Seasonal price. patterns in EU vegetable markets

Transcription

1 Seasonal price patterns in EU vegetable markets Student: Supervisor: Examiner: Joost Noorland Dr. R. Ihle Dr. J. Peerlings 1

2 Acknowledgements This MSc thesis was written to complete my Master Management, Economics and Consumer Studies. I really enjoyed writing this thesis, since the topic of this thesis connects well with my interests in agricultural markets. I am therefore very thankful for the opportunity I got to write a thesis in this field of research. I take this opportunity to express my gratitude to Dr. R. Ihle for his help and support. Especially his detailed comments were very useful. Furthermore, I really appreciated his guidance and encouragement he gave me when writing this thesis. I also want to thank my fellow students for the valuable advices and motivation. I hope you will enjoy reading this thesis! Joost Noorland Wageningen,

3 Abstract The aim of this thesis is to analyse seasonal price patterns of EU vegetables. The central concept in this thesis is seasonality, which is reflected by the price amplitude in price data. The price amplitude is the maximum price minus the minimum price in a given period. Special attention is paid to the differences in seasonal price patterns between storable and perishable vegetables. Futrell and Wisner (1982) argue that price movements of perishable vegetables may be greater than price movements of storable vegetables. This thesis builds on prior academic work relating to seasonal price patterns of agricultural products. This is applied to the EU vegetables market. In this thesis, it is analysed whether seasonality exists in the seasonal price patterns, and how stable these price patterns are over time, and across vegetables. Empirical evidence finds that perishable vegetables have, on average, more seasonality, and perishable vegetables are generally more unstable than storable vegetables. Furthermore, it was found that none of the seasonal price patterns are stable over time. These findings connect quite well with the existing theory. 3

4 Content Acknowledgements... 2 Abstract... 3 List of Figures... 7 List of Tables... 8 List of equations Introduction Problem definition Research objective Research questions Theoretical framework Methodology Data Content overview Determinants of seasonal patterns in vegetable prices Definitions Seasonality Storability Determinants of seasonality Demand Supply International trade Storing vegetables Structure of EU vegetables market Farm structure Production Consumption Trade Imports from outside the EU Exports outside the EU Destinations of EU vegetable exports Origins of EU vegetable imports Intra-EU trade in vegetables Common Market Organisation

5 3.5.1 CAP Reforms Evaluation of the Common Market Organisation Empirical evidence Descriptive analysis Data description Seasonal price patterns Annual price amplitudes Monthly price amplitudes Summary of key findings Methodology Dummy time series models Step functions Results Explanatory power of model (1) Presence of seasonality Seasonal price patterns by step functions Price stability Quarterly price stability Price stability across vegetables Price stability across years Conclusions Discussion References Appendices Appendix I Appendix II Appendix III Appendix IV Appendix V Appendix VI Appendix VII Appendix VIII Appendix IX Appendix X

6 Appendix XI Appendix XII

7 List of Figures Figure 2. 1 Tomato prices in France in EUR/ 100kg Figure 2. 2 Various demand elasticities Figure 2. 3 Various supply elasticities Figure 3. 1 Fresh vegetable production by Member State in Figure 3. 2 Fresh vegetable holdings by Member State Figure 3. 3 Percentage distribution of area and holdings, by class of area size in Figure 3. 4 Value of vegetable production by Member State in Figure 3. 5 Value of EU vegetable production by Member State Figure 3. 6 Main tomato producers by Member State in Figure 3. 7 EU27 Import structure of fresh vegetables Figure 3. 8 Extra-EU vegetable imports in EUR million Figure 3. 9 Extra-EU vegetable exports in EUR million Figure EU27 Export structure of fresh vegetables Figure Organisation rate F&V sector by Member State Figure 4. 1 Maximum storage time by vegetable in weeks Figure 4. 2 Seasonal price pattern of peppers from Turkey Figure 4. 3 Seasonal price pattern of garlic from France Figure 4. 4 Seasonal price pattern of tomatoes from Italy Figure 4. 5 Annual price amplitudes of tomatoes from Italy Figure 4. 6 Monthly price amplitudes of tomatoes from Italy Figure 4. 7 Significance of relation between R 2 and maximum storage time Figure 4. 8 Bivariate graph R 2 values of model (1) Figure 4. 9 Number of significant coefficients by max. storage time Figure Step functions Figure Estimated R-squares of model (1) Figure Estimation results regression (5) Figure Quarterly seasonal price analysis

8 List of Tables Table 2. 1 Division of vegetables Table 4. 1 Storability of vegetables Table 4. 2 Relative perishability Table 4. 3 Descriptive statistics data Table 4. 4 Vegetables ranked according to descending average price Table 4. 5 Vegetables ranked according to the range (descending) Table 4. 6 Vegetables ranked according to the coefficient of variation (descending) Table 4. 7 Monthly price amplitudes (by year) of tomatoes from Italy Table 4. 8 Step function analysis S Table 4. 9 Step function analysis S Table Step function analysis S Table Step function analysis S Table Quarterly price stability Table Price stability across vegetables Table Price stability across years Table Number of significant coefficients model (1) Table Vegetables ranked according to their R-square (descending) List of equations Equation 2. 1 Demand elasticity Equation 2. 2 Cross price elasticity Equation 2. 3 Supply elasticity

9 1 Introduction 1.1 Problem definition The vegetables sector in the European Union accounts for 10% of the total agricultural output value. The most important vegetables, in terms of volume harvested, are tomatoes, carrots and onions (European Parliament, 2015). Most of the EU s vegetable production value is consumed domestically and only 7% of the production value is exported outside the EU (European Commission, 2014). The intra-eu trade in vegetables is large. 80% of EU s vegetable production value is exported to other EU Member States (CBI, 2015). Vegetable prices show a large seasonal variation in the EU (European Commission, 2016). For example, producer price of tomatoes in France vary from 2.03 EUR/kg at its maximum to 0.82 EUR/kg at its minimum in This means that tomato prices can be multiplied or divided by factor 2.5 during the season. For the same commodity and the same year, this factor was 3.75 in the Netherlands. These price variations have impacts for both consumer and producers. Consumers may pay roughly four times as much when there is a price peak compared to price valleys. This may not be a problem to wealthy people, but these price variations can be a problem for people with a limited budget. People with a limited budget may chose unhealthy and cheap food instead of expensive vegetables. Prices of agricultural products, of which vegetables is part of, vary considerably more over time than prices of most industrial products (Tomek and Kaiser, 2014). The most common regularity that agricultural product prices exhibit is a seasonal pattern, which is characterized by a low price during harvest and a higher price during post-harvest. Imports and exports must be taken into account when analysing seasonal price patterns. Imports can cause domestic vegetable prices to drop, whereas exports can cause domestic vegetable prices to rise. But there are still seasonal price patterns, characterized by price peaks and price valleys. Seasonal price patterns are also determined by the storability of vegetables. Vegetables can be roughly divided into either storable or perishable vegetables. Storable vegetables can be stored over a long period after the moment of harvest, whereas perishable vegetables have a limited lifetime after the moment of harvest. Perishable vegetables thus must be consumed rather quick before they are deteriorated. The seasonal price patterns of storable vegetables differ from the seasonal price patterns of perishable vegetables. This can mainly be explained by the fact that storable vegetables can be stored and sold at any time, whereas perishable vegetables have to be sold quick after harvest to preserve a good quality of the vegetable. An implication of this is that prices of perishable vegetables will decrease when there is a production peak, and prices will increase when there is a drop in production. Prices of storable vegetables can be smoothened by storing vegetables. Storable vegetables can be stored when prices are low (and high supply) and thereafter sold when prices are high (and low supply). This mechanism of storing and selling storable vegetables causes seasonal price patterns of storable vegetables to be more stable compared to the seasonal price patterns of perishable vegetables. The price peaks and price valleys are therefore generally smaller for storable vegetables. 9

