hedging purposes. In other time periods, short positions of well-accepted view among economists and prompts a
investors and financial speculators are made responsible detailed analysis of these findings. This paper reviews 100
empirical papers.
for declining commodity prices. Irrespective of its sign,
The theoretical literature is not covered in this review.
speculation is often regarded as a major cause of increas-
However, the impact of speculation on financial markets in
ing price volatility, as well as price and volatility spill-
general, and on futures markets specifically, has long been
overs from financial markets to commodity markets
of interest to economists, as early as in the 19th century,2
(called “financialization”) with adverse effects on the real
was later elaborated by eminent economists like Kaldor,
economy. Such a negative impact would be particularly
Keynes, Hicks, Working, Friedman, Telser and others, and
fatal in the case of food prices and less developed econo-
became subject of rigorous theoretical analysis in the
mies. It is thus not surprising that the public concern of
second part of the 20th century. While the early papers
NGOs about commodity futures speculation is mainly
emphasized the risk-sharing aspect of futures markets and
about food security and hunger in third world countries. In
speculation, the more recent research focusses on the in-
November 2014, Pope Francis said that it “is also painful to
formational role of futures prices. The classical papers3
see that the struggle against hunger and malnutrition is
conclude that futures prices, while not perfectly, aggregate
hindered by market priorities, the primacy of profit, which
valuable information and help coordinating investment
have reduced foodstuffs to a commodity like any other,
and production decisions, thereby stabilizing the economy.
subject to speculation, also of a financial nature”. This
Other papers have argued rational speculation may have,
public debate is in no way new; as shortly addressed in the or always has, destabilizing price effects, depending on the
Conclusions of this paper, essentially the same debate took economic setting used in the model: examples include
place a century ago, in the US as well as in Europe. specific institutional constraints, behavioral patterns, and
The consequences of this public debate are very real: collective (i.e. general equilibrium and herding) effects of
Most European banks have stopped selling commodity speculative activity.4 Therefore, the predictions of the
related products, and regulators impose tighter constraints theoretical literature are far from unambiguous.
in trading commodity futures if they are not explicitly used Given the enormous output of empirical research
for commercial or non-financial hedging purposes, or by
end-users of commodities. In the US, the Dodd-Frank act
1
being released after the recent financial crisis, explicitly Research on Regulating Agricultural Derivatives Markets, November
2013, Directorate-General for Internal Policies, Policy Department B,
requires “strong measures to limit speculation in agri-
Structural and Cohesion Policies.
cultural commodities”, which materialized in tighter and 2
The debate among economists was particularly heated in German
more elaborate position limits on exchange-traded con- speaking Europe in the 1890 s in view of the new Stock Exchange Law
tracts of 28 commodities, implemented by the Commodity and the regulation of the Berlin produce exchange. A few remarks can be
found in the Conclusions.
Futures Trading Commission (CFTC). In Europe, the second 3
The seminal papers include Grossman (1977) and Danthine (1978).
Markets in Financial Instruments Directive (MIFID II) 4
Representative papers are by Hart and Kreps (1986), De Long et al.
mandates the European Securities and Markets Authority (1990), Froot et al. (1992), or Kyle and Xiong (2001) among many others.
 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15
addressing the impact of speculation on commodity fu-
tures markets, it is not surprising that a few survey papers
or metastudies have been released recently. These surveys
are important because the range of papers to which the
public debate refers is extremely broad, both in terms of
sophistication, methodology, and data, and ranges from
unpublished, non-technical memorandums to top ranked
journals. The review paper by Shutes and Meijerink (2012)
reviews a total of 40 “quantitative and qualitative studies
and opinion pieces” (21 and 19, respectively), released
between 2006–2011, published and unpublished. The re-
view only includes papers dealing with (index) speculation
and agricultural commodities. The papers are judged on
six quality criteria and classified into three quality cate-
gories. The authors find “no consistent evidence in the
academic literature that the large inflow of speculative
capital by, amongst others, index funds has led to higher
prices or more volatility on the mid to long term.” While
there is some evidence for “very small and short-term
volatility”, the authors conclude that these effects do not
have distortive effects on traders and producers of com-
modities, but rather on financial markets themselves.
The review paper by Brümmer et al. (2013) has been
released as part of the ULYSSES project5 co-funded by the
European Commission, seeking a better understanding of
the volatility of agricultural commodity markets. The sur-
vey covers theoretical and empirical research papers re-
leased between 2001 and 2012; the empirical papers ex-
amine volatility drivers and spill-over effects, as well as
the interaction between spot and futures price volatility.
Apart from peer reviewed journal articles, the survey
covers “a selected number of working papers, policy briefs,
and discussion papers (‘grey literature’) from international
organizations or research institutes”. The paper points to
several methodological difficulties of empirical testing
such as the interpretation of Granger causality tests, or the
definition and measurement of speculation, and is there-
fore extremely vague about its role on volatility and spill-
overs effects. The authors conclude that “the literature
seems in broad agreement the fundamental factors explain
most of the observed price volatility increases in recent
times.” Based on the methodological difficulties, the role of
speculation is not explicitly examined.
The survey by Will et al. (2016) has attracted particular
interest – and triggered controversy – in the public. The
paper reviews the major findings of 35 empirical studies
published from 2010 to 2012 addressing the impact of
speculation on agricultural commodity prices. They find
“little supporting evidence that the recent increase in fi-
nancial speculation has caused either (a) of the price level,
or (b) the price volatility in agricultural markets to rise.
Rather, fundamental factors are responsible for this.”
Methodologically, our paper is closely related to Brüm-
mer et al. (2013) in that the measurement of speculation is
regarded as a key aspect in evaluating the empirical find-
ings of the papers. Also, we take a close methodological
5
The name of the project stands for: “Understanding and coping
with food markets volatility towards more stable world and EU food
systems”.
3
look at the studies and try to evaluate its quality impact on
findings. With respect to the objective of our paper, how-
ever, our paper is more pragmatic and more in the spirit of
Will et al. (2016): We do not summarize the empirical
findings of the study, but classify the papers and their
findings according to various criteria. We therefore call it,
like in medicine, a metastudy, since it is not a typical survey
or review of papers, but rather an evaluation of the docu-
mented impact (or non-impact) of measures and actions, or
activities, on specific focus variables.
In this study, we evaluate a total of 100 empirical pa-
pers which analyze the relationship between financial
speculation and its impact on commodity futures prices.
The effects are classified for six focus variables (price le-
vels, returns, risk premiums, spreads, volatilities, and
spillover effects) on which speculation has a potential
impact. The papers include published and unpublished
papers which form the heart of the current debate, most of
them released, respectively published, after 2009. Apart
from a broader range of papers examined, the distin-
guishing feature of our study is the examination of the
impact of the measurement of speculation. There are stu-
dies which do not define speculation at all and still claim
to examine the impact of speculation, others use rough
proxy variables (like e.g. the stock market), and others
papers use commodity-market specific statistics (i.e. CFTC
position data). We also analyze whether the quality of the
papers, as measured by our own quality classification and
journal-based citation ranks, has an effect on the findings.
The paper is structured as follows. Section 2 describes our
methodological approach used to evaluate and classify the
papers. Section 3 contains the major findings for all papers
(first subsection), respectively for the papers of superior
quality (second subsection). Because speculation with food
commodities has been criticized for being responsible for the
price boom in the 2000s, we analyze the specific results for
these commodities in Section 4. The final Section 5 sum-
marizes the most important findings of this paper.
2. Methodological approach
In this section, we describe the methodological ap-
proach of our metastudy: How the papers are selected, the
specification of the focus variables (i.e. the variables on
which speculation has a potential impact), the classifica-
tion of direction and strength of impact, the measurement
of speculation, and the assessment of broad paper quality.
2.1. Selection of papers
The objective of this metastudy is to critically assess the
methodology and findings of empirical papers which con-
tribute to the “current debate” on the impact of speculation
on commodity futures markets, which accelerated with the
boom of commodity prices and index trading between 2006
and 2008. Therefore, all but six papers were released or
published in 2009 or later, the earliest is from 2002.6 The
6
The few papers which date before 2009 are selected because of
4 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15
temporal cut-off of the reviewed paper is at March 2015.
The range of papers considered for the metastudy is
comparable to Shutes and Meijerink (2012) and Will et al.
(2016) with the distinguishing feature that our papers are
not restricted to agricultural commodities, but include all
commodities included in typical commodity indexes
(crude oil, metals, live cattle). However, we exclude papers
which (exclusively) analyze precious metals. The papers by
Irwin and Sanders (2011) and Cheng and Xiong (2014)
contain references to a representative range of papers
considered for our study. We moreover include papers to
which we found frequent references in the core sample of
studies or in the public discussion. This also includes a few
unpublished working papers7 or studies released by or-
ganizations (such as the World Bank, IMF, UNCTAD, or
ECB) which have an unclear publication status. The inclu-
sion of papers which have not (yet) been published in
academic journals also mitigates a possible “publication
bias” which may understate non-findings in the empirical
literature. However, including a broader range of papers
also makes it necessary to evaluate – at least in a rudi-
mentary way – the quality of the papers. This leads us to
include a total of 100 papers in our metastudy.8
Of course, any selection remains to a certain extent
subjective. This has also to do with the fact that several
authors, or teams of authors, have published several pa-
pers under different titles with similar (sometimes up-
dated) results. We did our best to eliminate these studies
and include only the most recent version.
The 100 papers are classified across the following criteria:
2.2. Focus variables
Speculation is criticized of having an adverse impact on
several economic magnitudes. In this paper, the potential
effects are grouped into six categories, abbreviated by
“focus variables”:

