Any information reflects only the author(s) view and not that from the European Union. Literature review on price volati
Literature review on price volatility transmission in food supply chains, the role of contextual factors, and the CAP’s market measures Tsion Taye Assefa1, Miranda P.M. Meuwissen1, Alfons G.J.M. Oude Lansink1 Business Economics, Wageningen University
Working Paper No.4 ULYSSES “Understanding and coping with food markets voLatilitY towards more Stable World and EU food SystEmS” July, 2013 Seventh Framework Program Project 312182 KBBE.2012.1.4-05 www.fp7-ulysses.eu
ULYSSES project has received research funding from the European Commission Project 312182 KBBE.2012.1.4-05 Any information reflects only the author(s) view and not that from the European Union
ULYSSES project assess the literature on prices volatility of food, feed and non-food commodities. It attempts to determine the causes of markets' volatility, identifying the drivers and factors causing markets volatility. Projections for supply shocks, demand changes and climate change impacts on agricultural production are performed to assess the likelihood of more volatile markets. ULYSSES is concerned also about the impact of markets' volatility in the food supply chain in the EU and in developing countries, analysing traditional and new instruments to manage price risks. It also evaluates impacts on households in the EU and developing countries. Results will help the consortium draw policy-relevant conclusions that help the EU define market management strategies within the CAP after 2013 and inform EU’s standing in the international context. The project is led by Universidad Politécnica de Madrid. Internet: http://www.fp7-ulysses.eu/ Authors of this report and contact details
Name: Tsion Taye Assefa, Miranda P.M. Meuwissen, Alfons G.J.M. Oude Lansink Partner acronym: WU Address: Hollandseweg 1, 6706 KN WAGENINGEN, The Netherlands E-mail:
[email protected],
[email protected],
[email protected]
When citing this ULYSSES report, please do so as: Assefa,T.T., Meuwissen, M.P.M. and Oude Lansink, A.G.J.M. 2013. Literature review on price volatility transmission in food supply chains, the role of contextual factors, and the CAP’s market measures, Working Paper No. 4, ULYSSES project, EU 7th Framework Programme, Project 312182 KBBE.2012.1.4-05, http://www.fp7-ulysses.eu/ , 30 pp.
Disclaimer: “This publication has been funded under the ULYSSES project, EU 7th Framework Programme, Project 312182 KBBE.2012.1.4-05. Any information reflects only the author(s) view and not that from the European Union.” "The information in this document is provided as is and no guarantee or warranty is given that the information is fit for any particular purpose. The user thereof uses the information at its sole risk and liability."
ii
Table of Contents List of acronyms ...................................................................................................................... iii Executive summary .................................................................................................................. iv 1.
Introduction ..................................................................................................................... 1
1.
Price transmission versus price volatility transmission........................................................... 2
2.
Review of studies on price volatility transmission in food supply chains .................................. 4
3.
Factors with effects on the degree of price volatility transmission in food supply chains ............ 9
4.
Market measures of the CAP .............................................................................................13
5.
Conclusions ....................................................................................................................18
References .............................................................................................................................20
List of acronyms AR
Auto Regressive
ARMA
Auto Regressive Moving Average
BEKK
Baba-Engle-Kraft-Kroner
CAP
Common Agricultural policy
DCC
Dynamic Conditional Correlation
DVECH
Diagonal VECH model
EGARCH
Exponential Generalized Autoregressive Condition Heteroskedasticity
EU
European Union
GATT
General Agreement on Tariff and Trade
GJR-GARCH Glosten-Jagannathan-Runkle Generalized Autoregressive Condition Heteroskedasticity MGARCH
Multivariate Generalized Autoregressive Condition Heteroskedasticity
M-TAR
Momentum Threshold Auto Regressive
STCC-GARCH Smooth transition conditional correlation Generalized Autoregressive Condition Heteroskedasticity TVECM
Threshold Vector Error Correction Model
VAR
Vector Auto Regressive iii
VECM
Vector Error Correction Model
WTO
World Trade Organization
Executive summary The last decade, and particularly since the 2007/08 food crisis, food price volatility in world markets has seen an increasing trend. The successive reforms of the Common Agricultural Policy (CAP), which made EU agriculture more market oriented, led to the exposure of EU farmers and consumers to world price uncertainties. Given this background, investigating the extent to which the various actors of food supply chains are exposed to price volatilities is a timely matter. This deliverable has the general objective of reviewing studies that empirically investigated the level of price volatility transmission in food supply chains. The scope of this review is a comprehensive one in the sense that studies on all countries and types of commodities covered in the literature are addressed. The literature on energy-agricultural markets price volatility linkages are also reviewed given the increasing linkage of these two markets in the past decade and also because the stream of this literature is necessary to have a holistic understanding of the sources of price volatilities in food supply chains. This deliverable has three specific objectives: 1) to review the methodological applications and major findings of existing studies on price volatility transmission in food chains, and in the energyagricultural commodity chains, 2) to identify and discuss the factors with effects on the transmission of price volatilities in food chains, and 3) to review studies that assess the CAP’s price stabilization policies. Since the focus of this deliverable is the food chain (from farm to retail), examination of factors with effects on the degree of price volatility transmission is limited only to this chain. The inclusion of the last objective is justified given that the CAP stabilization policies, which focus on the farm sector, have their own bearing on the volatility that is transmitted from the farm stage to the rest of the food supply chain. One would expect stability achieved at the farm stage to reduce the exposure of the downstream sector to price uncertainties coming from the upstream sector. The review revealed that the number of studies on price volatility transmission in food supply chains (farm-retail) is very limited in terms of country coverage where the EU market is poorly covered, while more coverage is given to US markets. Product coverage focuses on the meat market with practically no studies for vegetable and cereal markets. Period coverage is also limited in that the last decade, and especially since 2006, is inadequately covered. In terms of methodologies used to investigate price volatility transmission, the application of multivariate general autoregressive conditional heteroskedasticity (MGARCH) models are common modelling approaches in studies for food supply chains (from farm to consumer) and for energy-agricultural commodity chains. The BEKK form of MGARCH models is in particular used in most of the studies. The major finding of the studies on farm-retail price volatility transmission is that price volatility does transmit across food supply chains. They further show that price volatility sourced from the downstream sector is as important as the volatility sourced from the upstream sector. The literature on energy-agricultural commodity price volatility transmission shows that energy price volatilities transmit to commodities used for the production of biofuel. These latter studies do not however extend the scope of their investigation to the rest of the food chain, and only study the impacts of the energy price volatilities on agricultural commodity prices. iv
The price volatility transmission literature for food supply chains does not attempt to empirically investigate the role of factors that potentially affect the transmission of price volatility. Nevertheless, the authors suggest that the degree of market power (particularly in the retail sector) is responsible for the low transmission of upstream price volatilities to the retail sector. The absence of contracts at the farm sector was also attributed as contributing to the transmission of downstream price volatilities to the farm sector. Factors such as the biological nature of agricultural production (time lag in production response) and lower price elasticity of farm-level demand than that of retail demand were attributed to the lower level of consumer price volatility compared to farm price volatility. The price transmission literature (in levels) was also investigated to see other factors that can affect the transmission of prices along the chain. It is hypothesised that the same factors that affect the transmission of prices in food supply chains can affect the transmission of price volatility as both deal with price linkages along different markets. The price transmission literature deals with the transmission of predictable price changes in food supply chains; on the other hand, the price volatility transmission deals with the transmission of unpredictable price changes or price uncertainties. Among other factors, the price transmission literature commonly mentions retail market power and retail adjustment costs (such as repricing costs) as major reasons for low or/and asymmetric transmission of prices. Future research can investigate if the same factors that affect the level of price transmission also affect the level of price volatility transmission in food supply chains. The review of studies that investigated the implementation of the CAP’s price stabilization measures reveals that the EU agriculture is increasingly exposed to world market price signals resulting from CAP reforms that made EU farmers market oriented and from WTO commitments to open borders and restricts export subsidies. These measures have traditionally kept EU agricultural prices of the main commodities at high and stable levels. The further reduction of border protection and the prospective removal of export subsidies are expected to further reduce EU prices. The reduction of intervention prices and the abolishment of production quotas might make imports uncompetitive in the EU because they lower EU prices and thus help avoid the further downward pressure on EU prices brought by imports. However, such effect greatly depends on world market prices and exchange rate conditions. As a result, EU farm gate prices might be further exposed to world price signals and not be as stable as in the past decade. All this implies that the downstream sector can also be increasingly exposed to price volatilities because farm prices volatilities can transmit to the rest of the chain. Overall the literature review illustrates the relevance of price volatility along food supply chains and the existing gap in the literature. The limited country, product, and sample period coverage of the current literature can be one avenue of future research. Empirical investigation of the role of factors with possible effects on the degree of price volatility transmission in food supply chains, as well as on the transmission of energy price volatilities in food chains, is another avenue of future research. The effects of current CAP market measures on the level of volatility transmitted in EU food chains is also a currently unexplored area and can be addressed in future research.