10 The role of seasonality in agricultural prices has been discussed in agricultural economics. However, the amount of empirical studies related to seasonality in vegetable prices is quite limited. A few authors have studied seasonality with respect to agricultural product prices, like Amikuzuno and von Cramon- Taubadel (2012); Noonari et al. (2015); Arnade and Vocke (2016). To conclude, there is no empirical study on seasonal price patterns of vegetables in industrialized countries, regions or areas like the EU. This research fills this knowledge gap. It is relevant for producers, because this research contributes to the understanding how seasonal patterns of vegetable prices behave in the EU. Insights in seasonal price patterns of vegetables provide them information about how prices behave during the production cycle. This information can help the producers to decide when contracts with wholesalers are favourable. Concluding contracts may be a good option when there is large price variation in a specific part of the season, which causes price uncertainty. Contracts can be a solution to deal with this price uncertainty. Besides the potential benefit for producers, there is also potential benefit for consumers. Especially for consumers with a limited budget, it may be beneficial to gain insight in the seasonal price patterns, because this can help them to decide what vegetable to buy in which period in order to save money as much as possible. The point of interest in this thesis is whether seasonal price patterns in the EU vegetable market exist, and how stable and uniform they are over time and across vegetables. 1.2 Research objective The objective of this thesis is to analyse seasonal patterns of vegetable prices in the EU. 1.3 Research questions 1. What are the potential determinants of seasonal patterns in vegetable prices? 2. What is the structure of the EU vegetables market? 3. What is the empirical evidence? 1.4 Theoretical framework According to Tomek and Kaiser (2014), a seasonal price pattern has a fixed period with fixed seasonal economic conditions, which implies that the amplitude of prices would be constant from year to year. The fixed period is a whole production cycle, which is often twelve months (Noonari et al. (2015); Tomek and Kaiser, 2014). The fixed economic conditions imply that storage costs remain constant during the production cycle of twelve months. The price amplitude is a very important concept in this thesis, because the price amplitude reflects seasonality in seasonal price patterns. The price amplitude is the maximum price minus the minimum price in a given period. The production of agricultural products, and consequently the agricultural price formation, follows a seasonal pattern. One of the explanations is that relative changes in supply and demand cause seasonal price variation, which results in a price amplitude, or seasonality. 10

11 Seasonal price patterns are not the same for every agricultural commodity, as the unique marketing conditions, that belong to a certain product, influence the impact of changing supply and demand on the seasonality of producer prices (Plattner et al., 2014). Another factor that affects seasonal price patterns of vegetables is the extent of storability that belongs to a certain vegetable (Tomek and Kaiser, 2014). 1.5 Methodology The first and second research question are answered by conducting a literature study. Books and scientific articles are consulted to answer the first research question. Official EU documents are the main information source that are consulted to answer the second research question. Furthermore, scientific articles also served as an information source to answer the second research question. The methodology used for the empirical analysis are multiple dummy time series models. Three models are constructed to analyse the seasonal price patterns of the vegetables included in this thesis. The first model analyses whether the average price in each of the 11 months significantly differs from the average price in January. This is done for each of the 24 time series. Step functions are used to graphically present the significance of the beta coefficients of the model. The benefit of step functions is that one can easily observe the course of the seasonal price pattern. It is easily observable if there are clear price peaks or price valleys. The first model is slightly adjusted to test for quarterly price stability by means of F-tests. The original model is a model for each of the 24 time series, whereas the model for the quarterly price stability is a model for each of the 16 vegetables. The second model is constructed to analyse whether there is price stability across vegetables. This model indicates deviations of seasonal price patterns with respect to the seasonal price pattern of the reference vegetable, which is mushrooms. Note that this model distinguishes between vegetables, instead of different time series. This implies that some vegetables, such as tomatoes, cover represent more than one time series. This is because there are 24 time series, while there are only 16 different vegetables. The third model analyses the price stability of vegetables over time. This model indicates whether there are significant price changes over the years. This gives information on the evolution of the seasonal price pattern. 1.6 Data Two datasets are used in this thesis. One dataset contains weekly data and is supplied by the Zentrale Markt- und Preisinformationen GmbH (ZMP). The other dataset is supplied by Agrarmarkt Informations- Gesellschaft (AMI). This dataset also contains weekly price data. These German agencies collect data about agricultural markets. ZMP and AMI have collected price data for a wide variety of vegetables, which is used as a source for this research. The dataset from AMI contain data from , whereas the dataset from ZMP contain data from The dataset of AMI contains more than 800 observations for their selected vegetables, whereas the dataset of ZMP contains around 450 observations for their selected vegetables. The two datasets deliver, in total, 24 different time series. Note that the number of time series is not equal to the number of vegetables, because there are only 16 vegetables included. This implies that some vegetables have more than one time series. 11

12 1.7 Content overview The first chapter of this thesis is an introduction to the topic of this thesis. This introductory chapter gives the problem statements, background information, the research objective and the research questions. The second chapter is a theoretical chapter in which seasonality in vegetable markets is explained. The determinants of seasonality are presented, as well as definitions of seasonality. Furthermore, the concept of storability is introduced. The third chapter gives an overview of the EU vegetables market. Production, consumption, trade is the content of this chapter. The fourth chapter presents the empirical evidence of this research. First, an analysis of the data is presented, in which basic statistics are given. Second, the dummy time series models are presented. The empirical evidence, with respect to seasonality and stability of the seasonal price patterns, is presented in the last part of chapter four. Chapter 5 presents the conclusions of this thesis in which a reference is made to the research questions, and chapter 6 presents a discussion in which a reflection is made on the results, as well as the connection between the theory chapter and the empirical evidence in this thesis. 12