Spot or futures price levels, including deterministic
price trends or structural breaks (“Price”)

Realized returns on futures contracts (“Return”)

Risk premiums on futures contracts, based on an equi-
librium relationship between e.g. hedging or storage
costs, risk appetite, scarcity, backwardation and ex-
pected commodity futures returns (“Risk premiums”)

Price spreads between futures contracts, mostly be-
tween long and short dated (eventually spot) contracts
(“Spreads”)

Volatility of spot or futures returns (“Volatility”)

Spill-over effects between financial markets and com-
modity prices respective returns (“Spill-over”).
Most paper use futures prices in their empirical ana-
lysis, because they reflect the most accurate information
(footnote continued)
their relevance for subsequent studies, as reflected in the citations.
7
However, only unpublished papers from 2010 and later are
considered.
8
In fact, our selection ended with 99 papers, and one was added to
complete one hundred.
about the underlying commodities. It is common practice
in the empirical literature to use the price of the nearby
contract9 as spot price, and a few papers analyzing “spot”
prices (e.g. in calculating spot-futures spreads) effectively
use nearby prices. In just a few cases, the nature of ana-
lyzed prices cannot be determined unambiguously.
Therefore, by far most of the prices analyzed in the papers
are futures prices.
Several of the papers analyze the effects for a multitude
of focus variables. In this case, the papers appear multiple
times and are treated as different “studies”. Therefore,
with respect to the focus variables, the 100 papers re-
present 122 individual studies.
2.3. Impact direction, strength, and uniformity
The empirical findings of each study are classified into
five categories, depending on the uniformity and strength
the reported effects of speculation: the impact will be re-
ferred to as the “impact direction”, rated as “ þ2″ for
clearly reinforcing, “ þ1″ for mixed reinforcing, “0″ for
none, “  1″ for mixed weakening and “  2″ for a clearly
weakening impact.
A study is classified as ( þ2) or ( 2) if all essential
findings support a specific effect, in terms of impact di-
rection and strength (statistical significance). A study is
classified in the “mixed” category ( þ1) or (  1) if apart
from non-findings a statistically significant impact is re-
ported for at least a subset of results, for example for in-
dividual commodities or for specific time periods. If a
study finds reinforcing and weakening effects, it is classi-
fied as “mixed reinforcing” (þ 1) which intentionally
overstates the criticized effect of speculation.
The impact direction can be exemplified as follows: A
reinforcing (weakening) price effect means that specula-
tion has an increasing (decreasing) effect on the price level
or its trend. For volatility, it means that the documented
effect of speculation is in favor of a volatility increase. A
reinforcing spill-over effect of speculation occurs, for in-
stance, if speculation accelerates the transmission of vo-
latility from financial to commodity markets.
For several studies, the impact direction or strength
cannot be determined based on the available results; for
example, a paper only contains significance levels of test
statistics, but not the estimated coefficients. In these cases,
the study is classified as “undetermined”.
We evaluate the effects of speculation based on the
actual empirical findings reported in the empirical part of
the papers, which is in some instances not identical to the
authors’ own interpretation or concluding summary of
findings.
2.4. Speculation measures
“Speculation” is measured in various ways across the
different studies. We differentiate between four categories,
9
Among all outstanding contracts of a commodity, the one with the
shortest maturity is called “nearby” contract; for most futures contracts,
the nearby contract expires in the current month.
 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15
which will hereafter be referred to as “speculation
measures”:

Unspecific

Indirect

Proxy

Direct
“Unspecific” studies discuss speculators or speculation
without defining it. “Indirect” studies employ empirical
observations such as structural breaks or structural shifts
in correlations which are indirectly attributed to specula-
tion without substantiating the relationship. “Proxy” stu-
dies include specific variables which are representative for
speculative activities, such as stock or oil prices. Of course,
in both cases, it is questionable whether the measures are
actually related to speculation. “Direct” studies apply ex-
plicit speculation measures, ideally based on position data
and trader classifications by supervisory authorities or
exchanges.
2.5. Broad paper quality
The broad range of papers, published and unpublished,
makes it necessary to apply a rough quality assessment of
the papers. At first glance it may be surprising to apply a
somehow subjective criterion in our review. However, a
broad methodological triage is vital, particularly when
considering unpublished papers (but not only for this
reason).
We therefore classify the papers into two categories (A
and B) based on three main criteria: appropriateness of
empirical methodology, traceability of test results, and
adequacy of test interpretation. The following examples
are representative of identified shortcomings:
– Volatility effects are meaningless if volatility is esti-
mated from nonstationary or spatially-averaged data.
– When a variable indicates a shock percentage of 10%
under variance decomposition, this does not mean that
this variable explains 10% of the price increase.
– When the correlations between commodity prices in-
crease during the same time period that index trading
increased significantly, this does not mean that the in-
creased correlations are causally linked to index trading.
– What cannot be described by supply and demand factors
or inventory changes does not necessarily need to be
attributed to an unspecified effect of speculation.
Traceability of results requires that the empirical tests
must be replicable, i.e. a clear description of the data, the
estimated model, and a state-of-the-art presentation of
empirical results including the relevant statistical tests.
Those paper qualifying these criteria are graded A, the
other B.
The advantage of a proprietary – albeit somehow sub-
jective – evaluation of papers is that it allows to apply con-
sistent evaluation criteria to all papers, published and un-
published. However, as shown in Section 3.2, our triage is
rather moderate and excludes less than 20% of the studies,
i.e. eliminates merely those below a minimum standard.
5
Therefore, we alternatively apply a journal-ranking criterion
for the published papers, the Elsevier based SCImago Journal
Rank (SJR). We find 55 papers published in journals with a
minimum ranking of 0.5. We also report the results for an
SJR-breakpoint of 1.0, and for an alternative journal ranking,
the Impact Factor (IF) released by Thomson Reuters, with the
same breakpoint. In contrast to the SJR, the IF tends to put
more emphasis on the number of citations than the quality
and prestige of a journal.10
A potential problem of journal-based quality evaluation
is a “publication” bias: it is well known that published
work requires methodological innovation, which sup-
presses research using well-established methods. In addi-
tion, there may be a bias towards papers with significant
test results in published work.
3. The impact of speculation: all commodities
In this section, the results of the impact of speculation
on the full spectrum of commodities are discussed. Section
3.1 includes all papers, while Section 3.2 only contains the
results of the papers with quality grade A, respectively,
with superior journal ranking.
3.1. All papers
In this section, the studies in our overall sample are
classified based on the six focus variables (122 studies) and
based on the measurement of speculation (131 studies).
We also analyze the speculative impact for each focus
variable for the subsample of 52 studies which use direct
speculation measures.
3.2. Analysis based on focus variable
A breakdown of the results for the six focus variables
can be found in Table 1. It is evident that speculation most
likely has a reinforcing (clear or mixed) influence on fu-
tures prices (11% þ 33%), in contrast to (clear or mixed)
weakening effects (11%). Notice, however, that the number
of studies with undetermined effects is substantial for the
price variable (14 of 32). Predominant weakening effects
can most clearly be observed for returns and risk pre-
miums. For returns, some 50% of the studies find weak-
ening effects (33% þ 17%), and about 20% find reinforcing
effects (8% þ13%). These are essentially the same figures as
for the risk premiums (weakening effects: 33þ17%, re-
inforcing effects: 17 þ17%). For the other focus variables,
the reinforcing and weakening effects balance out.
3.3. Impact of the speculation measure
The role of the speculation measure used in the studies
is analyzed in Table 2. The results reveal that the majority
10
“… – popular journals cited frequently by journals of low prestige
have high impact factors and low SJRs, whereas journals that are pres-
tigious may be cited less but by more prestigious journals, giving them
high SJRs but lower impact factors.” From: Nature News 451.7174 (2008):
6–6 (D. Butler).
6 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15

Table 1 Table 3
Impact of speculation on focus variables: all papers. Impact of speculation on focus variables: direct speculation measures
only.
Focus variable Impact direction Total undet
Focus variable Impact direction Total undet
2 1 0 1 2
2 1 0 1 2
Price 11% 33% 44% 11% 0% 18 14
Returns 8% 13% 29% 33% 17% 24 1 Price 33% 17% 33% 17% 0% 6 2
Risk Premiums 17% 17% 17% 33% 17% 6 2 Returns 10% 10% 30% 30% 20% 20 1
Spreads 0% 25% 50% 25% 0% 4 1 Risk Premiums 0% 25% 25% 50% 0% 4 1
Volatility 10% 33% 14% 33% 10% 21 7 Spreads 0% 50% 0% 50% 0% 2 1
Spill-Over 23% 18% 27% 23% 9% 22 2 Volatility 15% 15% 15% 38% 15% 13 2
Spill-Over 29% 0% 0% 57% 14% 7 0