v
1. Introduction World agricultural prices have experienced an increasing degree of volatility in the last decade (FAO et al., 2011). In the EU, the recent rise of volatility in international agricultural commodity prices has intensified the debate about risk related agricultural policies. Such debate was already underway in response to increasing volatility of agricultural prices as a result of successive reforms of the CAP, since early 1990s that have led to the exposure of EU domestic prices to international price signals (Tangermann, 2011). Market measures under the CAP such as border protection through import barriers, export subsidies, production quotas and intervention buying have traditionally kept EU agricultural prices at high and stable levels (Bardají et al., 2011; Tangermann, 2011). The mounting pressure of such market interventions on the EU budget and the increasing tensions with the EU’s trading partners have led to the CAP reforms in 1992 and 2003 which gradually shifted market interventions to decoupled single farm payments and opened up the EU to international markets (Sckokai and Moro, 2009; Tangermann, 2011). Such reforms have led to the spillover of the recent international agricultural price volatility to the EU market, as evidenced in several empirical studies (Hernandez et al., 2011; Piot-Lepetit, 2011; Bardarjí et al., 2011; Rabobank, 2011). Price volatility, characterized by unexpected price changes, entails risk to farmers who may react by reducing output supply and investments in productive inputs (Seal and Shonkwiler, 1987; Rezitis and Stavropoulos, 2009; Sckokai and Moro, 2009; Piot-Lepetit, 2011; Tangermann, 2011; and Taya, 2012). The downstream sector of food supply chains is additionally subject to sourcing uncertainties arising from unexpected price fluctuations in agricultural production inputs (Rabobank, 2011). All this implies that price volatility is a supply chain matter and its management requires understanding its levels and transmission across the whole food chain. Exploring the magnitudes and direction of price volatility transmission across the chain enables to direct price stabilization policies to the sectors that exhibit the highest level of price volatility. As stated by Buguk et al. (2003), policy changes in primary input markets that alter price volatility will have impacts on price volatility through the vertical chain. And where it is found that volatility is not being transmitted across chain stages, one cannot expect that stabilizing one market will lead to stability in other related markets (Serra, 2011a). This suggests that different types of intervention are required at different levels of the supply chain (Serra, 2011a). The transmission of volatility from one stage to another also increases the accuracy of forecasts made by agents about prices in other markets (Apergis and Regitis, 2003; Buguk et al., 2003). This in turn affects the hedging decisions of chain actors. According to Buguk et al. (2003), volatility spillover could introduce cross-hedge relationships across chain stages. For instance, if price volatility spills from a chain stage where futures market exist to the next stage where such markets do not exist, then actors in the latter market stage can use the futures market in the former stage to hedge against price volatilities in their own market. Agricultural economists have traditionally been more interested in the chain-wide transmission of prices in levels than in the transmission of price volatilities. One may assert that price transmission in levels also implies the transmission of price volatilities since both deal with price linkages. But empirical evidence (for instance Rezitis and Stavropoulos (2011) for the Greek broiler chain) indicates that this is not necessarily the case. This deliverable serves as evidence of the limited coverage of price volatility linkages in food supply chains, and draws the attention of researchers with interests in the area. The coverage of the EU market is even more limited, suggesting the need to include studies made for other countries in order to make the review a comprehensive and more informative one. In 1
addition to studies that cover the food supply chain (from farm to retail), this deliverable also reviews studies that deal with energy-agricultural commodity price volatility linkages. This stream of literature is included given the increased price linkage between the energy and agricultural sector since the mid 2000s as a result of the rise in biofuel production (Serra and Gil, 2012). This increased linkage between energy and agricultural commodity price volatilities is believed to have an implication on the volatility that is transmitted along food supply chains. Lastly, the review includes studies that assessed the price stabilizing potentials of the CAP market measures within the EU. This is justified given that these market measures, which focus on the farm sector (Van Meijl and Van Tongeren, 2003), have their own bearing on the volatility that is transmitted from the farm stage to the rest of the food supply chain. One would expect stability achieved at the farm stage to reduce the exposure of the downstream sector to price uncertainties coming from the upstream sector. The general objective of this deliverable is to review studies on price volatility transmission in food supply chains. More specifically, the objectives are 1) to review the methodological applications and major findings of existing studies on price volatility transmission in food chains, and in the energyagricultural commodity chains, 2) to identify and discuss the factors with effects on the transmission of price volatilities in food chains, and 3) to review studies that assess the CAP’s price stabilization policies1. Section 2 presents a definition of price volatility transmission and its difference with price transmission (in levels), Section 3 reviews current studies on price volatility transmission, Section 4 identifies and discusses the major factors with effect on the degree of volatility transmission in food supply chains, Section 5 reviews the studies that assess the CAP approach in its agricultural price stabilization policies and Section 6 concludes this deliverable.
1. Price transmission versus price volatility transmission Definition Both price and price volatility transmission are similar in that they both deal with price linkages along the chain. Their difference lies in the fact that price transmission refers to the linkages between the conditional mean prices while price volatility transmission refers to the linkages between the conditional variance of prices (Natcher and Weaver, 1999). In a less technical term, price transmission deals with the vertical relationship between the predictable ‘portions’ of prices whereas price volatility transmission deals with the vertical relationship between the unpredictable ‘portions’ of prices. Price volatility transmission can also be defined as the degree to which price uncertainty in one market affects price uncertainty in the others (Apergis and Rezitis, 2003). If prices (and price volatilities) are fully and instantaneously transmitted along the chain, one would expect a near to unity correlation between prices (and price volatilities) at different market levels (Serra, 2011a). It can be the case that prices (in levels) are perfectly transmitted whereas price volatility transmission is not perfect.
Even though the original title of the deliverable is ‘Price transmission, food chain organisation, CAP's approach in the EU food supply chains’, aspects that characterize the ‘food chain organisation’ of the EU are not included in this deliverable. These aspects, such as contractual relations in the chain and the degree of vertical integration will be included in more detail in the second deliverable on ‘Alternative management and strategic responses to deal with agricultural markets instability: going beyond futures and options contracts in EU agri-food sector’. 1
2
Methodological differences in empirical studies of price transmission and price volatility transmission Price transmission models (in levels) can take several specifications depending on their intended use. The most commonly used econometric model applied in recent papers is the vector error correction model (see for instance Lloyd et al., 2006; Falkowski, 2010; Kuiper and Oude Lansink, 2013) where the error correction mechanism was first suggested by Von Cramon-Taubadel (1997). Variations of this model can be used to test for asymmetries in price transmission (see for survey of asymmetric price transmission models in Meyer and Von Cramon-Taubadel (2004) and Frey and Manera (2007)), and to test for threshold error correction behaviour. For applications of threshold vector error correction model see for instance Goodwin and Holt (1999) and Rezitis and Stavropoulos (2011), and Brummer et al. (2009) for application of the Markov switching vector error correction model. To help identify the main methodological differences between price transmission and price volatility transmission models, a vector error correction model (VECM) is specified as follows for a chain consisting of two stages: ∑ ∑
∑ ∑
(
) (
(1) (2)
)
Equation (1) and (2) , which form the VECM, are solved simultaneously to determine the degree of price transmission between the two stages. In these equations, is the retail price at time t and is the farm price at time t. The coefficients show the degree of transmission of current and lagged short term farm price changes to the retail prices while the coefficients show the degree of transmission of current and lagged short-term changes in the retail price to the farm price. The degree to which retail prices and farm prices adjust to the long-term equilibrium relation are given by coefficients and respectively. This vector error correction model shows how the predictable portions of changes in farm and retail prices transmit to each other. The error terms and in equation (1) and (2) are the basis for estimating the commonly used price volatility transmission models called Multivariate Generalized Autoregressive Heteroskedasticity models (MGARCH). A family of MGARCH models exist, which similarly with price transmission models, can have different specifications depending on the intended use and estimation efficiencies. For the purpose of illustration of the difference of price transmission and price volatility transmission models, a specification of an MGARCH model called the BEKK (Baba, Engle, Kraft, Kroner, 1991) model is given below. The full BEKK model takes the following form (3) In this equation, and for two price series for instance (i.e., farm and retail prices), is a 2 x 2 covariance matrix, is a lower triangular matrix of constants, is a 2 x 2 matrix of ARCH term coefficients and is a 2 x 2 matrix of GARCH term coefficients. To clearly see the matrix elements, the equation can be written as: [ [
] ][
[
][ ][
]
][
[ ]
][
] (4)
3
After including the mean price transmission equations (assuming a VECM specification) and after doing the matrix multiplication of the BEKK model, the full price volatility transmission model leads to the simultaneous estimation of the following system of five equations: ∑ ∑
∑ ∑
(
) (
( (
)
)
)
The transmission of recent shocks from farm to retail is captured by the coefficient and the transmission of past wholesale price volatility is captured by . Similarly, the transmission of recent retail price shocks to the farm stage is captured by the coefficient and the transmission of past retail price volatility is captured by . While the coefficients are called the ARCH terms (Autoregressive Conditional Heteroskedasticity), the coefficients are called the GARCH terms. Since one-period lag is used for the variables associated with these coefficients, this BEKK model is an MGARCH (1,1) model. The time varying covariance between the two prices is given by . It can be seen that the price volatility transmission model uses the error terms and estimated in the price transmission model (VECM), and sees the relationship between the two error terms. These error terms indicate the price changes that cannot be predicted by the price transmission model. In other words, the error terms measure the degree of volatility in prices. The price volatility transmission model regresses the error terms from one stage on the error terms on the others stages and tests if recent shocks and past volatilities of one stage affects those of other stages.