13 2 Determinants of seasonal patterns in vegetable prices 2.1 Definitions This chapter outlines the central theoretical concepts of seasonality and storability. Subchapter 2.1 provides an overview of the existing definitions, and subchapter 2.2 provides an explanation of the concepts that are mentioned in subchapter2.1. In the existing literature vegetables is included in the term agricultural products, since vegetables are agricultural products. Besides vegetables, agricultural products do also refer to fruits, as well as every food product produced in the agricultural sector. The term agricultural products thus serves as a kind of umbrella term for the abovementioned food products Seasonality The production of vegetables is often subject to seasonality, which affects prices in case of perishable vegetables. But what does the term seasonality exactly mean? This section provides some definitions of seasonality. Tomek and Kaiser (2014) argue that prices of agricultural products are typically dependent on changes linked with seasonal, cyclical, trend and random factors. However, the most observed factor in agricultural product prices is a seasonal pattern. The cyclical and trend factor do influence price patterns of EU vegetable prices. Therefore, these concepts are briefly described in this section. Tomek and Kaiser (2014) define seasonal price behaviour as a systematic pattern that occurs within a year (p. 169). This is a more general definition. The authors outline the causes of seasonality, which is mainly changes in supply, and to a less extent changes in demand. The definition of seasonality by Tomek and Kaiser has three characteristics. First, the seasonal price pattern has a fixed period of 12 months (p. 174). Second, the seasonal price pattern assumes fixed economic conditions within that period of 12 months. That is, it assumes that storage costs remain the same for the period of 12 months. Third, based on the previous characteristics, the amplitude of agricultural product prices would remain constant over time. Figure 2. 1 Tomato prices in France in EUR/ 100kg (European Commission, 2017) 13

14 The black dotted line represents the monthly maximum price within a given time period of 5 years. The black stripe line represents the monthly minimum price within the same given period. The green line reflects the monthly average price during this given period of 5 years. Figure 2.1 presents the concept of the amplitude, which is used in the definition of seasonality by Tomek and Kaiser. The amplitude, which is represented by the arrow, in figure 2.1 applies to the year The amplitude is basically the maximum minus the minimum price in a given period, which is a whole year in figure 2.1. Assuming that storage costs are the same throughout the year, the amplitude reflects the seasonality in the figure. Figure 2.1 tells that there is a large price amplitude in 2016 for tomatoes originating from France. However, the large amplitude does not only apply to 2016, as the monthly 5 year average price also shows a large amplitude, which amounts to (223-85) = 138 EUR/ 100 kg Noonari et al. (2015) define seasonality as a phenomenon that occurs over one production cycle for crops, which is generally twelve months (p.3). They also make a distinction between seasonal, cyclical and trend factors when considering crop price behaviour. They argue that the seasonal factor a special form of a cycle is, which is defined as a continuous and self-sustaining price pattern which can occur over any length of time (p.3). Tomek and Kaiser (2014) define the cyclical factor as a price pattern that repeats itself over a time period that is longer than 1 year (p.177). This cycle has two characteristics. First, when a cycle is a true cycle, it is self-energizing, and not driven by other factors. Second, a cycle has a fixed period and a fixed amplitude, which is matching with the second characteristic of seasonal price patterns. In fact, cyclical price behaviour is only determined by the law of supply and demand. This is an endogenous process in the sense that no other factors are involved in determining the price behaviour. However, Tomek and Kaiser (2014) stress that cyclical price behaviour for agricultural products is more complex than only the law of demand and supply. Exogenous disturbances as droughts reduce supply and raises prices. Cyclical price behaviour neglects these exogenous disturbances. The trend factor is defined as persistent and systematic upward (or downward) movements in economic variables (Tomek and Kaiser, 2014 (p.185)). These movements in economic variables are called deterministic trends. An example is a regression with a dependent variable and an independent variable. The coefficient of the independent variable indicates a positive (upward) of negative (downward) trend. This trend factor can shift agricultural product prices upward or downward, but it does not affect the amplitude and so the seasonality factor. In case of perishable agricultural products, Hudson defines seasonality as a type of trend that occurs in agricultural products that emanates, at least in part, from the fact that agricultural production is a biological process (Hudson, 2007 (p.111)). He makes a distinction between the times when crops are harvested, and times when crops are being sold. The period between harvest time and consumption causes seasonality. In case of storable products that are produced seasonally, there is a pattern of prices over time, which is called normal backwardation (Hudson, 2007). This pattern appears when the price of a certain product, after harvest time, increases until the new harvest. The price of storable products reaches its maximum right before the harvest, and its minimum just after harvest. 14

15 This is a repetitive pattern that occurs from harvest to harvest. Storable and perishable products do have another amplitude in prices. Perishable products are subject to the actual market conditions, in which supply and demand determine the price. When the new harvest enters the market, there is more supply than demand, which results in low prices. Just before the new harvest, demand exceeds supply, which results in a high price. This results in a large price amplitude throughout the year for perishable products. Contrastingly, storable products are harvested at any point in time, but these products can enter the market when the producer wants to sell his product. The profit maximizing producer sells his product when demand exceeds supply, in order to receive a high price. Therefore, storable products can be sold in such a way that supply continuously meets demand. This results in a smaller price amplitude for storable products. The definitions that are given do support each other quite well. Almost all definitions attach a limited time horizon related to seasonality. 12 months is the most used time horizon, but other definitions use a time horizon that is described by the production cycle (i.e. from harvest to harvest). Harvest to harvest time horizons may deviate from a 12 months production cycle. Another similarity is that all definitions indicate that seasonality is driven by cyclical changes in supply and demand. These changes in supply and demand exhibit the feature that they are predictable. Thus, exogenous shocks as sudden changes in supply due to droughts, diseases are not included in explaining seasonality. One has to take into account that a peak or low in vegetable prices is not necessarily the result of seasonality, but it can also be the result such an exogenous shock. There are also differences between the given definitions. Noonari et al. (2015) argue that the seasonal factor of agricultural prices a special form of a cycle is, whereas Hudson (2007) argues that seasonality a type of trend is. Thus, Noonari et al. (2015) links seasonality to a self-energizing process (i.e. an endogenous process of supply and demand), and Hudson (2007) links seasonality to the biological process of agricultural production. The definition of seasonality by Tomek and Kaiser is used for this thesis, because this definition is the most specific, and it captures the amplitude of agricultural prices, which is a main concept in this thesis Storability Storability indicates the extent to which agricultural products can be preserved over time. The extent of storability can be explained by the rate of deterioration. Perishable vegetables deteriorate faster than storable vegetables. Although both perishable and storable agricultural products ultimately deteriorate (Kohls and Uhl, 2002), there is a difference. Perishable products are generally sold and consumed immediately after harvest, whereas storable products could be sold and consumed directly after harvest, but they can also be stored and sold at another time. When an agricultural product deteriorates, it loses value and that is the reason why perishable products are generally sold and consumed immediately after harvest, because the deterioration starts after harvest. As Kohls and Uhl outlined, storable products do also deteriorate in the end, but the process of deterioration of these products goes much slower. 15