All 13% 23% 28% 26% 9% 95 27

All 15% 13% 21% 37% 13% 52 7
The focus variables are those economic magnitudes on which speculation
is supposed to have a statistical impact; they are described in Section 2 of The focus variables are those economic magnitudes on which speculation
this paper. The impact direction indicates how speculation affects the is supposed to have a statistical impact; they are described in Section 2 of
focus variable. Thereby, 2 means a clear reinforcing impact, 1 a mixed this paper. The impact direction indicates how speculation affects the
reinforcing impact, 0 no impact,  1 a mixed weakening impact and  2 a focus variable. Thereby, 2 means a clear reinforcing impact, 1 a mixed
clear weakening impact. The percentages refer to the Total number of reinforcing impact, 0 no impact,  1 a mixed weakening impact and  2 a
observations displayed in the second last column. “undet” in the last clear weakening impact. The percentages refer to the Total number of
column refers to studies which cannot be categorized to the 5 impact observations displayed in the second last column. “undet” in the last
categories. column refers to studies which cannot be categorized to the 5 impact
categories.

of studies (52 þ11) measures the influence of speculation (28%) to weakening effects (50%). Of course, direct measures
directly, i.e. use a specific well-defined speculation mea- are clearly preferable from a methodological perspective.
sure (such as the Working T index); a fewer number of Overall, the specification of the speculation measure (if spe-
papers uses proxies (24) or indirect quantities (10 þ5). cified at all) has a strong impact on the findings.
Surprisingly, 28 studies do not operationalize speculation
at all (classified as “unspecified”) but still come up with 3.4. Direct speculation measure: analysis based on focus
conclusions about the role of speculation. variable
Table 2 reveals that most (20 out of 28) studies with un-
specified speculation measures or proxies get results which In Table 3, the speculative impact for each focus vari-
cannot be clearly attributed to an impact category, which is able is displayed for the subsample of 52 studies which use
hardly surprising. The results of the remaining studies are direct speculation measures. Compared to Table 1, the
mixed and do not reveal clearly weakening or reinforcing ef- number of clearly reinforcing effects on price is larger (33%
fects. If indirect or proxy variables are used to measure vs. 11%), while the effect on returns is unchanged. This
speculation, about half of the respective studies report re- contrasts the findings for the volatility and spill-over ef-
inforcing effects (45% resp. 50%), compared to 18% resp. 21% fects: While the number of clearly reinforcing effects in-
which support weakening effects. Interestingly, if speculation creases only slightly (from 10% to 15%, and from 23% to
is measured directly in the empirical tests, the distribution 29%), there is an overall shift towards (mixed or clearly)
shifts from positive to negative scores, i.e. from reinforcing weakening effects here: The percentage of weakening
volatility effects is 53% compared to 43% in the overall
Table 2 sample, and the percentage of weakening spill-over effects
Impact of speculation based on speculation measures: all papers. is 71% compared to 31%. This shift can also be observed for
risk premiums and spreads; no study reports clearly re-
Speculation measure Impact direction Total undet inforcing or weakening effects any more.
2 1 0 1 2
In general, these observations reveal that reinforcing
price level effects are supported by those studies that use
Unspecified 0% 25% 63% 13% 0% 8 20 superior speculation measures; but at the same time, these
Indirect 18% 27% 36% 18% 0% 11 5 studies find more evidence for weakening effects of
Proxy 8% 42% 29% 13% 8% 24 0
speculation on volatility, and in particular on spill-overs.
Direct 15% 13% 21% 37% 13% 52 11

3.5. Impact of broad paper quality

ALL 13% 23% 28% 26% 9% 95 36

The variables used to measure speculation are classified into four cate-
gories; they are described in Section 2 of this paper. The impact direction
indicates how speculation affects the focus variable. Thereby, 2 means a
clear reinforcing impact, 1 a mixed reinforcing impact, 0 no impact,  1 a
mixed weakening impact and  2 a clear weakening impact. The per-
centages refer to the Total number of observations displayed in the sec-
ond last column. “undet” in the last column refers to studies which
cannot be categorized to the 5 impact categories.
In this section, our analysis is restricted to those papers
which meet some minimum quality standards, either
based on our own criteria as defined in Section 2 (grade A,
qualified by appropriateness of empirical methodology,
traceability of test results, and adequacy of test inter-
pretation), or based on minimum journal rankings for the
published papers.
 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15 7

Table 4 We test for stability of our findings by applying a more

Impact of speculation on focus variables: studies from quality category A. restrictive SRJ-breakpoint of 1.0 instead of 0.5, and by
adding the “Impact Factor” (IF) of the journals as an al-
Focus variable Impact direction Total undet
ternative journal-ranking based quality criterion, applying
2 1 0 1 2 a breakpoint of 1.0 as well. This leads to a selection of,
respectively, 38 and 52 studies. The results of this analysis
Price 13% 20% 53% 13% 0% 15 8
are displayed in the Appendix, Tables 10 and 11. They are
Returns 9% 9% 30% 35% 17% 23 1
Risk Premiums 17% 17% 17% 33% 17% 6 1 very similar to the previous findings. Again, the relative
Spreads 0% 25% 50% 25% 0% 4 1 number of studies that report mixed reinforcing effects on
Volatility 5% 26% 16% 37% 16% 19 3 volatility exceeds the relative number of studies with
Spill-Over 21% 21% 32% 16% 11% 19 2
mixed weakening effects (50% vs. 17%, and 43% vs. 29%).
Again, no study reports unambiguous effects on volatility.
All 12% 19% 31% 27% 12% 86 16 Also notice that the relative number of studies with zero-
classified (i.e. ambiguous) results with respect to price
The focus variables are those economic magnitudes on which speculation
is supposed to have a statistical impact; they are described in Section 2 of
levels or price trend tends to increase with stricter quality
this paper. The impact direction indicates how speculation affects the standards (71% and 67%). However, the number of studies
focus variable. Thereby, 2 means a clear reinforcing impact, 1 a mixed that find (mixed or clearly) reinforcing spill-over-effects of
reinforcing impact, 0 no impact,  1 a mixed weakening impact and  2 a speculation tends to increase with quality (30%þ30%, and
clear weakening impact. The percentages refer to the Total number of
observations displayed in the second last column. “undet” in the last
27%þ27%), but the overall picture is the same as in the full
column refers to studies which cannot be categorized to the 5 impact sample.
categories. Papers in quality category A are classified based on the criteria Therefore, the surprising finding of this subsection is
described in Section 2. that quality screening does not lead to different conclu-
sions about the effects of speculation on our six focus
3.6. Analysis based on focus variable variables compared to the full sample of studies. An ex-
ception applies to volatility where stricter quality criteria
Tables 4 and 5 show the influence of speculation on show slightly more reinforcing (but mixed) effects. Al-
each focus variable. Our own grading identifies a total of though the high quality papers tend to find more reinfor-
102 (86 þ16) A-ranked studies, the SJR-classification with cing effects on volatility, they do not find unambiguous
a minimum score of 0.5 identifies 55 (44 þ11) studies; evidence for it.
they form a subsample of our A-ranked studies with a
single exception. The major focus variables of our 3.7. Analysis based on speculation measure
A-ranked studies are returns (24 studies),11 prices (23),
and volatilities (22), while the SJR-selected papers focus on Tables 6 and 7 provide an overview of the studies in
prices (16), spill-over effects (12), and volatilities (9). quality categories A and SRJ ≥ 0.5, respectively, grouped by
Although the size of the sub-samples is fairly different, impact direction and speculation measure. While the
the insights from Tables 4 and 5 are quite similar, except majority of A classified studies measure the influence of
with respect to the impact of speculation on volatility. The speculation directly (as in Table 2 for the overall sample),
SRJ-classification reveals that 43% of the studies show a the use of proxy variables becomes equally important for
mixed reinforcing impact on volatility compared to 31% the SRJ-selected studies; this can also be observed for the
(mixed or clearly reinforcing) in our own grading (A). In subset of studies selected by SRJ ≥ 1.0 and IF ≥ 1.0 (in the
contrast, our grading identifies 53% of the studies with Appendix, Tables 12 and 13).
mixed or clearly weakening effects on volatility compared A comparison of the results between Tables 6 and 7
to 29% (mixed weakening) for the SRJ-classification. gives a fairly consistent picture. We observe a slightly
The results are also largely consistent with those for the larger number of studies with a reinforcing effect for the
full sample of papers (Table 1), again with the exception of SRJ-ranked papers compared to the A ranked. This is due
the volatility effect for which the studies reveal a fully to the fact that relatively more studies use proxy variables
symmetric picture on (clearly or mixed) reinforcing and in SRJ-sample, and proxy variables tend to identify re-
weakening effects (10% and 33%, respectively). Therefore, inforcing effects – this was already observed in Table 2 and
and somehow ironically, the quality screening based on is confirmed here. Moreover, for the studies using direct
the two quality criteria merely confirms the mixed effect of measures of speculation, the SRJ-sample includes a few
speculation on commodity price volatility. Across all focus more papers with mixed reinforcing effects than our A
variables, the number of no-findings is only slightly larger sample (22% vs. 10%). The same can be observed for the
here (31% and 36%) compared to the overall sample in studies selected by SRJ ≥ 1.0 and IF ≥ 1.0 (in the Appendix,
Table 1 (28%).12 Tables 12 and 13).
Compared to Table 2, i.e. the full sample of studies, we
11
This number also includes the papers where the impact direction do not find fundamentally different results here. The pre-
is undetermined (last column in the Tables). dominance of weakening effects, if direct measures of
12
The observation is not much different between our own grading
(applied to all papers, including unpublished) and the SJR-based grading
(applicable to published papers only). This is inconsistent with the pub- (footnote continued)
lication bias which predicts that published work has a bias towards (significant) findings.
8 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15

Table 5 Table 7
Impact of speculation on focus variables: studies from journal rankings Impact of speculation based on speculation measures: studies from
SRJ Z 0.5. journal rankings SRJ Z 0.5.