2. Review of studies on price volatility transmission in food supply chains This section reviews studies on the transmission of volatility in food supply chains (from farm to retail) and studies on the transmission of price volatility between energy and agricultural commodity markets. The energy-agricultural commodity price volatility linkage results from the growth in ethanol production as alternative source of energy and from the use of agricultural commodities as inputs to its production (Trujillo-Barrera et al., 2012, Gardebroek and Hernandez, 2012). This interrelationship implies that energy price volatility transmits in the food chain and translates into increased farm input price volatility (inputs such as animal feed, fertilizers and transportation costs) which in turn can lead to farm output price volatility. Even though the current literature on the energy-agricultural commodity price volatility linkage focuses on feedstock commodities (soya, maize, wheat, rapeseed), this literature is also of interest from a food supply chain perspective. This is because energyagricultural commodity price volatility linkages can translate into risk and uncertainty for the rest of the food chain since biofuel feedstocks are not only used to produce energy, but also as food products such as meat or flour (Serra and Gil, 2012). 4
Price volatility transmission in food chain (farm-retail) An overview of the studies on price volatility linkages along the food chain – from farm to retail - is provided in Table 1 below. Table 1- Empirical studies on price volatility transmission along the food supply chain Author (date)
Khan and Helmers (1997)
Country (Product) US (Feed (corn), pork, beef and poultry)
Sample Period1
Mean equations specification
Specification of variance equations
19701981
---
VAR on moving variances of prices
US (Beef)
19701998
VECM
VAR on univariate GARCH conditional variances
Buguk et al. (2003)
US (Catfish)
19802000
AR
Univariate EGARCH with spillover effects
Apergis and Rezitis (2003)
Greece (Agricultural products)
19851999
VECM
Variant of VECH model
VECM
A Variation of BEKK and VECH
VECM
Cholesky2 decomposition
AR
Univariate EGARCH
VAR
Cholesky2 decomposition
Natcher and Weaver (1999)
Rezitis (2010) Chavas and Mehta (2004) Zheng et al. (2008) Mehta and Chavas (2008) Uchezuba et al. (2010) Serra (2011)
Greece (Lamb, beef, pork and poultry)
US (Butter)
US (45 retail food items)
US (Coffee)
South Africa (Broiler) Spain (Beef)
Alexandri (2011)
Romania (Agricultural price indices)
Rezitis and Stavropou
Greece (Broiler)
19882000
19802001
19802004
19752002
20002008 19962005 20062010 19932009
M-TAR VECM
---
TVECM
Univariate EGARCH with spillover effects Univariate STCC-GARCH Comparison of coefficient of variation of prices DVECH and BEKK model
5
Transmissi on of volatility detected
Direction of detected volatility transmission
Yes
From feed to farm
Yes
Bidirectional across all chain stages (farmwholesale-retail)
Yes
From feed to farm; From wholesale to farm; From farm to wholesale
Yes
From feed to farm; From consumer to farm
Yes
From farm to consumer; From consumer to farm
Yes
N/A3
Yes
N/A
Yes
N/A
Yes
From farm to retail
Mixed4
N/A
Yes
N/A
No
----
los (2011)
Rezitis (2012)
Greece (Beef, lamb, pork, poultry)
Khiyavi et al. (2012)
Iran (Poultry)
19932008 19972010
AR and ARMA
Diagonal VECH
Yes
N/A2
VECM
Variant of VECH
Yes
From feed to farm; From retail to farm
1
all data frequencies are monthly. based on effect of prices on covariance of price volatilities. 3 not applicable. 4 based on covariance (correlation) of price volatilities. 2
Table 1 provides a clear indication of the limited coverage of European food supply chains in the price volatility transmission litterature. The case of Greece is an exception, and it takes up four out of the six studies on the European market. Overall, the studies are dominated by the US market. Product coverage is also limited, as the focus is on the meat market (beef, pork, lamb and poultry). It is also striking that the vegetable and cereal sectors are not covered in the literature. Even more striking is the sample periods that the studies cover. The extreme level of volatilities observed in world and EU markets particularly since 2006 did not seem to have initiated increased interest in studies related to price volatility linkages along the chain. Many of the studies included in Table 1 end their sample periods as of 2005, including studies that are undertaken as early as 2008 (Zheng et al., 2008; Mehta and Chavas, 2008), and 2011 (Serra, 2011a). Only the last four studies listed in Table 1 cover the period from 2006 onwards. A commonality can be observed in the methodological approaches used in the existing literature. The focus is on a class of univariate and multivariate GARCH models which are a combination of mean and variance modelling of the prices under consideration. The mean part is commonly specified in the VECM (or AR or VAR) form and the variance part in a GARCH form. This type of volatility modelling gives by itself an insight on how the authors define volatility. Almost all authors model price volatility as a price component that remains unexplained in the mean part of the whole model. None of the studies use lags higher than one period in specifying the GARCH component with GARCH(1,1) being the chosen specification. A closer look at the trend in the methods adopted reveals that multivariate GARCH models (such as VECH and BEKK) are more used in recent periods than in earlier periods, whereas univariate GARCH models with adjustment for spillover effects are more adopted in the earlier periods. An example of univariate GARCH application to study spillover effects is the paper by Buguk et al. (2003). They estimate a univariate EGARCH (that also allows for asymmetries) for this market and the most recent innovations from other markets are included as exogenous variables in the conditional variance equation of the wholesale market. Moreover, VAR models (which are normally used to specify predictable price linkages) to specify volatility linkages seem to be used in earlier studies (Khan and Helmers, 1997; Natcher and Weaver, 1999) rather than in recent studies. Sample sizes as selected by the studies listed in Table 1 range from 10-20 years with all data frequencies being monthly. Not every study listed in Table 1 directly studies the transmission of volatility from one chain stage to another. For instance, Chavas and Mehta (2004) and Mehta and Chavas (2008) investigate the effect of changes in prices (in levels) on the covariance of price volatilities in two chain stages; Serra (2011a) and (Rezitis 2012) use the covariance of volatilities in two chain stages to infer about spillover effects; Alexandri (2011) compares volatilities at the farm and retail stages; and Zheng et al. (2008) 6
estimate only a univariate EGARCH from which we can infer how retail price volatilities respond to negative price shocks (possibly coming from upstream price shocks). The rest of the studies listed in Table 1 explicitly study the transmission of price volatility from one stage to another. Except for the paper by Rezitis and Stavropoulos (2011), all reviewed papers indicate that price volatility does transmit across chain stages. The farm sector is shown by many of the studies to be vulnerable to price volatilities not only transmitted from the agricultural input markets but also from the consumers market. A tendency to be responsive to positive price shocks rather than to negative ones is shown by Chavas and Mehta (2004), Zheng et al. (2008), Chavas and Mehta (2008), Uchezuba et al. (2010). These studies and other studies (Natcher and Weaver, 1999, Buguk et al. 2003, and Rezitis, 2012) are also an indication that upstream price shocks do transmit to the downstream sector. But there are also few studies that assert that this is not the case (for instance, Apergis and Rezitis, 2003; Serra, 2011a; and Kyiyavi et al., 2012). Overall, the reviewed studies indicate that price volatility can transmit in different directions and in different magnitudes depending on the commodities, countries and chain stages. Understanding the peculiarities in the ways price volatility transmits can help design public and private price risk management measures that are commodity, country and chain stage specific. From energy markets to agricultural commodity markets An overview of the studies on energy-agricultural commodity price volatility transmission is provided in Table 2 below. Table 2-Empirical studies on energy-agricultural commodity price volatility transmission
Author (date)
Country (Product)
Sample Period
Data frequency
Mean equations specifications
Specification of variance equations
Transmissio n of volatility detected
Direction of detected volatility transmission
Balcombe (2008)
(19 agricultural commodities and oil prices)
19732008
Monthly and annually1
---
---
Yes
From oil to food
20032009
Daily
VECM and VAR for different subperiods
BEKK and Causality in Variance of Cheung and Ng (1996)
Yes
From oil to corn
19892007
Weekly
VECM
BEKK
No
N/A2
19992009
Daily (forward and future prices)
ARMA
DCC
Yes3
N/A
VAR
A class of univariate GARCH models per sub-sample period
Yes
From oil to food
Harri and Hudson (2009) Zhang et al. (2009)
Busse et al. (2012)
Alom et al. (2011)
US (Crude oil and corn) US (Ethanol, gazoline, oil corn, soya) EU ( Rapeseed and rapeseed oil, soya bean and soya oil and Brent crude oil) Asia and Pacific countries (Oil and food products-price indices)
19952010
Daily
Du et al. (2011)
US (Corn, wheat and crude oil)
Kaltalioglu and SOyta (2011)
(Oil, food consumption item, and agricultural raw material)
19802008
Monthly
Onour and Sergi
World (Crude oil, fertilizers,
19922011
Monthly
19982009
Weekly (futures prices)
---
Stochastic volatility model
Yes
Bidirectional between oil and corn; Bidirectional between oil and wheat
Constants, ARMA
Univariate GARCH and Granger causality in variance approach
No
N/A
BEKK
Yes
From oil to corn and wheat;
7
(2011)
wheat and corn)
Ott (2012)
World (Crude oil, fertilizer, wheat, corn, soya and rice)
From fertilizers to corn and wheat 19602010
Monthly
Comparison of standard deviations of prices
---
Yes
N/A
Nazlioglu et al. (2012)
(Corn, wheat, soya, sugar and oil)
19862011
Daily
Constants
Univariate GARCH and causality in variance test
Yes
From oil to corn; Bidirectional between oil and soya; Bidirectional between oil and wheat
Serra et al. (2011)
Brazil (Crude oil, ethanol and sugar
20002008
Weekly
VECM
BEKK
Yes
Bidirectional across all products
Serra (2011b)
Brazil (International crude oil price, Brazilian ethanol and sugar)
20002009
Weekly
VECM
Parametric and nonparametric BEKK
Yes
From crude oil to ethanol; From sugar to ethanol
Wu et al. (2011)
US (Oil and corn)
19922009
Weekly (futures prices)
VECM
Yes
From oil to corn
Serra and Gil (2012)
US (Ethanol and corn)
19902010
monthly
VECM
Yes
From ethanol to corn (and vice versa)
Gardebroe k and Hernandez (2012)
US (Oil, ethanol and corn)
19972011
Weekly
VAR
BEKK and DCC
Yes
From corn to ethanol
TrujilloBarrera et al. (2012)
US (Crude oil, ethanol and corn)
20062011
Mid-week (futures prices)
AR
GJR-GARCH and BEKK
Yes
From crude oil to corn and ethanol; From corn to ethanol
Du and McPhail (2012)
US (Corn, ethanol and gasoline)
20052011
Daily (futures prices)
Structural VAR
DCC
Yes3
N/A
Univariate Threshold GARCH and BEKK Parametric and nonparametric BEKK
1
Approach that decomposed price volatility into volatility in mean and volatility in cycle. A panel regression is also used. Not applicable Based on covariance (correlation) of price volatilities
2 3