16 Table 2. 1 provides an overview on the vegetables that are analysed, and to what category they belong. That is, vegetables can be either storable or perishable. Vegetable Beetroot a Carrot b Chicory Cucumbers Eggplant Garlic Horseradish Lamb s lettuce d Leek Lettuce Mushroom Onion Oyster mushroom e Peppers Tomatoes White cabbage f Table 2. 1 Category Storable Storable Perishable Perishable Perishable Storable Storable Perishable Storable Perishable Perishable Storable Perishable Perishable Perishable Perishable Division of vegetables (simplified version of Table 4. 1) Source: author Producers of perishable vegetables have no choice to when to sell their vegetable products. The producer of perishable vegetables is forced to sell his vegetables after harvest, because perishable vegetables start to deteriorate after harvest, which results in a loss of value. In order to earn money, the producer of perishable vegetables is forced to sell his products immediately after harvest. It could occur that supply of perishable vegetables enters the market more or less simultaneously when many producers of perishable vegetables supply their product. Supply exceeds demand in this situation, which results in a low price. This does not need to occur in case of storable vegetables, because owners of storable vegetables can decide to store their vegetables, and sell it another time. Futrell and Wisner (1982) argue that price movements of perishable vegetables are similar to those for storable vegetables, but the price movements for perishable vegetables may be greater. The price movements are reflected in the price amplitude of both perishable and storable vegetables. The price amplitude of perishable vegetables thus may be greater than vegetables. There are four distinctive forms of food storage, which are used for various purposes (Hudson, 2007). These forms of food storage do apply to storable vegetables. All these four forms of food storage relate to seasonality, because food storage is there to balance demand and supply, which would theoretically result in a stable price. This stable price leads to a lower amplitude, which in fact is the seasonality. The first form of food storage is called working inventory, which is necessary for an efficient market. These stocks make sure that there is a continuous supply of agricultural products, which prevents supply disruptions. Thus, this form of storage is meant to continuously meet demand. 16

17 The second form of food storage is seasonal food stocks. These stocks are being maintained, because some agricultural products are harvested in a short time, while these products are consumed over the whole year. These products are therefore stored in order to be able to supply these products on a yearround basis. The third form of food storage is carryover stocks. This form of food storage refers to the amount of agricultural products that has not been sold last year. The annual production of agricultural products is not always equal to the annual consumption. In case of a surplus, the amount from last year will be carried over to the next marketing season. The fourth form of food storage is called buffer food stocks. This form of storage is intended to balance agricultural product supplies with demand over the long run, especially between countries with a food shortage and a food deficit. 2.2 Determinants of seasonality This section deals with the main drivers of seasonal patterns in vegetable prices, which are changes in supply and demand. Furthermore, the concept of storability is discussed, and how this contributes to seasonal patterns in vegetable prices Demand Demand and supply together determine vegetable prices. This section presents demand characteristics in both the short and long run. Although seasonality is mainly caused by the supply side, demand is a minor determinant of seasonality (Tomek and Kaiser, 2014). They outline that this seasonal price variation, due to changes in demand, could be caused by increased or decreased demand during holidays and the seasons within a year. Increased demand leads to higher prices, which may increase the price amplitude (see 2.1.1). Therefore, short-term demand increases may lead to a larger seasonality factor in prices. In case of most agricultural products, it holds that demand is more stable than supply (Ramirez, 2009). He further argues that exogenous demand shifters like demographic and income changes usually show a smooth and gradual trend, which implies that demand does not change significantly with changes in demography and income. This is in accordance with the view of Schnepf (2006), who also argues that the long-run changes of population and income occur slowly and in accordance with well-known behavioural patterns. In short, long-term exogenous demand shifters do change demand and prices, but they occur slowly and predictable. These demand shifters do not determine short-run vegetable prices, because vegetable prices are determined by actual demand and supply. Long-term exogenous demand shifters are therefore not taken into account in seasonality analysis of vegetable prices. The impact of demand and supply shift depends on the elasticities that characterize demand and supply (Ramirez, 2009). Demand elasticity is therefore an important concept, because the demand elasticity of vegetables has an impact on vegetable prices, which may also affect the price amplitude or seasonality of vegetable prices. Hudson (2007) defines demand elasticity as the percentage change in the quantity demanded given a percentage change in own price (p.21). 17

18 The following formula represents the demand elasticity: d = % change in quantity demanded % change in price = % Q % P Equation 2. 1 Demand elasticity Figure 2. 2 presents various demands elasticities graphically. The range from the price elasticity coefficient ranges from zero to minus infinity. This range can be divided into three parts (Tomek and Kaiser, 2014). The first part represents an absolute value of the demand elasticity coefficient larger than one, then demand is elastic. This is indicated with a red line in Figure An elastic demand means that the percentage change in quantity demanded is greater than the corresponding change in price. A horizontal demand curve represents a perfect elastic demand curve. The second part represents an absolute value of the demand elasticity coefficient lower than one, then demand is inelastic. This is indicated with a green line in Figure An inelastic demand means that the percentage change in quantity demanded is less than the corresponding percentage change in price. A vertical demand curve, corresponding with a demand elasticity of zero, represents a perfect inelastic demand curve. The third part represents an absolute value of the demand elasticity coefficient which is exactly one. In this case, the percentage change in quantity demanded exactly equals the corresponding percentage change in price. Unit elasticity is measured at the intersection point of the three lines in Figure Figure 2. 2 Various demand elasticities (Source: author) 18