Focus variable Impact direction Total undet Speculation measure Impact direction Total undet

2 1 0 1 2 2 1 0 1 2

Price 11% 11% 67% 11% 0% 9 7 Unspecified 0% 0% 100% 0% 0% 4 11

Returns 0% 14% 43% 29% 14% 7 1 Indirect 25% 0% 75% 0% 0% 4 0
Risk Premiums 20% 20% 0% 40% 20% 5 0 Proxy 11% 33% 28% 17% 11% 18 0
Spreads 0% 25% 50% 25% 0% 4 1 Direct 11% 22% 22% 39% 6% 18 4
Volatility 0% 43% 29% 29% 0% 7 2
Spill-Over 25% 25% 25% 17% 8% 12 0
All 11% 23% 36% 23% 7% 44 15

All 11% 23% 36% 23% 7% 44 11
The focus variables are those economic magnitudes on which speculation
is supposed to have a statistical impact; they are described in Section 2 of
this paper. The impact direction indicates how speculation affects the
focus variable. Thereby, 2 means a clear reinforcing impact, 1 a mixed
reinforcing impact, 0 no impact,  1 a mixed weakening impact and  2 a
clear weakening impact. The percentages refer to the Total number of
observations displayed in the second last column. “undet” in the last
column refers to studies which cannot be categorized to the 5 impact
categories. Papers are selected based on the Elsevier SCImago Journal
Rank (SJR) above 0.5.
The variables used to measure speculation are classified into four cate-
gories; they are described in Section 2 of this paper. The impact direction
indicates how speculation affects the focus variable. Thereby, 2 means a
clear reinforcing impact, 1 a mixed reinforcing impact, 0 no impact,  1 a
mixed weakening impact and  2 a clear weakening impact. The per-
centages refer to the Total number of observations displayed in the sec-
ond last column. “undet” in the last column refers to studies which
cannot be categorized to the 5 impact categories. Papers are selected
based on the Elsevier SCImago Journal Rank (SJR) above 0.5.
change our findings.

Table 6
Impact of speculation based on speculation measures: studies from
4. The impact of speculation for food commodities
quality category A.

Speculation measure Impact direction Total undet Speculation with agricultural and food commodities
has been criticized for being responsible for the com-
2 1 0 1 2 modity price boom in the 2000s. In the US, the criticism
got public attention by Michael Masters, an investment
Unspecified 0% 25% 63% 13% 0% 8 11
Indirect 20% 20% 40% 20% 0% 10 5 manager, in his testimony before a Senate committee ar-
Proxy 5% 35% 35% 15% 10% 20 0 guing that futures market speculation caused a bubble in
Direct 15% 10% 23% 35% 17% 48 8 oil prices in 2007–2008. The criticism was quickly ex-
tended to grain commodities in a report of the U.S. Senate
All 12% 19% 31% 27% 12% 86 24 (U.S. Senate, 2009) and got its attention across the Atlantic
in public statements by the British Prime Minister, the
The variables used to measure speculation are classified into four cate-
French President, the German Finance Minister, and the
gories; they are described in Section 2 of this paper. The impact direction
indicates how speculation affects the focus variable. Thereby, 2 means a
Pope.
clear reinforcing impact, 1 a mixed reinforcing impact, 0 no impact,  1 a The adverse effect of commodity speculation was gra-
mixed weakening impact and  2 a clear weakening impact. The per- dually extended to agricultural commodities in general,
centages refer to the Total number of observations displayed in the sec- food prices, and hunger in third world countries, and
ond last column. “undet” in the last column refers to studies which
quickly provoked a public debate about the ethics of
cannot be categorized to the 5 impact categories. Papers in quality ca-
tegory A are classified based on the criteria described in Section 2. speculative behavior. In Switzerland, the young socialist
party even launched a public referendum on “No spec-
ulation on foodstuffs” which was declined in February
speculation are used, is only slightly impaired by the im- 2016. However, this issue remains a hot topic, in particular
posed quality standards – the relative number of studies in the context of the ongoing regulatory debate on tigh-
with weakening effects still exceeds the number of studies tening position limits of commodity contracts on- and off-
with reinforcing effects. Still, if proxy variables are used, exchanges.
the reverse observation is true and is stable across the For this reason, the results have been broken down to
different quality standards applied. Therefore, the mea- individual food commodities as far as the papers permit.
surement of speculation has a substantial impact upon the Each specific food commodity analyzed in the papers de-
empirical results, even after controlling for paper quality. fines a specific “case”. This gives a total of 212 cases, dis-
We conclude from these observations: While the tributed among the following food products: wheat, corn,
overall picture indicates that the number of studies which soybeans, rice, coffee, cocoa, feeder cattle, live cattle, lean
support and contradict the criticized effects of speculation hogs, sugar and orange juice. Given the small number of
is about the same, the results shift against the criticized studies about orange juice and rice, we drop these from
(mixed or clearly reinforcing) effects if the studies use the subsequent analysis.
direct measures of speculation, except for price. Applying In the subsequent Tables, we analyze the speculative
different paper quality standards does not fundamentally impact on the six focus variables for the overall sample
Table 8
Impact of speculation on focus variables: food commodities.

M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15

"\wide-
hattbchmode-
val"Max
Focus Wheat Corn Soybean Coffee Cocoa Feeder Cattle Live Cattle Lean Hogs Sugar Total
variable
Impact direction Impact direction Impact direction Impact direction Impact direction Impact direction Impact direction Impact direction Impact direction Impact direction

1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1

Price 25% 75% 0% 50% 50% 0% 50% 25% 25% 33% 67% 0% 0% 100% 0% 0% 100% 0% 100% 0% 0% 50% 25% 25% 0% 50% 50% 12 15 3
Return 27% 53% 20% 6% 65% 29% 27% 47% 27% 20% 40% 40% 25% 50% 25% 0% 78% 22% 22% 56% 22% 11% 56% 33% 13% 63% 25% 16 54 25
Risk 0% 50% 50% 0% 0% 100% 0% 0% 100% – – – – – – – – – – – – – – – – – – 0 1 3
Premium
Spread 50% 50% 0% 50% 0% 50% 50% 50% 0% – – – – – – – – – – – – – – – – – – 3 2 1
Volatility 10% 40% 50% 27% 55% 18% 33% 33% 33% 20% 60% 20% 40% 20% 40% 50% 50% 0% 0% 40% 60% 40% 60% 0% 0% 40% 60% 14 26 19
Spill-Over 67% 33% 0% 50% 25% 25% 0% 100% 0% 100% 0% 0% 0% 100% 0% 0% 100% 0% 0% 100% 0% 0% 100% 0% 100% 0% 0% 6 8 1

All 25% 50% 25% 23% 51% 26% 31% 41% 28% 29% 50% 21% 25% 50% 25% 12% 76% 12% 30% 45% 25% 26% 53% 21% 13% 50% 38%
Number 9 18 9 9 20 10 10 13 9 4 7 3 4 8 4 2 13 2 6 9 5 5 10 4 2 8 6 51 106 52

Explanations: The focus variables are those economic magnitudes on which speculation is supposed to have a statistical impact; they are described in Section 2 of this paper. The impact direction indicates how
speculation affects the focus variable. Thereby, 1 means a clear or mixed reinforcing impact, 0 no impact,  1 a clear or mixed weakening impact. The percentages refer to the Total number of observations displayed
in the last column.

9
 10
Table 9
Impact of speculation on focus variables, direct speculation measures only: food commodities.