The number of studies on energy-agricultural commodity price volatility linkages shows the equal or even higher importance of this stream of literature compared to the literature that address food chain price volatility linkages (Table 1). The last decade marked by rising agricultural price volatilities is thoroughly covered in the former literature. Country coverage is dominated by the US. This can be just justified on the ground that a large share of global ethanol production (approximately 54%) is manufactured in the US (Gardebroek and Hernandez, 2012). However, since the inputs for ethanol production are globally traded, it is not unreasonable to expect price shocks to these inputs to be transmitted to the EU market. Methodological approaches are similar to those used in the price volatility transmission literature for food chains, indicating that a combination of VECM, AR or VAR specification for the mean part and specification of the variance as a GARCH process is standard in volatility modelling. The GARCH specification is in most of the studies in its multivariate form, in contrast to the food chain literature which rather uses a mixture of univariate and multivariate specifications. The fact that the literature focused on energy-agricultural commodity volatility linkages is recent literature (from 2009 onwards) further supports the argument that multivariate volatility models are more applied in recent times. Sample period coverage ranges from 10-15 years, with data frequency from daily to weekly. Use of
8
higher data frequency arises due to use of futures prices, which can easily be obtained since the products considered are easily traded energy and agricultural commodities. The findings of the papers support the conclusion that energy price volatility does transmit to agricultural commodities. The papers by Zhang et al. (2009) and Kaltalioglu and Soyta (2011) are exceptions, since they find no transmission of volatility between energy and agricultural markets. According to Table 2, it can be seen that there are more studies on the oil-agricultural commodity linkage than on the ethanol-agricultural commodity linkage. Since the effect of energy prices on food prices either comes through the use of energy as input in agricultural production or through the use of agricultural commodities for the production of ethanol (whose demand increases with rising oil prices), oil prices can be considered as reflecting these two effects. Moreover, in contrast to the studies that include ethanol in the energy-agricultural commodity linkage, all the studies that focus on oil prices find a positive effect of oil price volatility on agricultural commodity volatility (except Zhang et al. 2009, and Kaltalioglu and Soyta, 2011). Among the studies focused on ethanol prices, only Serra et al. (2011), Serra and Gil (2012) find that ethanol price volatility affects the volatility of agricultural commodity prices. Those studies that find a positive correlation between ethanol prices and agricultural commodity prices attribute this positive correlation not to the effect of ethanol price volatility on corn but to the effect of corn price volatility on ethanol (Gardebroek and Hernandez (2012), Du and Mcphail, 2012). The effect of energy price volatilities through the prices of agricultural means of production, rather than through the shift towards biofuel production, seems to better explain the increased energy-agricultural commodity price volatility linkages. Nevertheless, the general finding that energy prices fluctuations do transmit to agricultural commodity markets points to the necessity of considering these fluctuations in the attempt to stabilize food prices.
3. Factors with effects on the degree of price volatility transmission in food supply chains Stabilizing prices in food supply chains requires an understanding of the factors that contribute to the increase (or decrease) in the transmission of price uncertainties along the food chain. The current price volatility transmission literature (Table 1) lacks systematic identification and empirical testing of such factors. Nevertheless, some authors attribute the identified transmission (or non-transmission) of food price volatility to specific factors characterizing the chains being investigated. Table 3 provides a list of such factors as suggested in the literature listed in Table 1.
9
Table 3 – Factors with effects on price volatility transmission Author (Date)
Country (Product)
Suggested factors
Khan and Helmers (1997)
US (corn – pork, beef and poultry)
Large scale farm contract production
Buguk et al. (2003)
US (catfish)
Market power
Apergis and Rezitis (2003)
Greece (agricultural products)
Effect of suggested factor on price volatility transmission
Lack of contracting; biological nature of agricultural production (time lag in production response); Lower price elasticity of farm-level demand than that of retail demand Concentration in distribution and retailing; Low share of farm price in retail output price
As a result of contract production, the transmission of volatility in corn prices to beef and poultry farm prices is reduced. Cooperative organized farmers use market power to asymmetrically transmit positive input price shocks to the next stage.
The factors are attributed to the higher price volatility at farm level compared to agricultural input and retail output price volatility.
Large retailers with market power absorb upstream price shocks. Low cost share of the farm input leads to low transmission of farm price volatility.
Zheng et al. (2008)
US (45 retail food items)
Uchezuba et al. (2010)
South Africa (broiler)
Retail market power
Retailers use power to asymmetrically transmit unexpected positive farm price shocks.
Market power
Retailers with market power transmit farm price volatility in calmer periods (low news of BSE crisis) but do not transmit in turbulent times (detection of BSE cases leading to higher farm price volatility).
Serra (2011a)
Spain (beef)
Marketing strategy of retailers; Lower elasticity of farmlevel demand than that of retail demand
Alexandri (2011)
Romania (agricultural price indices)
Rezitis (2012)
Greece (beef, lamb, pork, poultry)
Retail concentration
Khiyavi et al. (2012)
Iran (poultry)
As a marketing strategy, retailers stabilize prices due to customers’ sensitivity to frequent price changes. Low elasticity of farm level demand causes farm prices to be more volatile than retail prices.
Retailer concentration leads to transmission of volatility from retail to farm in the poultry and lamb sector.
Lack of contracting; Biological nature of agricultural production (time lag in production response); Lower elasticity of farmlevel demand than that of retail demand
10
Factors are attributed to the higher price volatility at farm level compared to agricultural input and retail output price volatility
Many of the studies in Table 3 attribute market power at the retail stage to the low transmission of upstream price volatility, and in particular farm price volatility, to the consumer stage. Equivalent power exercised at the farmer stage is also argued to hinder the symmetric transmission of negative farm price shocks to the downstream stage (Buguk et al., 2003). The importance of contract farm production in reducing the transmission of farm input and retail output price volatilities to the farm stage is also stressed by some of the studies (Khan and Helmers, 1997; Apergis and Rezitis, 2003; Khiyavi et al., 2012). The assertions made in relation to retail market power seem an echo of the more numerous studies on price transmission (in levels), which in contrast to the price volatility transmission literature, have empirically tested and theoretically shown the relationship between market power and price transmission. This is an indication that factors that affect the degree of price transmission can also affect the transmission of price volatilities. To give a better idea of other factors that can also potentially affect the degree of price volatility transmission, Table 4 presents a brief summary of studies that have theoretically shown and empirically tested for several factors with effects on price transmission (in levels). Whether these studies have found asymmetric price transmission or not is also mentioned in the Table. Factors that cause asymmetries in the transmission of predictable price changes can potentially lead to asymmetries in the transmission of price volatilities, even though empirical tests are needed to confirm this hypothesis. Table 4 – Factors with effects on price transmission Author (Date)
Country (Product)
Detection of APT1
Reagan and Weitzman (1982)2 Mankiw (1985)2
---
Yes
Positive APT(+)/Negative APT (-) /(other findings) +
---
Yes
+
Shonkwiler amd Taylor (1988)
US (Orange Juice)
No
(Price rigidity for some price range)
Adjustment costs
Weaver et al. (1989)
US (Meat, poultry, eggs, fish, cereals)
---
Market concentration
Schroeter and Azzam (1991)
US (Pork)
---
McCorriston Sheldon (1996)
and
EU (Banana)
Yes
(Consumer Price explained by Prod. Price) (Retail margin explained by market power) +
Willet
US (Broiler)
Yes
+
Market power
Yes
(Price rigidity)
Adjustment costs
Bernard (1996)
and
low
Firm inventory management strategy Adjustment costs
Market power
Number of stages in the chain & market power
Levy et al. (1997)
US (Various food products)
Azzam (1999)2
---
Yes
---
Level of retail competition and repricing costs
Coffee
---
Cost share of farm input and Market power
Mc Corriston et al. (2001)2
---
---
(Incomplete transmission of farm price decrease to retail) ---
Miller and Hayenga (2001) Romain et al. (2002)
US (Pork)
Yes
+
US (Fluid milk)
Mixed
+/-
Bettendorf Verboven (2000)
and
retail
Cause of transmission/APT
11
Market power and processing technology in food industry Adjustment costs & market power Government regulation & market power
Fishman and Simhon (2005)2
---
---
Sticky downstream prices
Menu costs
Lloyd et al (2006)
UK (Beef)
---
Market power
Herrmann and Moeser (2006)
Germany (Various retail food products)
---
(Retail margin explained by market power) (Price rigidity)
Weber and Anders (2007) Acharya et al. (2011)
Germany (Beef and Pork) US (Strawberry)
Yes
(Price rigidity)
Mixed
+
Market power and firm pricing strategy Market power
Falkowski (2010)
Poland (Milk)
Yes
+
Market power
Bolotova and Novakovic (2012) Moro et al. (2012)
US (Milk)
Mixed
+
Government regulation
Italy (Beef, chicken & pork)
---
(Large % of retail margin due to market power)
Market power
Psychological pricing
1
Asymetric price transmission Indicates theoretical papers
2
Table 4 is a clear indication of the importance of market power as a factor determining the degree of price transmission. The effect of retail market power leading to positive asymmetric price transmission is empirically shown in many of the studies listed in Table 4. Nevertheless, there are also studies that assert that retail market power can lead to negative asymmetric price transmission, with upstream price decreases transmitted more than price increases due to efficiency achieved through increased market concentration (Weaver et al., 1989; Azzam, 1999). The price volatility transmission literature (see Table 3), surprisingly, does not mention the second important factor that determines the degree of price transmission: adjustment costs at the retail sector. These costs, which are also referred to as menu costs or repricing costs by some authors, refer to the costs that retailers incur in changing their prices (such as relabeling costs). Miller and Hayenga (2001) argue that, in the presence of menu costs, firms do not react to transitory price changes. Price volatilities by definition are price changes that are unanticipated in nature and which represent uncertainty over their future occurrence and on their magnitude. Therefore, one would expect these costs to also affect the degree of price volatility transmission to the retail sector. Among the other reasons for low price transmission (in levels) is the use of psychological pricing, which consists of setting prices just below some particular pricing points. These pricing points contribute to price rigidity because consumers are believed to react easily if prices go beyond those pricing points (Hermann and Moeser, 2006). The theory of psychological prices is related to the theory of kinked demand curve which asserts that firms’ gain is limited if they reduce prices, but loose largely if they raise prices (more elastic demand curve above the kink) (Hermann and Moeser, 2006). Therefore, firms choose to fix prices at levels slightly below the pricing points. Weber and Anders (2007) associate the rigidity of prices in hard discounters due to their price strategy of Every Day Low Pricing. These types of pricing strategies, the lost goodwill due to frequent price changes and the use of psychological pricing result in the rigidity of retail prices and thus in a low response to frequent input price changes. A large number of stages in the vertical chain, on the other hand, can result in a delay in the transmission of input price changes to the far downstream stages. Kinnucan and Forker (1987) attribute the positive asymmetric price transmission by retailers to government policy that sets price floors to the US dairy chain. They argue that middlemen may view farm price increases caused by government price support as permanent whereas they view the infrequent farm price decreases as temporary, resulting in a less complete pass-through of farm price decreases. 12