19 Schnepf (2006) argues that demand elasticity reflects the willingness of a consumer to change consumption when prices of a particular product rises or falls. He further argues that consumers consider the own price elasticity, as well as cross price elasticity of both complementary and substitute products in their decision-making process what products to buy. Cross-price elasticity of demand measures how the quantity demanded of one product responds to changes in the price of another product (Tomek and Kaiser, 2014). The following formula represents the cross-price elasticity of demand: c = % change in quantity demanded of good 1 % change in price of good 2 Equation 2. 2 Cross price elasticity A consumer s willingness to substitute, when a product price rises, depends on the number of substitutes, the importance of the product, which is measured by the share of consumer s expenditure, and the strength of consumer s tastes and preferences. Demand for agricultural products is characterized by inelasticity (Noonari et al (2015), Kohls and Uhl (2002b)). Demand inelasticity of agricultural products has two implications. First, the quantity demanded does respond very much with a change in price. But when demand changes, the related price rises much more than the related demand change. In figure 2.2, inelasticity of demand means an almost vertical demand curve. When the demand curve is almost vertical, the demanded quantity does not change that much in response to the corresponding price change. But if the demanded quantity changes only a little bit, the price change is larger than the change in the corresponding demanded quantity. However, the quantity demanded does not change suddenly, because agricultural products are primary goods (i.e. basic necessities of life), which implies that these products will be bought anyway Supply Prices of agricultural products are mainly determined by seasonal supply (Borovkova and Geman (2006); Noonari et al (2015)). David-Benz et al. (2006) state that, according to vegetable producers, price variations are generally caused by instability in supply. They even argue that empirical data proves the assertion that seasonal price patterns results from seasonal supply. These views are in line with Tomek and Kaiser (2014), who also argue that the main source of seasonality is caused by the supply side of agricultural products. Jayaramu (2015) also finds that deviations of actual production from expected production can have a significant effect on seasonal price patterns. In the literature, there is a consensus that seasonality is mainly caused by the supply side. Demand for vegetables is a given. In other words, it cannot be regulated. However, the supply of vegetables can be regulated to a certain extent. Especially in case of storable vegetables, the supply of vegetables can be determined artificially. In case of perishable vegetables, the extent to which supply can be regulated is much lower. Perishable vegetables are sold after harvest, and are not stored. Supply levels of perishable vegetables can be regulated in such a way that producers can decide what vegetable and how much to produce before a new production cycle starts. In short, supply levels of vegetables can be regulated to a certain extent, especially in case of storable vegetables. The price amplitude of vegetables can be influenced by determining the level of supply. 19

20 This influence is larger in case of storable vegetables, because supply of these vegetables can be regulated in the short term, whereas this is not possible for perishable vegetables. Supply is the main determinant of seasonality in vegetable prices. It is therefore important to analyse the determinants of supply, because supply determines seasonality. Factors that determine of shift supply do indirectly cause seasonality in vegetable prices. In the remainder of 2.2.2, supply shifters are outlined, as well as the concept of supply elasticity. Figure 2. 3 Various supply elasticities (Source: author) Supply elasticity reflects the percentage change in quantity supplied as a result of a 1% change in price, while other factors are held constant (Tomek and Kaiser, 2014). The formula of supply elasticity is as follows: s = % change in quantity supplied % change in price = % Qs % P Equation 2. 3 Supply elasticity In the short run, supply is assumed to be fixed, or perfect inelastic. Inelastic supply refers to a supply elasticity coefficient between zero and one, which means that the percentage change in quantity supplied is less than the corresponding percentage change in price. Elastic supply refers to a supply elasticity coefficient larger than one, which means that the percentage change in quantity supplied is larger than the corresponding percentage change in price. Supply elasticity of agricultural products reflects the speed with which new harvest become available in response to a price change of a particular agricultural product (Schnepf, 2006). 20

21 He further argues that supply does not change in the short term, assuming low inventories. However, in the long run supply can increase or decrease as a result of changing the cultivated area. The reason for stable supply in the short term is that once a producer has seeded a crop, there is limited or no ability to change crops. It takes a new production cycle before new crop choices can be made. This is why supply of agricultural products is assumed to be inelastic in the short term. This inelastic supply of agricultural products has an implication for agricultural product prices, because inelasticity of supply indicates that an increase in demand leads to a higher increase in prices. In case of nonstorable, or perishable vegetables, it holds that the supply cannot be changed in the short term, because it takes a new production cycle to make new output decisions. This makes the supply of these vegetables to be fixed, which can also be called perfectly inelastic. Given demand, which is assumed to be more or less stable, price is low when supply is high. Conversely, it holds that agricultural product prices are high when supply is low. In case of storable vegetables, it holds that the total available quantity supply includes both the currents harvest and the stocks available from prior harvests (Hudson, 2007). When the market price for storable vegetables rises, then stockholders are incentivized to take out stocks of storage and sell them. When stockholders are taking out stocks of storage, the total amount of stocks decreases until there is no stocks left. In case of no stocks, total supply is perfectly inelastic. Vegetable prices can no longer be influenced by stocks in this case. Unpredictable events, such as weather conditions, diseases do change supply, which affects the price of agricultural products. Unpredictable means that the vegetable producer cannot control these events. This makes the price variation of vegetables to have a natural character, because nature can affect supply of vegetables. In section it is explained what is meant by cyclical price behaviour. Changes in cyclical behaviour can be explained by taking an example of unpredictable events. Unpredictable events, such as a drought, decreases supply and increases the price of agricultural product prices. Next to unpredictable events that change supply, there are also events going on that are predictable. At least, the producer is aware of output results of last year, which will influence decisions for the following crop cycle. Producer output decisions are made based on output results of last year. These developments include output prices of the produced crops, but the relationship between input and output is also one to be mentioned. Changes in prices of inputs that are used to grow crops do affect supply (Hudson, 2007). When inputs for a certain crop rises significantly, it may be better to produce another crop that uses less expensive inputs. For instance, when costs of labour rise, it may be wise to grow a crop that requires a little labour. However, Tomek and Kaiser (2014) argue that higher input prices shift the supply curve to the left. This results in higher agricultural product prices. Thus, costs and benefits will be analysed before a producer will choose to grow another crop. Risk also plays an important role in the decision-making process of producers concerning output (Hudson, 2007). Output decisions are being made by both looking at the past, as well as looking ahead. This is, prices of last year will be a source that will be consulted in the output decision making process, but the producer will also use as much as possible information to form expectations of future prices. Producers form an expectation of what the price is likely to be for a certain product next year. 21

22 If the price for this certain product is highly variable, then it will be hard for the producer to form an expectation of the price, which will make output decisions very hard. When the producer is known to be risk averse, he will choose to produce another crop. Institutional factors do affect output levels as well (Hudson, 2007). Supply control mechanisms have played an important role, but the importance of these mechanisms has decreased. In contrast to supply control mechanisms, environmental programs and regulations imposed by governments have gained importance. There are regulations that take land out of production for the sake of nature conservation. Other regulations are in the scope of pesticide and herbicide use, nonpoint-source pollution and other environmental regulations. These environmental regulations do affect land availability, cropping patterns, costs and therefore market supply (Hudson, 2007). Timing of harvest plays also influences the seasonal price pattern of agricultural products (Tomek and Kaiser, 2014). This is specifically the case for storable products, because storable agricultural products are stored to be able to meet demand throughout the year. When potatoes are being stored, it is crucial to determine the timing of the new harvest. If potato stocks run out of stock, prices will rise. If there is still enough stock, the prices will drop when a potato producer decides to harvest. It is therefore important to take the amount of stock into account when deciding when to harvest. Technology plays also a role during the seasonal patterns of agricultural product prices (Tomek and Kaiser, 2014). Technological improvements are defined as something that enables firms to produce more output with the same quantity of inputs as previously. Advances in technology make also sure that agricultural output can be supplied over the year, whereas in the past the same agricultural output was used to be a seasonal product. Advances in technology thus extend the period in which agricultural output can be supplied. However, one must take into account that advances in technology tend to be a long-run process (Hudson, 2007). This indicates that advances in technology cannot explain supply shocks in the short-run. In the long-run however, advances in technology might explain gradual increases in supply due to increasing productivity of machines, mechanisation and more knowledge International trade Another notion is international trade. International trade contributes to domestic price variations in agricultural products when demand and supply change on the world market (Kohls and Uhl, 2002b). International trade does therefore also influence the shape of price patterns of vegetables. Existing price amplitudes may thus be partly explained by an international trade factor. In particular, when a specific country is a large exporter of a certain agricultural product, then demand changes for this certain agricultural product on the world market can have a destabilizing effect on the domestic prices of this product. Trade distortions like import bans impede international trade, which leads to more supply in the exporting country. Subsequently, this results in lower product prices in the exporting country. Although most of EU s produced vegetables (93% production value) are consumed within the EU (European Commission, 2014), there may be individual EU producers of vegetables with a contract with trading partners originating from a country outside the EU. To these producers it makes sense to take international trade into account. 22