M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15

"\wide-
hattbchmode-
val"Max
Focus Wheat Corn Soybean Coffee Cocoa Feeder Cattle Live Cattle Lean Hogs Sugar Total
variable
Impact direction Impact direction Impact direction Impact direction Impact direction Impact direction Impact direction Impact direction Impact direction Impact direction

1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1 1 0 1

Price 25% 75% 0% 50% 50% 0% 50% 25% 25% 33% 67% 0% 0% 100% 0% 0% 100% 0% 100% 0% 0% 50% 25% 25% 0% 50% 50% 12 15 3
Return 29% 57% 14% 6% 69% 25% 31% 54% 15% 25% 50% 25% 29% 57% 14% 0% 88% 13% 25% 63% 13% 13% 63% 25% 14% 71% 14% 16 54 15
Risk 0% 50% 50% 0% 0% 100% 0% 0% 100% – – – – – – – – – – – – – – – – – – 0 1 3
Premium
Spread 50% 50% 0% 50% 0% 50% 50% 50% 0% – – – – – – – – – – – – – – – – – – 3 2 1
Volatility 10% 40% 50% 27% 55% 18% 33% 33% 33% 20% 60% 20% 40% 20% 40% 50% 50% 0% 0% 40% 60% 40% 60% 0% 0% 40% 60% 14 26 19
Spill-Over 50% 50% 0% 33% 33% 33% 0% 100% 0% 100% 0% 0% 0% 100% 0% 0% 100% 0% 0% 100% 0% 0% 100% 0% 100% 0% 0% 4 8 1

All 24% 53% 24% 22% 54% 24% 33% 43% 23% 31% 54% 15% 27% 53% 20% 13% 81% 6% 32% 47% 21% 28% 56% 17% 13% 53% 33%
Number 8 18 8 8 20 9 10 13 7 4 7 2 4 8 3 2 13 1 6 9 4 5 10 3 2 8 5 49 106 42

Explanations: The focus variables are those economic magnitudes on which speculation is supposed to have a statistical impact; they are described in Section 2 of this paper. The impact direction indicates how
speculation affects the focus variable. Thereby, 1 means a clear or mixed reinforcing impact, 0 no impact,  1 a clear or mixed weakening impact. The percentages refer to the Total number of observations displayed
in the last column.
 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15
(209 cases) and the subsample of studies using direct
speculation measures (197). We do not display a break-
down of the results for quality classification because, as
discussed in the previous section, the quality dimension
does not change the interpretation of the results sub-
stantially when direct speculation measures are used, and
the number of “cases” for each commodity would become
fairly small.
The number of cases and their relative frequencies
across three impact directions are presented in Tables 8
and 9. Here, in contrast to the findings in Section 2, the
influence of speculation (impact direction) is only re-
presented by degrees of 1 for reinforcing impact, 0 for no
impact and  1 for a weakening impact. This is due to the
low number of observations.13
If all papers are included (see Table 8), there is no
evidence for predominantly weakening or reinforcing im-
pacts. However, the percentage of reinforcing impacts is
slightly higher than for weakening impacts in the case of
soybeans, coffee, live cattle and lean hogs.
This result does not change if only those studies which
measure speculation directly are considered (Table 9). With
the exception of wheat, corn and sugar, slightly more re-
inforcing effects are found. But overall, across all commodities,
the cases which find no impact of speculation (classification
“zero”) dominate. This is particularly true for the agricultural
products: wheat (53%), corn (54%) and soybean (43%).
With respect to the individual focus variables, given the
number of commodities, our analysis only permits reliable
statements related to price (30 studies), returns (85), and
volatilities (59). Again, for most commodities, the majority of
studies do not find a systematic speculative impact (classifi-
cation “zero” is the most frequent). Exceptions include: Re-
inforcing price effects for corn, soybean, live cattle, and lean
hogs; and reinforcing volatility effects for feeder cattle.
Weakening volatility effects for wheat, live cattle, and sugar.
And contradicting volatility effects for cocoa.
Overall, where studies do find effects of speculation,
they predominantly identify reinforcing price effects, but
indeterminate effects on returns and volatility (the figures
are displayed in the last column). This finding is consistent
with the conclusions from.
Table 1 and, in particular, from Table 3 where the re-
sults indicate more reinforcing than weakening price ef-
fects, but indeterminate effects on volatility.
5. Conclusions
In the past decade, non-commercial and non-hedging
motivated investments (“speculation”) in commodity futures
have been blamed by several organizations and public opinion
leaders for being responsible for increasing prices of agri-
cultural products, or commodity and food prices in general.
There are several instances in the history of economics where
speculation was made responsible for increase or declining
13
Another consequence of this is that since the following analysis is
based on a relatively low number of cases, the overall picture seems more
important to us than overly detailed results.
11
commodity prices, or price volatility unwarranted by funda-
mental factors. In the US, this debate emerged towards the
end of the 19th century after futures contracts on many
commodities have been established and “bucket shops” at-
tracted a lot of speculative capital of small “ill-educated” in-
vestors, which provoked established brokers, exchanges, and
agrarians to engage against these trading activities, which
were finally forbidden by rulings at the state and Federal
levels.14 Theoretical and empirical research about the benefit
or damaging effects of futures trading and speculation was not
available in these days. A similar debate took place at about
the same time in Germany, where speculation was blamed for
commodity price disruptions in the 1890s after grain trading
became increasingly globalized, prices declined and became
more volatile. The new Exchange Act (Börsengesetz) of 1896
prohibited the publication of spot and future prices, and grain
futures trading was banned outright. This was against the
recommendation of major economists in the Imperial Com-
mission (Max Weber, Arthur Cohn), who commented the
measures as the result of “ill-guided mass instincts”. Another
author complained that these regulatory measures are driven,
against all historical experience, by the pressure of public
opinion which has almost always been more powerful than all
scientific evidence (Fröchtling 1909). While his own empirical
analysis is merely based on price trends, averages and cycles,
other researchers used more advanced statistical tools. A
particularly interesting example is the study of Hooker (1901),
a then famous British statistician, who analyzed the volatility
effects of the suspension of futures trading at the Berlin ex-
change by estimating different volatility measures and corre-
lation coefficients (including their sampling error) to a large
international dataset of daily spot and futures prices of com-
modities. He concluded that the suspension has “not induced
a rise in the prices of grain, and they have not imparted
greater stability to those prices; if, indeed, they have not ex-
ercised a deleterious effect in both these directions.” In the US,
large-scale theoretical and empirical research on the behavior
of commodity futures prices and the economic role of hedging
and speculation started with the pioneering work of H.
Working at Stanford University's Food Research Institute in
the 1920s
Today, huge datasets are more readily available and
econometric methods made significant progress, so one is
tempted to expect the public debate to be more influenced
by academic findings. This presumption is not un-
warranted; the results from empirical research are indeed
part of the current debate about speculation. The survey
paper by Will et al. (2016) provoked much public con-
troversy, and several NGOs, political parties or interna-
tional organizations provide own detailed reviews and
appraisals of empirical studies.15
Our contribution in this paper is threefold: First, empirical
findings are separated for six distinct focus variables, on which
speculation is supposed to have reinforcing effects; second,
we discriminate between different measures of speculation
that are used the empirical work; and third: we address the
14
Hochfelder (2006) provides a discussion of this episode.
15
Examples include: WEED (World Economy, Ecology and Develop-
ment), JUSO Switzerland, or UNCTAD.
12 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15
impact of broad paper quality. Moreover, the focus of our
studies is not restricted to agricultural commodities and as
such, the overall sample is larger than in comparable studies.
The following conclusions can be drawn from our
metastudy:
– In the full sample of studies, speculation most likely has
a reinforcing influence on prices. Predominant weaken-
ing effects can most clearly be observed for returns and
risk premiums. For the other focus variables (spreads,
volatility and spill-over) the reinforcing and weakening
effects balance out.
– If indirect or proxy variables are used to measure spec-
ulation, about half of the respective studies report re-
inforcing effects, compared to a minority of studies
supporting weakening effects. If speculation is measured
directly the distribution shifts from reinforcing to
weakening effects. In general, the specification of the
speculation measure has a strong impact on the findings.
– Quality screening does not lead to fundamentally dif-
ferent conclusions about the effects of speculation. Al-
though the high quality papers tend to find more re-
inforcing effects on volatility, they do not find un-
ambiguous evidence for it. However, the analysis reveals
that the measurement of speculation has a substantial
impact upon the empirical results, even after controlling
for paper quality.
– For food commodities, if all papers are included, there is
no evidence for predominantly weakening or reinforcing
effects. However, the percentage of reinforcing impacts
slightly dominates for soybean, coffee, live cattle and
lean hogs.
– With respect to the individual focus variables for most
food commodities, the majority of studies do not find a
systematic speculative impact (classification “zero” is the
most frequent). Where studies do find speculative effects,
they predominantly identify reinforcing price effects, but
indeterminate effects on returns and volatility.
How are these findings related to those of other survey
papers? First, the three survey papers to which we refer in
the Introduction analyze a narrower range of commodities,
namely agricultural goods. Although the public spotlight in
the past years has been much on this particular segment,
we are convinced that the evidence for other commodities
on which futures contracts are traded is highly relevant
too when it comes to analyze the impact of speculation.
Second, other survey papers such as Will et al. (2016)
review the empirical evidence in a very general way, not
for specific focus variables or uniformity of impact, and
hence their conclusions are rather general.16 The other two
survey papers have their focus on price and volatility ef-
fects (Shutes and Meijerink (2012)), respectively, only vo-
latility effects (Brümmer et al. 2013). Our results show that
the effects of speculation are markedly different across
16
The most specific statement from their conclusions is that an al-
ternative view of speculation “is supported by the findings arrived at in
the vast majority of empirical studies, that financial speculation has
contributed to making the agricultural markets function not worse, but
better.”
focus variables, in particular, if the measurement of spec-
ulation is taken into account.
Third, we do not narrowly restrict or define the nat-
ure of speculation. Other papers, such as Shutes and
Meijerink (2012), analyze the specific role of index re-
lated investments. While it is questionable whether
these investments should be regarded as “speculation”
at all,17 our study is not focused on any specific form of
speculation: we take whatever the papers claim to mean
by speculation. However, we take a detailed look how
speculation is measured. We show that the measure-
ment of speculation has a substantial impact on the
documented effects: The number of reinforcing effects is
generally reduced. Interestingly, the results are fairly
different whether proxy variables or direct measures
(position data) are used, even after controlling for paper
quality.
Fourth, one might argue as done by Will et al. (2016)
that for quality reasons, only papers published in peer-
reviewed journals should be included in a survey. We
include, as the other two survey also do, a few un-
published papers or work not published in journals. At
the same time, we analyze the impact of broad paper
quality upon the findings. Paper quality is also a key
issue in the survey of Shutes and Meijerink (2012); we
basically agree with their overall conclusion that “not
all publications on the role of speculation on price le-
vels are of the same quality. The assessment of the lit-
erature suggests that a number of (often cited) pub-
lications contain quantitative analyses that can be
questioned. Care should be taken by interpreting these
conclusions.” However, we find that applying different
paper quality standards does not fundamentally change
the interpretation of the findings, with the exception of
volatility where the results tend to become more
indeterminate.
Our overall results are in line with those of the three
studies cited, but provide a more detailed picture. While
the overall picture indicates that the number of studies
which report reinforcing and weakening effects is about
the same, the results shift in favor of finding (clear or
mixed) weakening effects if the studies use direct mea-
sures of speculation, except for price.
Furhter Reading
(Butler, 2008; Cheng and Xiong, 2014; Acharya et al.,
2013; Adammer and Bohl, 2015; Ai et al., 2006; Algieri,
2012; Alquist and Kilian, 2010; Aulerich et al., 2010,, 2013;
Baffes and Haniotis, 2010; Basak and Pavlova, 2015; Beck-
mann and Czudaj, 2014; Bohl et al., 2013; Bohl and Ste-
phan, 2013; Bos and Van der Molen, 2012; Brunetti et al.,
2011; Brunetti and Reiffen, 2014; Bruno et al., 2015;
Büyüksahin et al., 2010; Büyüksahin and Harris, 2011;
Büyüksahin and Robe, 2014; Capelle-Blancard and Couli-
baly, 2011; Carter et al., 2011; Cashin et al., 2002; Chen
17
Several authors, such as Stoll and Whaley (2011), strongly doubt
that index investing is classical speculation.
 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15 13