As both price transmission and price volatility transmission deal with price linkages, the price transmission literature can serve as a guide for empirical tests of the importance of the factors identified in Table 4 in explaining the degree of price volatility transmission in food supply chains. The transmission of prices (in levels) along the chain is necessary for an efficient market (Chavas and Mehta, 2004), for the protection of producers and consumers welfare, and for the effective transmission of policy induced price measures (Meyer and von Cramon-Taubadel, 2004; Vavra and Goodwin, 2005; Ben-Kaabia and Gil, 2007). Price volatility transmission, on the other hand, is the transmission of price uncertainty and risk from one market to another (Apergis and Rezitis, 2003). Therefore, one would rationally attempt to minimize such transmission to avoid the negative consequences. The measures that are intended to improve the transmission of prices (for example increasing the competitiveness of the retail sector, improving the bargaining power of farmers) may also lead to the perfect transmission of price volatility from one stage to another. One should therefore be aware of these conflicting objectives in attempting to improve the transmission of prices (in levels). A combination of measures might be necessary to achieve both objectives of reducing the degree of price volatility transmission and that of improving the degree of price transmission. Policy measures that reduce price volatility itself and private coping measures (such as contracts and derivative markets) can achieve the objective of reducing the exposure of actors to price volatility. On the other hand, improving the competitiveness of the chain can help transmit predictable price signals along the chain.
4. Market measures of the CAP Measures that aim at reducing price volatility have an implication on the level of volatility that is transmitted to the next stage (Buguk et al., 2003). Therefore, one would expect the stabilization measures of EU farm-gate prices, which have long been an important feature of the CAP (Van Meijl and Van Tongeren, 2003), to have an implication on the price volatilities faced by actors in the rest of the food supply chain. Studies have investigated the implementation of relevant CAP market stabilization measures for EU agriculture and evaluated the degree to which these measures can and have achieved price stability within the EU. The studies cover the main price stabilization measures that were implemented and those that are still being implemented since 1992. Market measures within the CAP are not only limited to market stabilization, but also include measures that aim at improving the competitiveness of the chain. These later measures have their own implications on the transmission of prices and volatilities along the chain. This section first presents a review of studies that investigated issues related with the CAP’s price stabilization measures. A second subsection will briefly present the CAP measures intended to improve the competitiveness of food supply chains.
4.1.
Price stabilization measures
Export refunds One of the measures used in the CAP to achieve agricultural price stability within the EU is export refund that allows excess supplies to be disposed of on world markets (Van Meijl and Van Tongeren, 2003). This instrument particularly benefits the dairy, sugar, pork and poultry sectors since these products were uncompetitive in the world market as a result of large difference between world and EU markets. (Chatellier, 2011). Van Meijl and Van Tongeren (2003) argue that the use of this instrument 13
is challenged by the increasing export refund restriction in terms of volume and expenditure brought by WTO negotiations. The use of all forms of export subsidies is expected to be eliminated by WTO member States by the end of 2013 (Chatellier, 2011). Van Meijl and Van Tongeren (2003) show through a partial equilibrium analysis that the degree to which the Agenda 2000 reform (which reduced intervention prices, increased milk quota and reduced area set-aside) helps the EU stay within its subsidized export limits depend on world prices and on euro-dollar exchange rate. That is, low world prices and a depreciating dollar can make EU exports more expensive in the world market and may therefore require the use of export subsidies above the set limits. Binfield et al. (2010) show through a stochastic modelling applied to the FAPRI-UK modelling system that export subsidies are needed more for butter than for cheese and skimmed milk powder as these last two products are competitive in the world market. They stress that the abolition of export subsidies in WTO future agreements will have a significant impact on the EU dairy sector. Intervention buying Public authorities may store certain agricultural products, or provide aid to private storage, when their price falls below a threshold and release the stock in the EU or international market when the need arises (Chatellier, 2011). This tool is applicable for storable products like cereals, rice, meat and meat products, powdered milk and milk products (Bardají et al., 2011). The success of public storage lies, among others, on the possibility of disposing stock surpluses in the world market through the use of export subsidies, on world market price levels, and on the budgetary burden brought by further reduction of tariffs (since intervention should include imported produce as well) (Bardají et al., 2011). The impact of the reduction in intervention prices on the stability of farmers’ income depends on whether such reduction is accompanied by other compensating measures. Fraser (2003) shows for instance that a 5% reduction in the support price of cereals from the Agenda 2000 level in the case of the ‘Fishler’ proposal (which also retains land set-aside), as well as complete removal of price support, leads to increase variability of EU prices and farmers’ income. This is particularly true when world prices are volatile and at low level, in which case farmers require large compensation payments. But if the reduction in the support price is accompanied by a reduction in the set-aside land, then income variability is low for farmers that are cost-efficient and who therefore require low compensations (Fraser, 2003). The effect of lower intervention price was also shown by Bouamra-Mechemache et al. (2008) for the case of milk to have no depressing effect on milk price if the demand for the milk product is high and growing. While this might be true for milk products for which demand is higher (for instance, higher for skimmed milk powder compared to butter) it is acknowledged that the steady restriction of intervention buying together with an approximately 20% reduction in intervention prices brought by the 2003 Luxembourg agreement created a much greater variability of prices, particularly during downside market movements (Keane & O Connor, 2009). Keane and O Connor (2009) further stress the importance of intervention buying and timely release of stocks to avoid extreme volatilities. They show that the large upward swing in butter prices in 2007/2008 was met with a virtual exhaustion of stocks during the same period, thereby exacerbating the price swing. Binfield et al (2010) show that, in contrast to butter, the world skimmed milk prices is almost always above the intervention prices making EU skimmed milk powder prices more competitive. This however results in a higher variability of skimmed milk powder prices as more of the world price volatility is transmitted to the EU market. Production quotas
14
A study by Bouamra-Mechemache et al. (2008) for the milk sector showed that the gradual increase in the milk quota from 2009 until 2015 at higher rates than the rates proposed in the Luxembourg agreement led to a significant increase in production and strongly depressed EU prices. They showed that the effect was particularly strong for butter prices compared to skimmed milk powder prices. This was attributed to the low demand for butter both at the local and world markets. This finding is an indication that prices can be stable and high even in the event of an increase in quota levels if demand levels are high. Keane and Konnor (2009) show that an inflexible quota system can lead to increased price volatility both with respect to price peaks and price falls, therefore supporting the view that quotas should not be rigid. The same view is supported by Nolte et al. (2012) who show by taking the case of the EU sugar sector that EU prices are more stable when quotas are abolished, as planned after 2014/2015, and are able to respond to world price signals. Border protection A further reduction of import levies to protect the EU market is expected given the agreements in WTO negotiations. In the EU, while consolidated duty on agricultural and food products stands at an EU average of just under 20%, the 80% threshold is surpassed for sugar, beef and butter (Chatellier, 2011). The use of levies to restrict imports to the EU is challenged by the WTO commitments to abolish any form of export subsidies post 2013 (Van Meijl and Van Tongeren, 2003). The surplus that can potentially be created in case of low world market prices can further increase with the WTO commitments to reduce import restrictions. According to Van Meijl and Van Tongeren (2003), the EU should therefore focus its attention on means of reducing internal prices and make the EU stand in the face of competition from imported products. While the abolishment of quota production in the dairy sector and the reduction in guaranteed prices in the cereal sector might reduce the downward pressure on EU prices brought by tariff reductions (since imports become more expensive than EU internal prices), the meat sector is more prone to downward price instability as a result of such action (Chatellier, 2011). This is because of the high EU prices for meat products compared to imported products. For the sugar sector, Nolte et al. (2012) shows that the effect of imports on EU prices depends on the level of world market prices. In a situation of quota abolishment, if world market prices are high and increasing, imports to the EU are discouraged because the EU market becomes unattractive to exporters as a result of low and strongly competitive EU prices. For the fruits and vegetable sector, Cioffi et al. (2011) show that the price stabilizing effect of the entry price system introduced in 1995 is mixed. This system consists of charging an import duty that is roughly equal to the difference between a reference EU price and the Standard Import Value of the imported product. While Cioffi et al. (2011) find that this system is effective in sheltering the EU domestic market of fruits and vegetables from cheap imports for the case of imported tomatoes from Morocco and lemons from Turkey, this price stabilization effect is not obvious for tomatoes imported from Turkey and lemons imported from Argentina. Thompson et al. (2000) show with a theoretical model that the EU traditional grain policy instruments of variable import level and use of intervention prices protected the EU market from domestic price instabilities. However, they show that perfect tariffication using ad valorem tariffs led to instability of EU prices, which became linked to variations in world demand levels (Thompson et al., 2000). On the other hand, they show that the implementation of a minimum import price equal to “155% of the intervention” price does not lead to EU price instabilities. But this happens if export refunds are not restricted, as the EU domestic prices will be set exogenously (Thompson et al., 2000). Thompson et al (2002) show the CAP reform in 1992 that among others, reduced intervention prices, and the 15