23 2.2.4 Storing vegetables The feature of storability of storable agricultural products enables producers to allocate the supply over the year, in order to meet continuous demand (Tomek and Kaiser, 2014). When agricultural food products have the ability of being stored, then this ability affects the seasonal pattern of agricultural product prices. If there is no storage possible at all, then agricultural product prices are fully determined by supply and demand at the moment of entering the market. However, if storage is possible, a producer may decide to store a product after harvest, and sell it at another time. Thus, storage of agricultural products enables producers to choose a selling time at which they think they would get the highest price. Producers of storable vegetables can partly determine the price, because they are able to decide how much to sell, and how much to store. In other words, they are able to control the supply level to a certain extent. Decisions concerning the supply levels do influence the price of storable vegetables, which also affects the price amplitude. When storable vegetables are harvested, it does not automatically mean that they are harvested. Storage of storable vegetables depend on the actual market conditions, which are demand and supply, but also storage costs. In the beginning of this section, it was outlined that storage is there to meet demand throughout the year, but it is more complex than that. The remainder of this section outlines the concept of storage costs. Noonari et al. (2015) outline that crop prices are lowest at harvest, and then prices start to increase until the new harvest. The rate by which prices rise is exactly equal to the marginal cost of storage per unit of time, assuming an annually produced crop that is storable and a competitive market. (Tomek and Kaiser, 2014). Agricultural product prices are lowest at harvest, because then the available supply exceeds the short-term demand (Kohls and Uhl, 2002b). After harvest, there is an upward trend in prices, which is called a post-harvest rally. Noonari et al (2015) explain this process by stating that supply of a crop is fixed, and this supply will be used up gradually by consumption. The storage costs between the current period and the next period can be divided into four parts (Tomek and Kaiser, 2014). The first part represents conventional costs of inputs, for instance investments in warehouse, wages and insurance premiums. The second part represents the opportunity costs related to the current value of the stock, which depends on two factors; the product price, as well as the interest rate. Costs in the first part, such as insurance premiums, investment costs and wages are relatively stable from year to year, whereas costs in the second part, such as opportunity costs based on interest rates and product prices are not stable. The third part of storage costs can be considered as a type of benefit, which is called convenience yield of stock holdings. This means that the expected price for the new harvest is below the current price when the new harvest is coming. One may wonder if this implies that the return to storage is negative in this situation, but as Tomek and Kaiser (2014) argue, it is presumable that merchants and processors save at least part of their inventories from one year to the other, because they obtain a benefit from avoiding a situation in which there are no inventories anymore. The fourth part of storage costs is linked to uncertainty related to the expected future price. In a perfect world with perfect information, the product price for next period is known. However, in the real world the presumption of perfect information does not hold, which implies that uncertainty arises. The realized price change in the storage interval may not be sufficient to cover the storage costs. 23

24 Therefore, uncertainty and price risk is a type of cost. All these parts have its influence on the price formation of agricultural products, which logically also contribute to the price patterns of agricultural products. One has to take into account that a certain seasonal price pattern does not automatically prevail every year, since the agricultural product prices change, as well as the cost of storage. A vegetable producer may decide not to store when storage costs are too high and vegetable prices are not expected to rise in the future. In this situation, actual demand and supply determine the price. Vegetable prices cannot be manipulated anymore by artificially determining supply levels. This producer of storable vegetables now acts as a producer of perishable vegetables, because his storable vegetables are not stored, but supplied after harvest. This results in larger price movements as Futrell and Wisner (1982) argue. Storage costs can therefore indirectly affect vegetable prices, and the seasonality of it. 24

25 3 Structure of EU vegetables market The aim of this chapter is to introduce the main characteristics of the EU vegetables market. Furthermore, facts and figures are presented to visualize these main characteristics. In this chapter, there are some references to the EU fruit and vegetables (F&V) sector. In official EU documents, fruit and vegetables are often captured in one policy. EU policies concerning the fruit and vegetables sector thus apply to EU vegetable producers. Subchapter 3.1 presents the key characteristics concerning the structure of EU vegetable farms. Subchapter 3.2 presents facts and figures on EU vegetable production. Tomatoes are highlighted in this subchapter, because this vegetable represents the largest production value of all EU produced vegetables. Subchapter 3.3 presents information concerning vegetable consumption in the EU. Furthermore, the importance of vegetable consumption is emphasized. Subchapter 3.4 presents facts and figures on EU vegetable trade, which includes both intra, as well as extra EU trade. Subchapter 3.5 presents the key information of the Common Agricultural Policy that applies to the EU fruit and vegetables sector. Subchapter 3.6 presents information on EU vegetable prices. This subchapter further contains a short analysis on different seasonal price patterns of some vegetables from the used databases. 3.1 Farm structure This section provides information on the structure of EU vegetable farms. First, information on the production area per Member State is presented. Second, vegetable farms per Member State is given. Third, a figure is presented that visualizes and categorizes the production area per vegetable farms. In the end of this section, there are some links between the figures presented. Almost 2.2 million hectares in the EU were used for fresh vegetable production in 2015 (Eurostat, 2016b). 2.2 million hectares represents 1.9% of EU s total arable land. Figure 3. 1 presents the fresh vegetable production area by Member States in Figure 3. 1 Fresh vegetable production by Member State in 2015 (Eurostat, 2017a) Figure 3. 1 indicates that almost half of the total EU fresh vegetable production is represented by only three Member States, which are Italy, Spain and Poland. 25