Table 10 Table 12
Impact of speculation on focus variables: studies from journal rankings Impact of speculation based on speculation measures: studies from
SRJ Z1.0. journal rankings SRJ Z 1.0.

Focus variable Impact direction Total undet Speculation measure Impact direction Total undet

2 1 0 1 2 2 1 0 1 2

Price 14% 0% 71% 14% 0% 7 3 Unspecified 0% 0% 100% 0% 0% 2 7

Returns 0% 0% 33% 33% 33% 3 0 Indirect 25% 0% 75% 0% 0% 4 0
Risk Premiums 20% 20% 0% 40% 20% 5 0 Proxy 13% 33% 33% 7% 13% 15 0
Spreads 0% 33% 33% 33% 0% 3 0 Direct 15% 23% 15% 38% 8% 13 1
Volatility 0% 50% 33% 17% 0% 6 1
Spill-Over 30% 30% 30% 0% 10% 10 0
All 15% 24% 35% 18% 9% 34 8

All 15% 24% 35% 18% 9% 34 4 The variables used to measure speculation are classified into four cate-
gories; they are described in Section 2 of this paper. The impact direction
The focus variables are those economic magnitudes on which speculation indicates how speculation affects the focus variable. Thereby, 2 means a
is supposed to have a statistical impact; they are described in Section 2 of clear reinforcing impact, 1 a mixed reinforcing impact, 0 no impact,  1 a
this paper. The impact direction indicates how speculation affects the mixed weakening impact and  2 a clear weakening impact. The per-
focus variable. Thereby, 2 means a clear reinforcing impact, 1 a mixed centages refer to the Total number of observations displayed in the sec-
reinforcing impact, 0 no impact,  1 a mixed weakening impact and  2 a ond last column. “undet” in the last column refers to studies which
clear weakening impact. The percentages refer to the Total number of cannot be categorized to the 5 impact categories. Papers are selected
observations displayed in the second last column. “undet” in the last based on the Elsevier SCImago Journal Rank (SJR) above 1.0.
column refers to studies which cannot be categorized to the 5 impact
categories. Papers are selected based on the Elsevier SCImago Journal
Table 13
Rank (SJR) above 1.0.
Impact of speculation based on speculation measures: studies from
journal rankings IF Z1.0.
Table 11
Impact of speculation on focus variables: studies from journal rankings IF Speculation measure Impact direction Total undet
Z 1.0.
2 1 0 1 2
Focus variable Impact direction Total undet