subsequent Uruguay round that resulted in the tariffication through the use of minimum import price (equal to 155% of the intervention price) did not result in an increase in EU price volatilities. They attribute this to the presence of variable import subsidies that helped to stabilize EU prices. Hayes et al. (1992) show through simulations for the beef sector that tariff reduction formulas that do not take into account world price conditions lead an increase variability of EU prices. He shows that the use the Swiss formula (proposed in the Tokyo round of GATT negotiations) that gradually reduces tariffs helps to delay the transmission of world price shocks to the EU. The review reveals that the EU agriculture is increasingly exposed to world market price signals resulting from CAP reforms that made EU farmers more market oriented and from WTO commitments to open borders and restrict export subsidies. These measures have traditionally kept EU agricultural prices at high and stable level. The further reduction of border protection and the prospective removal of export subsidies is expected to increase the downward variability in EU prices. The reduction of intervention prices and the abolishment of production quota might make imports uncompetitive in the EU because they lower EU prices and thus help avoid the further downward pressure on EU prices brought by imports. However, such effect greatly depends on world market prices and exchange rate conditions. As a result, EU farm gate prices might be further exposed to world price signals and not be as stable as in the past decade. All this implies that the downstream sector will also be increasingly exposed to price volatilities because farm prices volatilities can transmit to the rest of the chain. It is worth noting that the opening of the EU market to the world market has its own implication on international food price volatilities. Studies have shown that EU market measures such as border protection lead to increases in volatilities in the world food prices (Cantore, 2012; Tangermann, 2011). Thompson et al. (2000) show that EU trade liberalization helps to reduce world food price volatilities. In the long term, this implies that world market prices can become more stable than with a protected EU market, thereby reducing the volatility that is transmitted from the world to the EU market.
4.2.
Measures improving the competitiveness of the chain
Encouragement of producer organizations and inter-professional organizations Producer organizations, which mostly are cooperatives but also groups of individuals or groups of companies, are one of the means to increase the market position of producers and stabilize producer prices (Bardají et al., 2011). Legally recognized inter-professional organizations, which coordinate vertical actions (such as supply concentration and marketing) in a supply chain, are another means that among others are beneficial to stabilize producer and consumer prices. Inter-branch organizations help producer prices to be stable through the use of contracts, through the ability to regulate prices by setting minimum upstream prices (ex, fruit and vegetable sectors in France), and through the ability to regulate supply (ex, the wine sector in France) (Bardají et al., 2011). Producer organizations and interprofessional organizations, which tend to be sector specific, can improve the level of transparency and price discovery in the chain (Tothova and Velazguez, 2012). These organizations are granted exemption from a series of competition rules (Tothova and Velazguez, 2012). Collusive behaviour such as for price fixing is prohibited in the EU in Articles 101 of the Treatment on the Functioning of the EU (TFEU) (Bardají et al., 2011). However, Articles 175 and 176 of EC regulation 1234/2007 allows for exemptions to the agricultural sector provided that, among others, the negotiation among farmers involves sharing of production or marketing facilities (Bardají et al., 2011). 16
An example of such exemption is the recent milk proposal that enables producer organizations to collectively negotiate contract terms including prices (Bardají et al., 2011). Such exemption not only allows farmers to get a fair price for their produce, but also helps stabilize prices both for the farmers and the buyers. A precise definition of the types of producer and inter-branch organization that qualify for legal protection from competition is crucial for these organizations to flourish (CopaCogeca, 2013). Contracts standardization Contracts allow contracting parties to fix negotiated prices for certain duration of time and minimize their exposure to price volatility. Bardají et al. (2011) argue that standardization of contracts by specifying the basic elements to be included helps ease negotiations, reduces transaction costs, better organize market transactions and minimize unfair practices. In the EU, Spain has implemented standardized contracts since 2000 and makes use of Contract Monitoring Committees ensuring the adherence to fodder, fruit, vegetables, potatoes, tobacco and milk contracts (Bardají et al., 2011).
CAP (market prices related) measures to be maintained post 2013 Market management measures (LEI, 2013; European Comission, 2011): Internal measures Private storage aid for cereals, rice, sugar, olive oil and table olives, beef and veal, milk and milk products, pigmeat, sheepmeat and goatmeat. Public intervention for cereals, rice, skimmed milk powder, butter and beef. Production quotas for dairy (until 2015), sugar (until 2017) and wine. Trade measures Export subsidies. Abolition in 2013 contingent on WTO agreements. Border protection through import duties, import quotas and requirement of import licenses. Measures to improve chain competitiveness (Bardají et al., 2011) Encouragement of producer and inter-branch organizations. Standardization of contracts.
17
5. Conclusions The last decade, and in particular the period since 2007/2008, was marked by increased price volatility for food and agricultural products both at the world and EU levels. Studies have shown that the increased market orientation of EU agriculture achieved through successive market reforms since 1992 have resulted in the transmission of international food price volatilities to the EU market. Empirical studies have provided evidence that the effect of price volatility is not limited to the farm stage, but that it also extends to the downstream sector of food supply chains. In light of this, this deliverable provided an exhaustive review of the existing literature on food and agricultural price volatility transmission in food supply chains. The factors with potential effects on the degree of price volatility transmission were also identified and discussed. In addition, the review included studies that deal with the transmission of price volatilities from energy markets to agricultural commodity markets (mainly those used for feedstock). The following general conclusions can be drawn from the review:
With few exceptions, Generalized Autoregressive Conditional Heteroskedasticity models (GARCH) are applied in most of the reviewed studies (both literature on food chains, and that on energy-agricultural commodity markets), suggesting that these models are standard in price volatility modelling. More recent studies use multivariate GARCH models (MGARCH) to investigate volatility transmissions, with a particular focus on the BEKK model. Price data with higher frequency (up to daily) and shorter sample period (10-15 years) characterize energy-agricultural commodity price volatility transmission studies compared to food chain studies (from farm to retail). In the latter literature, the minimum frequency of price data is monthly and sample period ranges from 15-20 years. The literature on volatility transmission in food chains (from farm to retail) is limited in the period that it covers, in that the period from 2006 onwards is poorly covered. The literature on price volatility transmission in energy and agricultural commodity markets, on the other hand, thoroughly covers the last decade (from 2000-present). Country coverage is dominated by the US market in both literatures, and in particular the energy-agricultural commodity literature. The EU market is inadequately covered, with only three countries (Greece, Spain and Romania) covered in the literature that addresses the farm-retail chain. The commodity coverage for this same literature is also very limited, with a focus on the meat sector. The literature on energy-agricultural commodity price volatility linkage is limited to the upstream sector, and does not deal with the food chain implications of energy price volatilities. The food chain literature (from farm to retail) shows that food price volatility does transmit across the different stages of food supply chains, thereby exposing all chain actors to price uncertainties. The evidences suggest that price volatility transmitted from the consumer stage is as important, if not more, than the price volatility from the upstream stage. The meat farm sector was particularly found to be prone to price volatility sourced both from its inputs and outputs markets. The volatility in input markets (such as feed markets) is further exacerbated by price shocks occurring in energy markets. Contrary to general belief, the findings on increased price volatility linkage between energy and agricultural markets do not seem to arise from the use of agricultural commodities for biofuel (ethanol) production. While it is 18
generally found that increase in crude oil price volatilities increases volatilities in agricultural commodity prices, and that increase in agricultural commodity price volatilities lead to increase in the volatility of ethanol prices, the studies rarely find a positive effect of ethanol price volatilities on agricultural commodity prices. Nevertheless, the finding that energy price volatilities transmit to agricultural commodity markets is evidence of the need to consider such spillover in the attempt to stabilize food prices. The reviewed studies do not empirically test for the factors that affect the transmission of price volatility transmission in food supply chains. Yet, the assertions made by a majority of the authors suggest that the degree of market power and the availability of contracts determine whether price volatility transmits along the chain. These factors are asserted to reduce the transmission of price volatility. A review of the price transmission literature (for price levels), on the other hand, suggest that a number of other factors determine the degree of transmission of prices. Among these factors, the two major ones are market power and menu costs in the downstream sector. Future research can investigate if the same factors empirically tested for in the price transmission literature also affect the degree of price volatility transmission along the food supply chain. Past successive reforms of the CAP and further opening up of the EU market as part of WTO commitments has exposed and can further expose the EU farm sector to price volatilities. The use of price stabilization measures which include border protection, export refunds, production quota and intervention buying is gradually being reduced, with prospective discontinuation of the use of export refunds and production quota for some products in the near future. The empirical evidence that price volatilities transmit along the chain suggest that the potential further increased of farm price volatilities can translate into increase in downstream sector price volatilities as well.