26 Almost 920,000 vegetable farms produced fresh vegetables in the EU in 2013, which was 12% of all European farms with an arable area (Eurostat, 2016b). Nearly half of these vegetable farms originate from only three countries, which are Romania, Poland and Spain. 15 Member States accounted for shares less than 1% of total EU vegetable farms. Figure 3. 2 Fresh vegetable holdings by Member State (Eurostat, 2017b) The average acreage of the EU vegetable farms was 1.7 ha (Eurostat, 2016b). The largest average acreage was observed in the United Kingdom (17.4 ha), followed by the Netherlands (10.3 ha). Both of these countries are not explicitly presented in Figure 3. 2, which means that these Member States belong to the category of 22 Member States. One can conclude that the amount of vegetable farms in the United Kingdom and the Netherlands is relatively lower than in other Member States, but the average acreage is relatively larger. Figure 3. 3 is presented to illustrate the average acreage per vegetable farm in the EU. Figure 3. 3 Percentage distribution of area and holdings, by class of area size in 2013 (Eurostat, 2017c) 26

27 Remark: melon and strawberry are included in Figure Figure 3. 3 indicates that the majority of the EU vegetable farms (79.5%) has an acreage of less than 1 ha, while only 6.3% of the EU vegetable farms represents 68.3% of the total EU vegetable acreage. This implies that there are a lot of relatively small vegetable farms in the EU, and a small share (6.3%) accounts for large vegetable farms of more than 5 ha. There are links between Figure 3. 1, Figure 3. 2 and Figure One of the remarkable things is that Italy only represents 8.7% of the total EU vegetable farms, but it represents 19.5% of the total EU fresh vegetable production area. Conversely, Romania represents the largest share of total EU fresh vegetable producers in the EU, which amounts to 22.1% of EU s total (Figure 3. 2), but it only represents 7.1% of the total EU vegetable production area. This of course has to do with the production area per vegetable farm. Table 3. 1 presents the facts on production area, holdings, and the production area per holding. This table illustrates the difference in production area per holding between Italy and Romania. The production area per holding is (5.4/0.8) = 7 times higher in Italy. This indicates that the farm structure within the EU can be very different. Italy Romania Area (in hectares) 429, ,200 Holdings 80, ,320 Area per holding (ha/holding) Table 3. 1 Production area per holding of Italy and Romania Source: author 3.2 Production Section 3.2 presents key facts and figures on EU vegetable production. The main EU vegetable producers are presented. Special attention is paid to tomatoes, because of its economic importance (i.e. production value) in the EU. The EU vegetables sector accounts for 13.6% of the total EU agricultural output in production value (Eurostat, 2016a). In 2014, the total value of EU vegetable production was EUR 30.7 billion (Eurostat, 2016b). The largest producing Member States, in terms of production value, were the Netherlands (17.8%), Spain (16.7%) and Italy (16.5%) in These three countries thus produce together half of the total EU vegetable production value. Figure 3. 4 presents the largest vegetable producing Member States in the EU according to their production value. 27

28 Figure 3. 4 Value of vegetable production by Member State in 2014 (Eurostat, 2017d) Figure 3. 5 Value of EU vegetable production by Member State (based on Eurostat data; Eurostat, 2017d) 28

29 The most important vegetables, in terms of quantities, produced in the EU were tomatoes, carrots and onions in the EU. The production of these three vegetables amounted to 16.1, 5.1 and 5.4 million tons in 2010 (European Parliament, 2015). Tomatoes were mainly produced in Italy and Spain. Carrots were a major vegetable output in Poland and the United Kingdom, and Onions were mainly produced in the Netherlands and Spain. The fruit and vegetables (F&V) sector is of great importance in the Mediterranean Member States (European Parliament, 2015). This has to do with the favourable weather conditions to grow fruit and vegetables. Tomatoes is the most important vegetable product in the EU in terms of production value. EU tomato production in 2014 was responsible for EUR 7 billion. This illustrates the importance of the tomato sector in the EU, because it accounts for more than 20% of the total EU vegetable production value. Figure 3. 6 therefore presents the main tomato producers in the EU according to their production value. Figure 3. 6 Main tomato producers by Member State in 2014 (Eurostat, 2017d) This production value is measured at basic prices, that is, prices are measured including subsidies, but without taxes. One can derive from Figure 3. 6 that more than half of the tomato production, measured in economic output, originates from only three Member States, which are Spain, Italy and Poland. 29

30 3.3 Consumption Section 3.3 provides information on the consumption of vegetables. First, the importance of vegetable consumption is stressed. Second, consumer behaviour with respect to vegetable consumption is analysed. Third, differences between consumption figures within the EU is given. A high daily consumption of fruit and vegetables is important, because it is associated with a lower risk to get a major chronic disease, especially the incidence of a cardiovascular disease (Hung et al, 2004). The Consumption Monitor 2008, which is a report by Freshfel Europe, revealed that in half of the Member States, on average, the intake of fruit and vegetables is below or just above 400 grams a day, which is recommended by the World Health Organisation (Freshfel Europe, 2008). Freshfel concludes that there are still important efforts to be made to achieve satisfactory fruit and vegetables consumption figures throughout the EU, which is 400 grams of fruit and vegetables per day. Consumer behaviour is not homogeneous with respect to vegetables within the EU (Kalaitzis et al., 2007). Homogeneous consumer behaviour implies that consumption preferences and patterns are the same among the EU population. Cultural differences, as well as lifestyle differences contribute to heterogeneous consumer behaviour within the EU. This heterogeneity of consumer behaviour can be described by a trend in Northern European countries for fast food, a high concentration of supermarkets and vertical co-ordinated supply chains. In contrast to the Northern Europe, fruit and vegetable markets in Southern Europe are mainly dominated by street markets. That is, vegetables in Northern Europe are mainly sold in supermarkets, whereas vegetables in Southern Europe are mainly sold on street markets. The largest per capita consumption of vegetables is concentrated in de South of the EU (European Parliament, 2011). This is remarkable, because the main EU vegetable producers are located in the South (Figure 3. 4). Both production and consumption is highest in the South of the EU. Further, there is a great difference between per capita consumption of vegetables within the EU, ranging from 74.6 kg/capita/year(bulgaria) to 241 kg/capita/year (Greece). The per capita consumption of vegetables is more than three times as high as in Bulgaria. Romania (151 kg/capita/year) and Poland (130 kg/capita/year) have also reached a consumption level that matches with some Mediterranean Member States. Kalaitzis et al. (2007) expect two factors to influence demand for fresh fruit and vegetables. These two factors include aging and a decrease in household size Aging will probably increase demand for fresh fruit and vegetables, because aged people belong to the highest fruit and vegetable consuming category. 3.4 Trade This subchapter presents the key facts and figures concerning EU vegetable trade. Section presents key information on EU vegetable imports, and presents key information with respect to EU vegetable exports. Section subsequently presents the main export destinations of EU vegetables, and pays presents the origins of EU vegetable imports. Section pays special attention to Intra-EU trade in vegetables. 30