2 1 0 1 2 Unspecified 0% 0% 100% 0% 0% 4 11
Indirect 25% 0% 75% 0% 0% 4 0
Proxy 11% 33% 28% 17% 11% 18 0
Price 11% 11% 67% 11% 0% 9 7 Direct 13% 20% 20% 33% 13% 15 4
Returns 0% 0% 40% 20% 40% 5 1
Risk Premiums 20% 20% 0% 40% 20% 5 0
Spreads 0% 25% 50% 25% 0% 4 1 All 12% 22% 37% 20% 10% 41 15
Volatility 0% 43% 29% 29% 0% 7 2
Spill-Over 27% 27% 27% 9% 9% 11 0 The variables used to measure speculation are classified into four cate-
gories; they are described in Section 2 of this paper. The impact direction
indicates how speculation affects the focus variable. Thereby, 2 means a
All 12% 22% 37% 20% 10% 41 11 clear reinforcing impact, 1 a mixed reinforcing impact, 0 no impact,  1 a
mixed weakening impact and  2 a clear weakening impact. The per-
The focus variables are those economic magnitudes on which speculation centages refer to the Total number of observations displayed in the sec-
is supposed to have a statistical impact; they are described in Section 2 of ond last column. “undet” in the last column refers to studies which
this paper. The impact direction indicates how speculation affects the cannot be categorized to the 5 impact categories. Papers are selected
focus variable. Thereby, 2 means a clear reinforcing impact, 1 a mixed based on the Thomson Reuters Impact Factor (IF) above 1.0.
reinforcing impact, 0 no impact,  1 a mixed weakening impact and  2 a
clear weakening impact. The percentages refer to the Total number of
observations displayed in the second last column. “undet” in the last et al., 2011; Kilian and Lee, 2014; Kilian and Murphy, 2014;
column refers to studies which cannot be categorized to the 5 impact Kilian and Park, 2009; Kim, 2015; Lagi et al., 2011; Lehecka,
categories. Papers are selected based on the Thomson Reuters Impact
Factor (IF) above 1.0.
2013a; Lehecka, 2013b; Liao-Etienne et al., 2012; Lombardi
and Van Robays, 2011; Manera et al., 2013; Mayer, 2012;
McPhail et al., 2012; Nazlioglu et al., 2013; Peterson and
et al., 2010; Cheng et al., 2014; Cifarelli and Paladino, 2011;
Tomek, 2005; Pfuderer and Gilbert, 2013; Phillips et al.,
Cornaggia, 2013; Creti et al., 2013; Dorfman and Karali,
2011; Phillips and Yu, 2011; Pirrong, 2011; Rezitis and
2012; Du et al., 2011; Ederington and Lee, 2002; Etula,
Sassi, 2013; Roache, 2010; Rouwenhorst and Tang, 2012;
2013; Fattouh, 2010; Fattouh and Mahadeva, 2012; Fishe
et al., 2014; Fishe and Smith, 2012; Gilbert, 2010a,, 2010b,, Sanders et al., 2004; Sanders and Irwin, 2011a; Sanders
2012; Gilbert and Morgan, 2010; Gilbert and Pfuderer, and Irwin, 2011b; Sanders et al., 2009; Sanders et al., 2010;
2014a,, 2014b; Grosche, 2014; Gutierrez, 2013; Gutierrez Silvennoinen and Thorp, 2013; Smith, 2009; Sockin and
et al., 2014; Hamilton, 2009; Hamilton and Wu, 2014; Xiong, 2013; Stoll and Whaley, 2010; Sumner, 2009; Tang
Hamilton and Wu, 2015; Harri et al., 2009; Headey and and Xiong, 2012; Tilton et al., 2011; Tse and Williams,
Fan, 2008; Hernandez and Torero, 2010; Irwin et al., 2011; 2013; UNCTAD, 2011; Wang, 2003; Windawi, 2012;
Irwin and Sanders, 2010,, 2011,, 2012a,, 2012b; Irwin et al., Wright, 2009,, 2011).
2009; Janzen et al., 2013; Ji and Fan, 2012; Kawamoto
14 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15
Appendix
See Appendix Tables 10–13.
References
Acharya, V.V., Lochstoer, L.A., Ramadorai, T., 2013. Limits to arbitrage and
hedging: evidence from commodity markets. J. Financ. Econ. 109 (2),
441–465.
Adammer, P., Bohl, M.T., 2015. Speculative bubbles in agricultural prices.
Q. Rev. Econ. Financ. 55, 67–76.
Ai, C., Chatrath, A., Song, F., 2006. On the comovement of commodity
prices. Am. J. Agric. Econ. 88 (3), 574–588.
Algieri, B., 2012. Price Volatility, Speculation and Excessive Speculation
in Commodity Markets: Sheep or Shepherd Behaviour? ZEF-Discus-
sion Papers on Development Policy, 166.
Alquist, R., Kilian, L., 2010. What do we learn from the price of crude oil
futures? J. Appl. Econ. 25 (4), 539–573.
Aulerich, N.M., Irwin, S.H., Garcia, P., 2010. The Price Impact of Index
Funds in Commodity Futures Markets: Evidence from the CFTCs Daily
Large Trader Reporting System, Working Paper, Department of Agri-
cultural and Consumer Economics, University of Illinois at Urbana-
Champaign.
Aulerich, N.M., Irwin, S.H., Garcia, P., 2013. Bubbles, food prices, and
speculation: evidence from the CFTC's daily large trader data files,
The Economics of Food Price Volatility. University of Chicago Press,
pp. 211–253.
Baffes, J., Haniotis, T., 2010. Placing the 2006/08 Commodity Price Boom
into Perspective, Policy Research Working Paper Series, The World
Bank Development Prospects Group.
Basak, S., Pavlova, A., 2016. A model of financialization of commodities. J.
Financ. 71 (4), 1511–1556.
Beckmann, J., Czudaj, R., 2014. Non-linearities in the relationship of
agricultural futures prices. Eur. Rev. Agric. Econ. 41 (1), 1–23.
Bohl, M.T., Stephan, P.M., 2013. Does futures speculation destabilize spot
prices? New evidence for commodity markets. J. Agric. Appl. Econ. 45
(04), 595–616.
Bohl, M.T., Javed, F., Stephan, P.M., 2013. Do Commodity Index Traders
Destabilize Agricultural Futures Prices? Appl. Econ. Q. (formerly:
Konjunkturpolitik) 59 (2), 125–148.
Bos, J. W., Van der Molen, M., 2012. A Bitter Brew? Futures Speculation
and Commodity Prices, Working Paper.
Brunetti, C., Reiffen, D., 2014. Commodity index trading and hedging
costs. J. Financ. Mark. 21, 153–180.
Brunetti, C., Büyüksahin, B., Harris, J.H., 2011. Speculators, Prices and
Market Volatility, Working Paper.
Bruno, V., Büyüksahin, B., Robe, M.A., 2015. The Finnancialization of
Food? Working Paper.
Brümmer, B., Korn, O., Schlüssler, K., Jaghadani, T.J., Saucedo, A., 2013.
Volatility in the After Crisis Period – A Literature Review of Recent
Empirical Research, Working Paper.
Butler, D., 2008. Free journal-ranking tool enters citation market. Nature
451 (7174), 6.
Büyüksahin, B., Harris, J.H., 2011. Do speculators drive crude oil futures
prices. Energy J. 32 (2), 167–202.
Büyüksahin, B., Robe, M.A., 2014. Speculators, commodities and cross-
market linkages. J. Int. Money Financ. 42, 38–70.
Büyüksahin, B., Haigh, M.S., Robe, M.A., 2010. Commodities and equities:
’A Market of One’? J. Altern. Invest. 12 (3), 76.
Capelle-Blancard, G., Coulibaly, D., 2011. Index trading and agricultural
commodity prices: a panel granger causality analysis. Int. Econ. 126,
51–71.
Carter, C.A., Rausser, G.C., Smith, A., 2011. Commodity booms and busts.
Annu. Rev. Resour. Econ. 3, 87–118.
Cashin, P., McDermott, C.J., Scott, A., 2002. Booms and slumps in world
commodity prices. J. Dev. Econ. 69 (1), 277–296.
Chen, Y., Rogoff, K., Rossi, B., 2010. Can exchange rates forecast com-
modity prices? Q. J. Econ. 125 (3), 1145–1194.
Cheng, I.H., Xiong, W., 2014. Financialization of commodity markets.
Annu. Rev. Financ. Econ. 6 (1). 419 - 41.
Cheng, H., Kirilenko, A., Xiong, W., 2014. Convective risk flows in com-
modity futures markets. Rev. Financ., 1–43.
Cifarelli, G., Paladino, G., 2011. Hedging vs. Speculative Pressures on
Commodity Futures Returns, Working Paper.
Cornaggia, J., 2013. Does risk management matter? Evidence from the US
agricultural industry. J. Financ. Econ. 109 (2), 419–440.
Creti, A., Joets, M., Mignon, V., 2013. On the links between stock and
commodity markets' volatility. Energy Econ. 37, 16–28.
Danthine, J.-P., 1978. Information, futures prices, and stabilizing spec-
ulation. J. Econ. Theory 17 (1), 79–98.
De Long, J.B., Shleifer, A., Summers, L.H., Waldmann, R.J., 1990. Positive
feedback investment strategies and destabilizing rational speculation.
J. Financ. 45 (2), 379–395.
Dorfman, J.H., Karali, B., 2012. Have Commodity Index Funds Increased
Price Linkages between Commodities, Working Paper.
Du, X., Cindy, L.Y., Hayes, D.J., 2011. Speculation and volatility spillover in
the crude oil and agricultural commodity markets: a Bayesian ana-
lysis. Energy Econ. 33 (3), 497–503.
Ederington, L., Lee, J.H., 2002. Who trades futures and how: evidence
from the heating oil futures market. J. Bus. 75, 353–373.
Etula, E., 2013. Broker-dealer risk appetite and commodity returns. J. Fi-
nanc. Econ. 11 (3), 486–521.
Fattouh, B., Mahadeva, L., 2012. Assessing the Financialization Hypothesis.
Oxford Institute for Energy Studies Oxford Working Paper, Oxford.
Fattouh, B., 2010. Oil Market Dynamics through the Lens of the 2002–
2009 Price Cycle. Oxford Institute for Energy Studies, Oxford.
Fishe, R.P., Smith, A.D., 2012. Identifying informed traders in futures
markets. J. Financ. Mark. 15 (3), 329–359.
Fishe, R.P., Janzen, J.P., Smith, A., 2014. Hedging and speculative trading in
agricultural futures markets. Am. J. Agric. Econ. 96 (2), 542–556.
Froot, K.A., Scharfstein, D.S., Stein, J.C., 1992. Herd on the street: In-
formational inefficiencies in a market with short-term speculation. J.
Financ. 47 (4), 1461–1484.
Fröchtling, A., 1909. Über den Einfluss des Getreideterminhandels auf die
Getreidepreise. Jahrbücher für Nationalökonomie und Statistik (J.
Econ. Stat.) 37 (5), 577–622.
Gilbert, C.L., Morgan, C.W., 2010. Food price volatility. Philos. Trans. R. Soc.
Lond. B: Biol. Sci. 365 (1554), 3023–3034.
Gilbert, C.L., Pfuderer, S., 2014a. The financialization of food commodity
markets. In: Handbook on Food: Demand, Supply, Sustainability and
Security, pp. 122–48.
Gilbert, C.L., Pfuderer, S., 2014b. The role of index trading in price for-
mation in the grains and oilseeds markets. J. Agric. Econ. 65 (2),
303–322.
Gilbert, C.L., 2010a. How to understand high food prices. J. Agric. Econ. 61
(2), 398–425.
Gilbert, C.L., 2010b. Speculative Influences on Commodity Futures Prices
2006–2008. UNCTAD Working Paper No. 197.
Gilbert, C.L, 2012. Speculative impacts on grains price volatility, in: 123rd