19
References Acharya, R.N, Kinnucan, H.W and Caudill, S.B (2011). Asymmetric farm–retail price transmission and market power: a new test, Applied Economics, 43 (30), 4759-4768 Alexandri, C. (2011). Analysis of price transmission along the agri-food chains in Romania, Agricultural Economics and Rural Development, 8(2), 171-189 Alom, F., Ward, B., and Hu, B. (2011). Cross country mean and volatility spillover effects of food prices: multivariate GARCH analysis, Economic Bulletin, 31 (2), 1439–1450 Apergis, N. and Rezitis, A. (2003). Agricultural price volatility spillover effects: the case of Greece, European Review of Agricultural Economics, 30 (3), 389-406 Azzam, A.M. (1999). Asymmetry and rigidity in farm-retail price transmission, American Journal of Agricultural Economics, 81 (3), 525-533 Baba, Y., Engle, R. F., Kraft, D. F. and Kroner, K. F. (1991). “Multivariate Simultaneous Generalized ARCH,”manuscript, Dept. of Economics, UCSD Balcombe, K. and Rapsomanikis, G. (2008). Bayesian estimation and selection of nonlinear vector error correction models: the case of sugar-ethanol–oil nexus in Brazil, American Journal Agricultural Econonmics, 90 (3), 658–668 Bardaji, M., Garrido, A., Iglesias, E., Blanco, M. and Bielza, M. (2011). What market measures in the future CAP after 2013? Retrieved in October, 2012 from www.europarl.europa.eu/studies Ben-Kabbia, M. and Gil, J.M. (2007). Asymmetric price transmission in the Spanish lamb sector, European Review of Agricultural Economics, 34 (1), 53-80 Bernard, J.C., and Willett, L.S.(1996). Asymmetric price relationships in the US broiler industry, Journal of Agricultural and Applied Economics, 28 (2), 279–289 Bolotova, Y.V. and Novakovic, A.M. (2012). The Impact of the New York State milk price gouging law on the price transmission process and supermarket pricing strategies in the fluid whole milk market, Agribusiness, 28 (4), 377–399 Bettendorf, L. and Verboven, F. (2000). Incomplete transmission of coffee bean prices: evidence from The Netherlands, European Review of Agricultural Economics, 27 (1), 1-16 Binfield, J., Moss, J., Zhang, L., Patton,M. and Kim, I.S (2010). Impact of volatile world prices on the European dairy sector, Eurochoices, 9(2), 26-27 Bouamra-Mechemache, Z., Jongeneel, R. and Requillart, V. (2008). Impact of a gradual increase in milk quotas on the EU dairy sector, European Review of Agricultural Economics, 35(4), 461-491 Brummer, B., Von Cramon-Taubadel, S. and Zorya, S. (2009). The impact of market and policy instability on price transmission between wheat and flour in Ukraine, European Review of Agricultural Economics, 36(2), 203-230
20
Buguk C., Hudson, D., and Hanson, D.(2003). Price volatility in agricultural markets: An examination of U.S. catfish markets, Journal of Agricultural and Resource Economics, 28(1), 86-99 Busse, S., Brümmer,B. and Ihle, R. (2012). Price formation in the German biodiesel supply chain: a Markov-switching vector error correction modeling approach, Agricultural Economics, 43 (5), 545– 560 Cantore, N. (2012). Impact of the Common Agricultural Policy on food price volatility for developing countries. Retrieved in June 2013 from http://www.odi.org.uk/publications/6944-impact-commonagricultural-policy-food-price-volatility-developing-countries Chatellier, V. (2011). Price volatility, market regulation and risk management: challenges for the future of the CAP. Retrieved in April 2013 from http://ideas.repec.org/p/rae/wpaper/201104.html Chavas, J.P and Mehta, A. (2004). Price dynamics in a vertical sector, Agricultural and Applied Economics Association, 86 (4), 1078-1093 Cioffi, A., Santeramo, F.G. and Vitale, C.D. (2011). The price stabilization effects of the EU entry price scheme for fruit and vegetables, Agricultural Economics, 42 (3), 405-418 Copa-Cogeca (2011). The reaction of EU farmers and agri-cooperatives to the Comission’s legislative proposals. Retrieved in July 2013 from http://www.copacogeca.be/img/user/file/PAC2013/2012futureCAP_EN.pdf Du, X., Yu, C.L., Hayes and D.J. (2011). Speculation and volatility spillover in the crude oil and agricultural commodity markets: a Bayesian analysis, Energy Economics, 33 (3), 497–503 Du, X. and McPhail, L. (2012). Inside the black box: the rice linkage and transmission between energy and agricultural markets, Energy Journal, 33 (2), 171–194 European Commission (2011). Summaries of EU legislations. Retrieved in July 2013 from http://europa.eu/legislation_summaries/agriculture/agricultural_products_markets/l67001_en.htm FAO, IFAD, IMF, OECD, UNCTAD, WFP, the World Bank, the WTO, IFPRI and the UN HLTF (2011). Price volatility in food and agricultural markets: policy responses. Retrieved in October 2012 from http://www.oecd.org/dataoecd/40/34/48152638.pdf. Falkowski, J. (2010). Price transmission and market power in a transition context: Evidence from the polish fluid milk sector, Post-Communist Economies, 22(4), 513-529 Fishman, A. and Simhon, A. (2005). Can small menu costs explain sticky prices? Economics Letters, 87 (2), 227-230 Fraser, R. (2003). An evaluation of the compensation required by European Union cereal growers to accept the removal of price support, Journal of Agricultural Economics, 54(3), 431-445 Frey, G. and Manera, M. (2007). Econometric models of asymmetric price transmission, Journal of Economic Surveys, 2(21), 349-415 Gardebroek, C. and Hernandez, M.A. (2012). Do energy prices stimulate food price volatility? Examining volatility transmission between US oil, ethanol and corn markets. Paper prepared for the 123rd EAAE seminar, Dublin, 2012 21
Goodwin, B.K. and Holt, M.T. (1999). Price transmission and asymmetric adjustment in the U.S. beef sector, American Agricultural Economics Association, 81(3), 630-637 Harri, A., and Hudson, D. (2009). Mean and variance dynamics between agricultural commodity prices and crude oil prices. Paper presented at the Economics of Alternative Energy Sources and Globalization, The Road Ahead meeting, Orlando, Florida, 2009 Hayes, D.J., Wahl, T.I and Johnson, S.R.(1992). A proposal for the reduction of domestic price variability during the phase-in period of trade liberalization, Agricultural Economics, 7 (1), 55-64 Hernandez, M., Ibarra, R. and Trupkin, D. (2011). How far do shocks move across borders? Examining volatility transmission in major agricultural futures markets. Retrieved in September 2012 from http://www.um.edu.uy/docs/working_paper_um_cee_2011_09.pdf Herrmannn, R. and Moeser, A. (2006). Do psychological prices contribute to price rigidity? Evidence from German scanner data on food brands, Agribusiness, 22(1), 51-67 Keane, M. & O Connor, D. (2009). Price volatility in the EU dairy industry: Causes, consequences and coping mechanisms. Retrieved in May 2013 from http://www.agriculture.gov.ie/media/migration/agrifoodindustry/foodharvest2020/foodharvest2020/submissionsreceived/EDA%20Circulated%20Report2 60310.pdf Khan, M.A. and Helmers, G.A.(1997). Causality, input price variability and structural changes in the U.S. livestock-meat industry. Paper submitted to Western agricultural economics association meeting, Reno, Nevada, 1997 Khiyavi, P.K., Moghaddasi, R., Eskandarpur, B. and Mousavi, N. (2012). Spillover effects of agricultural products price volatilities in Iran, Journal of Basic and Applied Scientific Research, 2(8), 7906-7914 Kinnucan, H.W. and Forker, O.D.(1987). Asymmetry in farm-retail price transmission for major dairy products, American Journal of Agricultural Economics, 69 (2), 285-292 Kuiper, W.E. and Oude Lansink, A.G.J.M. (2013). Asymmetric price transmission in food supply chains: Impulse response analysis by local projections applied to U.S. Broiler and Pork prices, Agribusiness, 0(0), 1-19 Kaltalioglu, M. and Soytas, U. (2011). Volatility spillover from oil to food and agricultural raw material markets, Modern Economy, 2, 71-76 LEI (2013). CAP and EU trade policy reform, assessing impact on developing countries. The Netherlands, The Hague: Meijerink, G. and Achterbosch, T. Levy, D., Dutta, S., Bergen, M., and Venable, R. (1997). The magnitude of menu costs: Direct evidence from large U.S. supermarket chains, The Quarterly Journal of Economics, 112 (3), 791-825 Lloyd, T.T., McCorriston, S.,Morgan, C.W. and Rayner, A.J. (2006). Food scares, market power and price transmission: the UK BSE crisis, European Review of Agricultural Economics, 33(2), 119-147 Mankiw, N.G. (1985). Small menu costs and large business cycles: A macroeconomic model of monopoly, The Quarterly Journal of Economics, 100 (2), 529-537