Seminar, February 23–24, 2012, Dublin, Ireland, Working Paper No.
122540, European Association of Agricultural Economists.
Grosche, S.C., 2014. What does granger causality prove? A critical ex-
amination of the interpretation of granger causality results on price
effects of index trading in agricultural commodity markets. J. Agric.
Econ. 65 (2), 279–302.
Grossman, S.J., 1977. The existence of futures markets, noisy rational
expectations and informational externalities. Rev. Econ. Stud. 44 (3),
431–449.
Gutierrez, L., Piras, F., Roggero, P.P., 2014. A global vector autoregression
model for the analysis of wheat export prices. Am. J. Agric. Econ.. (p.
aau103).
Gutierrez, L., 2013. Speculative bubbles in agricultural commodity mar-
kets. Eur. Rev. Agric. Econ. 40 (2), 217–238.
Hamilton, J.D., Wu, J.C., 2014. Risk premia in crude oil futures prices. J. Int.
Money Financ. 42, 9–37.
Hamilton, J.D., Wu, J.C., 2015. Effects of index-fund investing on com-
modity futures prices. Int. Econ. Rev. 56 (1), 187–205.
Hamilton, J.D., 2009. Causes and Consequences of the Oil Shock of 2007–
08. Working Paper, National Bureau of Economic Research.
Harri, A., Nalley, L., Hudson, D., 2009. The relationship between oil, ex-
change rates, and commodity prices. J. Agric. Appl. Econ. 41 (02),
501–510.
Hart, O.D., Kreps, D.M., 1986. Price destabilizing speculation. J. Political
Econ. 94 (5), 927–952.
Headey, D., Fan, S., 2008. Anatomy of a crisis: the causes and con-
sequences of surging food prices. Agric. Econ. 39, 375–391.
Hernandez, M.A., Torero, M., 2010. Examining the dynamic relation be-
tween spot and futures prices of agricultural commodities. FAO
Commodity Market Review, no. 2009/2010, pp. 47–86.
Hochfelder, D., 2006. Where the common people could speculate: the
ticker, bucket shops, and the origins of popular participation in fi-
nancial markets, 1880–1920. J. Am. Hist. 93 (2), 335–358.
Hooker, R.H., 1901. The suspension of the Berlin produce exchange and its
effect upon corn prices. J. R. Stat. Soc. 64 (4), 574–613.
Irwin, S.H., Sanders, D.R., 2010. The Impact of Index and Swap Funds in
 M. Haase et al. / Journal of Commodity Markets 3 (2016) 1–15
Commodity Futures Markets. A Technical Report Prepared for the
Organization on Economic Co-Operation and Development (OECD).
Irwin, S.H., Sanders, D.R., 2011. Index funds, financialization, and com-
modity futures markets. Appl. Econ. Perspect. Policy 33 (1), 1–31.
Irwin, S.H., Sanders, D.R., 2012a. Financialization and structural change in
commodity futures markets. J. Agric. Appl. Econ. 44 (03), 371–396.
Irwin, S.H., Sanders, D.R., 2012b. Testing the masters hypothesis in com-
modity futures markets. Energy Econ. 34 (1), 256–269.
Irwin, S.H., Sanders, D.R., Merrin, R.P., 2009. Devil or angel? The role of
speculation in the recent commodity price boom (and bust). J. Agric.
Appl. Econ. 41, 393–402.
Irwin, S.H., Garcia, P., Good, D.L., Kunda, E.L., 2011. Spreads and non-
convergence in chicago board of trade corn, soybean, and wheat fu-
tures: are index funds to blame? Appl. Econ. Perspect. Policy 33,
116–142.
Janzen, J.P., Smith, A.D., Carter, C.A., 2013. Commodity Price Comovement
and Financial Speculation: The Case of Cotton, Working Paper, Uni-
versity of California.
Ji, Q., Fan, Y., 2012. How does oil price volatility affect non-energy com-
modity markets? Appl. Energy 89 (1), 273–280.
Kawamoto, T., Kimura, T., Morishita, K., Higashi, M., 2011. What Has
Caused the Surge in Global Commodity Prices and Strengthened
Cross-Market Linkage? Bank of Japan Working Paper, no. 11-E-3, May.
Kilian, L., Lee, T.K., 2014. Quantifying the speculative component in the
real price of oil: the role of global oil inventories. J. Int. Money Financ.
42, 71–87.
Kilian, L., Murphy, D.P., 2014. The role of inventories and speculative
trading in the global market for crude oil. J. Appl. Econ. 29 (3),
454–478.
Kilian, L., Park, C., 2009. The impact of oil price shocks on the US stock
market. Int. Econ. Rev. 50 (4), 1267–1287.
Kim, A., 2015. Does futures speculation destabilize commodity markets? J.
Futures Mark. 35, 696–714.
Kyle, A.S., Xiong, W., 2001. Contagion as a wealth effect. J. Financ. 56 (4),
1401–1440.
Lagi, M., Bar-Yam, Y., Bertrand, K.Z., Bar-Yam, Y., 2011. The Food Crises: A
Quantitative Model of Food Prices Including Speculators and Ethanol
Conversion, Working Paper.
Lehecka, G.V., 2013a. Do hedging and speculative pressures drive com-
modity prices, or the other way round? Empir. Econ. 49 (2), 575–603.
Lehecka, G.V., 2013b. Hedging and speculative pressures: an investigation
of the relationships among trading positions and prices in com-
modity futures markets, in: Proceedings of the NCCC-134 Conference
on Applied Commodity Price Analysis, Forecasting, and Market Risk
Management. St. Louis, MO (Working Paper).
Liao-Etienne, X., Irwin, S.H., Garcia, P., 2012. Price Explosiveness and In-
dex Trader Behaviour in the Corn, Soybean, and Wheat Futures
Markets, Working Paper.
Lombardi, M.J., Van Robays, I., 2011. Do Financial Investors Destabilize the
Oil Price? Working Paper.
Manera, M., Nicolini, M., Vignati, I., 2013. Futures Price Volatility in
Commodities Markets: The Role of Short Term vs Long Term Spec-
ulation, Working Paper.
Mayer, J., 2012. The growing financialisation of commodity markets: di-
vergences between index investors and money managers. J. Dev.
Stud. 48 (6), 751–767.
McPhail, L.L., Du, X., Muhammad, A., 2012. Disentangling corn price vo-
latility: the role of global demand, speculation, and energy. J. Agric.
Appl. Econ. 44 (03), 401–410.
Nazlioglu, S., Erdem, C., Soytas, U., 2013. Volatility spillover between oil
and agricultural commodity markets. Energy Econ. 36, 658–665.
Peterson, H.H., Tomek, W.G., 2005. How much of commodity price be-
havior can a rational expectations storage model explain? Agric. Econ.
33 (3), 289–303.
Pfuderer, S., Gilbert, C.L., 2013. Index Funds do Impact Agricultural Prices,
Working Paper.
Phillips, P.C., Yu, J., 2011. Dating the timeline of financial bubbles during
15
the subprime crisis. Quant. Econ. 2 (3), 455–491.
Phillips, P.C., Wu, Y., Yu, J., 2011. Explosive behaviour in the 1990s Nasdaq
when did exuberance escalate asset values? Int. Econ. Rev. 52 (1),
201–226.
Pirrong, C., 2011. Commodity Price Dynamics: A Structural Approach.
Cambridge University Press, New York, NY.
Rezitis, A.N., Sassi, M., 2013. Commodity food prices: review and empirics.
Econ. Res. Int., 1–15.
Roache, S.K., 2010. What Explains the Rise in Food Price Volatility? IMF
Working Papers, pp. 1–29.
Rouwenhorst, K.G., Tang, K., 2012. Commodity Investing. Annu. Rev. Fi-
nanc. Econ. 4 (1), 447–467.
Sanders, D.R., Irwin, S.H., 2011a. The impact of index funds in commodity
futures markets: a systems approach. J. Altern. Invest. 14 (1), 40–49.
Sanders, D.R., Irwin, S.H., 2011b. New evidence on the impact of index
funds in US grain futures markets. Can. J. Agric. Econ. (Rev. Can. d′
agroeconomie) 59 (4), 519–532.
Sanders, D.R., Boris, K., Manfredo, M., 2004. Hedgers, funds, and small
speculators in the energy futures markets: an analysis of the CFTC’s
commitments of traders reports. Energy Econ. 26 (3), 425–445.
Sanders, D.R., Irwin, S.H., Merrin, R.P., 2009. Smart money: the forecasting
ability of CFTC large traders in agricultural futures markets. J. Agric.
Resour. Econ. 34, 276–296.
Sanders, D.R., Irwin, S.H., Merrin, R.P., 2010. The adequacy of speculation
in agricultural futures markets: too much of a good thing? Appl. Econ.
Perspect. Policy 32 (1), 77–94.
Shutes, K., Meijerink, G.W., 2012. Food Prices and Agricultural Futures
Markets: A Literature Review, Working Paper.
Silvennoinen, A., Thorp, S., 2013. Financialization, crisis and commodity
correlation dynamics. J. Int. Financ. Mark. Inst. Money 24, 42–65.
Smith, J.L., 2009. World oil: market of mayhem? J. Econ. Perspect. 23,
145–164.
Sockin, M., Xiong, W., 2013. Feedback Effects of Commodity Futures Pri-
ces, Working Paper.
Stoll, H.R., Whaley, R.E., 2010. Commodity index investing and com-
modity futures prices. J. Appl. Financ. 20, 7–46.
Stoll, H.R., Whaley, R.E., 2011. Commodity index investing: speculation or
diversification? J. Altern. Invest. 14 (1), 50–60.
Sumner, D.A., 2009. Recent commodity price movements in historical
perspective. Am. J. Agric. Econ. 91 (5), 1250–1256.
Tang, K., Xiong, W., 2012. Index investment and the financialization of
commodities. Financ. Anal. J. 68 (5), 54–74.
Tilton, J.E., Humphreys, D., Radetzki, M., 2011. Investor demand and spot
commodity prices. Resour. Policy 36 (3), 187–195.
Tse, Y., Williams, M.R., 2013. Does index speculation impact commodity
prices? An intraday analysis. Financ. Rev. 48 (3), 365–383.
UNCTAD, 2011. Price Formation in Financialized Commodity Markets: The
Role of Information, United Nations Conference on Trade and De-
velopment. United Nations Publication, Geneva, Switzerland June
2011.
US Senate, Excessive speculation in the wheat market, Permanent Sub-
committee on Investigations, Committee on Homeland Security and
Governmental Affairs, Staff Report, June 24, 2009, Washington, D.C.
Wang, C., 2003. The behavior and performance of major types of futures
traders. J. Futures Mark. 23 (1), 1–31.
Will, M.G., Prehn, S., Pies, I., Glauben, T., 2016. Is financial speculation
with agricultural commodities harmful or helpful? A literature re-
view of current empirical research. J. Altern. Invest. 18 (3), 84–102.
Windawi, A.J., 2012. Speculation, Embedding and Food Prices: A Coin-
tegration Analysis, Working Paper.
Wright, B., 2009. International Grain Reserves: and other Instruments to
Address Volatility in Grain Markets. Policy Research Working Paper-
World Bank, 5028.
Wright, B.D., 2011. The economics of grain price volatility. Appl. Econ.
Perspect. Policy 33 (1), 32–58.