22
Mc Corriston, S. and Sheldon, I.M, (1996). The effect of vertical markets on trade policy reform, Oxford Economic Papers, 48 (4), 664-672 McCorriston, S., Morgan, C.W. and Rayner, A.J. (2001). Price transmission: The interaction between market power and returns to scale, European Review of Agricultural Economics, 28 (2), 143-159 Mehta, A. and Chavas, J.P (2008). Responding to the coffee crisis: What can we can we learn from price dynamics? Journal of Development Economics, 85(2), 282-311 Meyer, J. and von Cramon-Taubadel, S. (2004). Asymmetric price transmission: A Survey, Journal of Agricultural Economics, 55 (3), 581-611 Miller, D.J. and Hayenga, M.L. (2001). Price cycles and asymmetric price transmission in the U.S. market, American Journal of Economics, 83(3), 551-562 Moro, D., Sckokai, P. and Veneziani, M. (2012). Multi-stage market power in the Italian fresh meat industry (Working paper No. 7). Retrieved in November 2012 from http://www.transfop.eu/media/universityofexeter/businessschool/pdfs/research/transfop/TRANSFOP_ WP_No_7.pdf Natcher, W.C. and Weaver, R. (1999). The transmission of price volatility in the beef market: A multivariate approach. Paper selected for presentation at the American Agricultural Economics Association annual meeting, Nashville, Tennessee, 1999 Nazlioglu, A., Erdem, C. and Soytas, U. (2012). Volatility spillover between oil and agricultural commodity markets, Energy Economics, 36, 658-665 Nolte, S., Buysse, J. and Van Huylenbroeck, G. (2012). Modelling the effects of an abolition of the EU sugar quota on internal prices, production and imports, European review of agricultural economics, 39(1), 75-94 Onour, I. and Sergi, B. (2011). Global food and energy markets: volatility transmission and impulse response effects. MPRA Paper 34079.University Library of Munich, Germany Ott, H. (2012). Fertilizer markets and their interplay with commodity and food prices. Retrieved in June 2012 from http://ipts.jrc.ec.europa.eu/publications/pub.cfm?id=5059 Piot-Lepetit, I. (2011). Price volatility and price leadership in the EU beef and pork meat market (PiotLepetit, I. and M’Barek, R., Eds.). In Methods to Analyse Agricultural Commodity Price Volatility (pp.85-106). New York: Springer Science+Business Media Rabobank (2011). Rethinking the food and agribusiness supply chain; impact of agricultural price volatility on sourcing strategies. Retrieved in September 2012 from http://hugin.info/133178/R/1549493/476482.pdf Reagan P.B. and Weitzman M.L. (1982). Asymmetries in price and quantity adjustments by the competitive firm, Journal of Economic Theory, 27(2), 410-420 Rezitis, A. and Stavropoulos, K. (2009). Modelling pork supply response and price volatility: The case of Greece, Journal of Agricultural and Applied Economics, 41 (1), 145-162
23
Rezitis, A. (2010). Mean and volatility spillover effects in Greek producer-consumer meat prices, Applied Economic Letters, 10(6), 381-384 Rezitis, A.N. and Stavropoulos, K.S. (2011). Price transmission and volatility in the Greek broiler sector: A Threshold Cointegration analysis, Journal of Agricultural and Industrial Organization, 9(1), 1-35 Rezitis, A. N. (2012). Modeling and Decomposing Price Volatility in the Greek Meat Market, International Journal of Computational Economics and Econometrics, Forthcoming. Available at SSRN: http://ssrn.com/abstract=2113502 Romain, R. , Doyon, M. and Frigon, M., (2002). Effects of State regulations on marketing margins and price transmission asymmetry: Evidence from the New York City and Upstate New York Fluid milk markets, Agribusiness, 18(3), 301-315 Schroeter, J.R. and Azzam, A. M. (1991). Marketing margins, market power, and price uncertainty, American Journal of Agricultural Economics, 73(4), 990-999 Sckokai, P. and Moro, D. (2009). Modelling the impact of the CAP single farm payment on farm investment and output, European Review of Agricultural Economics, 36 (3), 395-423 Seal, J. and Shonkwiler, J. (1987). Rationality, price risk, and response, Southern Journal of Agricultural Economics, 19 (1), 111-118 Serra, T., Zilberman, D. and Gil, J. (2011). Price volatility in ethanol markets. European Review of Agricultural Economics, 38 (2), 259–280 Serra, T. (2011a). Food scares and price volatility: The case of the BSE in Spain, Food policy, 36 (2), 179-185 Serra, T. (2011b). Volatility spillovers between food and energy markets: a semiparametric approach, Energy Economics, 33 (6), 1155–1164. Serra, T. and Gil, J.M. (2012). Price volatility in food markets: can stock building mitigate price fluctuations? European Review of Agricultural Economics, 40 (3), 1-22 Shonkwiler, J.S. and Taylor, T.G. (1988). Food processor price behaviour: Firm level evidence of sticky prices, Agricultural and Applied Economics, 70(2), 239-244 Tangermann, S. (2011). Risk management in agriculture and the future of the EU’s Common Agricultural Policy. Retrieved in October 2012 from http://ictsd.org/downloads/2011/12/riskmanagement-in-agriculture-and-the-future-of-the-eus-common-agricultural-policy.pdf
Taya, S. (2012). Stochastic model development and price volatility analysis, OECD food, agricultura and fisheries working papers, No. 57, OECD publishing. Retrieved in October 2012 from http://dx.doi.org/10.1787/5k95tmlz3522-en Thompson, S.R., Herrman, R. and Gohout, W. (2000). Agricultural market liberalization and instability of domestic agricultural markets: The case of the CAP, American Journal of Agricultural Economics, 82 (3), 718-726
24
Thompson, S.R., Gohout, W, and Herrman, R. (2002). CAP reforms in the 1990s and their price and welfare implications: The case of wheat, Journal of Agricultural Economics, 53(1), 1-13 Tothova, M. and Velasquez, B. (2012). Issues and policy solutions to commodity price volatility in the European Union. Paper presented to the 2nd IFAMA annual symposium: The road to 2050, the China Factor, Shanghai, 2012 Trujillo-Barrera, A, Mallory, M. and Garcia, P. (2012). Volatility spillovers in U.S. crude oil, Ethanol, and corn futures markets, Journal of Agricultural and Resource Economics, 37(2), 247-262 Uchezuba, I.D., Jooste,A. and Willemse, J. (2010). Measuring Asymmetric Price and Volatility Spillover in the South African Broiler Market. Retrieved in June 2013 from http://ageconsearch.umn.edu/bitstream/96434/2/179briol.pdf Van Meijl, M. and Van Tongeren, F. (2003). The agenda 200 CAP reform, world prices and GATTWTO export constraints, European Review of Agricultural Economics, 29 (4), 445-470 Vavra, P. and B. K. Goodwin (2005). Analysis of Price Transmission Along the Food Chain, OECD Food, Agriculture and Fisheries Working Papers, No. 3, OECD Publishing. doi:10.1787/752335872456 Von Cramon-Taubadel, S. (1997). Estimating asymmetric price transmission with the error correction representation: An application to the German pork market, European Review of Agricultural Economics, 25 (1), 1-18 Weaver, R.D., Chatting, P. and Banerjee, A. (1989). Market structure and the dynamics of retail food prices, North Eastern Journal of Agricultural and Resource Economics, 18 (2), 160-170 Weber, S.A. and Anders, S.M. (2007). Price rigidity and market power in German retailing, Managerial and Decision Economics, 28 (7), 737-749 Wu, F., Guan, Z. and Myers, R.J.(2011). Volatility spillover effects and cross hedging in corn and crude oil futures, The Journal of Futures Markets, 31(11), 1052-1075 Zhang, Z., Lohr, L., Escalante, C and Wetzstein, M. (2009). Ethanol, corn, and soybean price relations in a volatile vehicle-fuels market, Energies, 2 (2), 320-339 Zheng, Y. Kinnucan, H.W. and Thompson, H.(2008). News and volatility of food prices, Applied Economics, 40(13), 1629-1635
25