31 3.4.1 Imports from outside the EU With respect to the value of EU fresh vegetable imports, there was an increase by more than 40% to EUR 2 billion in the period Two thirds of the increase in imports consisted of an increase in tomatoes, beans and sweet peppers. Figure 3. 7 summarizes the most important imported fresh vegetables in the EU. Figure 3. 7 EU27 Import structure of fresh vegetables (European Commission, 2012) Tomatoes, beans and sweet peppers jointly account on average for exactly half of the EU import value of fresh vegetables in the period Tomatoes appears to be the main vegetable imported to the EU from third countries. 31

32 Figure 3. 8 Extra-EU vegetable imports in EUR million (based on Eurostat data; Eurostat, 2017e) Exports outside the EU Only 7% of the total EU vegetable production value is exported outside the EU (European Commission, 2014). As mentioned before, tomatoes, carrots and onions are the most important EU vegetables in terms of production quantities. From total EU production of tomatoes and carrots, less than 5% is exported, while 12.5% of the EU production of onions is exported. The main export market for the EU vegetables is Russia, which buys one quarter of the EU fresh vegetable exports with a value of EUR 734 million. 32

33 Figure 3. 9 Extra-EU vegetable exports in EUR million (based on Eurostat data; Eurostat, 2017e) EU fresh vegetable exports was worth EUR 2.1 billion in 2011 (European Commission, 2012). The value of EU fresh vegetable exports has more than doubled since The expansion of EU trade in fresh vegetables is mainly due to an increased export in products as potatoes, tomatoes, onions and sweet peppers. Figure EU27 Export structure of fresh vegetables (European Commission, 2012) 33

34 Potatoes, sweet peppers and tomatoes jointly account for slightly more than half of the EU27 average export value of fresh vegetables in the period Figure 3. 7 shows that tomatoes are the largest imported vegetable, but tomatoes are the third largest exported vegetable, according to Figure Second, beans are an important imported vegetable, but beans do not play a role in vegetable imports (Figure 3. 10). Third, sweet peppers account for roughly the same percentages in both figures. Fourth, since the average EU vegetable imports and vegetable exports in the period are almost the same in value, one could conclude that the import value and export value for sweet peppers is quite similar. Fifth, Figure 3. 7 and Figure both contain three products that account for at least half of the total EU import or export value. In short, there are some important import and export vegetables in terms of value, and the rest of the vegetables account for relatively low shares. Although potatoes are included as a fresh vegetable in Figure 3. 7 and Figure 3. 10, their prices are not analysed in chapter 4, because the used datasets do not contain potato prices. However, sweet peppers, tomatoes, onions, garlic, mushrooms, cucumbers and cabbage lettuce are included in the datasets. Figure 3. 7 and Figure present trade information of some of the vegetables that is analysed in chapter Destinations of EU vegetable exports The main destinations for EU fruit and vegetable exports are Russia, Norway, Switzerland, the US and Japan (European Commission, 2012). Russia bought both fresh and processed vegetables amounting to EUR 886 million in 2011, while the export value of these products was worth EUR 157 million in Exports of fresh and processed vegetables to Russia grew by 465% in 2011 compared to the export value of these products in EU vegetable exports to Russia have grown rapidly over the past decade especially some products as tomatoes and potatoes. For instance, tomato exports have increased from EUR 6 million in 1999 to EUR 117 million in The value of EU potato exports to Russia increased from EUR 7 million to 199 million in the same period. These trade figures indicate the importance of Russia as an export market for EU vegetables. On 7 August 2014, the Russian federation imposed an import ban on fruit and vegetables originating from the EU (European Commission, 2017e). This import ban had significant implications for EU fruit and vegetable exports to Russia. The value of fruit and vegetable exports to Russia declined by more than $ 1.8 billion (Kutlina- Dimitrova, 2015). Switzerland is the second main export destination for EU vegetable exports. The value of fresh and processed EU vegetable exports, including potatoes, was worth EUR 395 million in 2011, whereas these products were worth EUR 249 million in The exports of EU vegetables to Switzerland has grown by 58% in the period The USA is another important export destination for EU vegetables. The value of fresh and processed EU vegetables was worth EUR 343 million, whereas the value for these products was EUR 367 million in The export of fresh and processed EU vegetables to the USA decreased by slightly more than 6%. Although this decline of 6% can be labelled as a small decline, there still is a negative trend, whereas the export of fresh and processed vegetables to Russia and Switzerland grew significantly. Norway was also a major export market for EU vegetables. The value of EU vegetable exports to Norway was worth EUR 222 million in 2011, whereas the value of the same products was worth EUR 76 million in 1999, which is an increase of 192%. 34

35 In other words, the EU vegetable export value was almost threefold in 2011 compared to The value of vegetable exports to Japan was worth EUR 159 million in 2011, whereas the value of the same products was worth EUR 87 million in 1999, which is an increase of 83% Origins of EU vegetable imports Most of the imported fresh vegetables comes from Mediterranean countries as Morocco, Israel, Turkey and Egypt (European Commission, 2012). Import of vegetables from these countries has grown, because these Mediterranean countries benefit from improved market access to the EU due to preferential trade agreements. These preferential trade agreements include tariff reductions (Emlinger et al., 2008). These preferential trade agreements were made to enhance the prosperity in the Mediterranean region. The Mediterranean countries supply a major part of the EU s vegetable imports. However, China is the biggest supplier of the EU s vegetable import in terms of value. The value of vegetable imports from China amounted to EUR 759 million in 2011, whereas the value of these products was EUR 324 million in The value of vegetable imports from China increased by 134%. In other words, the value of vegetable import from China has more than doubled in the period From the Mediterranean countries, Morocco is the biggest supplier of EU vegetable imports. In 2011, vegetable imports from Morocco was worth EUR 566 million, whereas the value of these products was worth EUR 251 million in The value of vegetable imports increased by 125%. Thus, also the value vegetable import from Morocco has more than doubled in the period Turkey was the second largest Mediterranean supplier of EU vegetable imports. The value of EU vegetable imports from Turkey was worth EUR 322 million in 2011, whereas the value of these products was worth EUR 212 million in 1999, which implies an increase of almost 52%. The third biggest Mediterranean supplier of EU vegetable imports is Israel. In 2011, the value of EU vegetable imports from Israel was worth EUR 292 million, whereas the value for the same products was worth EUR 106 million in 1999, which is an increase of almost 175%. The value of EU vegetable imports from Israel has more than doubled in this period. The fourth biggest supplier of EU vegetable imports is Egypt, which exported vegetables to the EU with a value of EUR 268 million in 2011, whereas the value of the same products was worth EUR 94 in The major export destinations for EU vegetables include: Russia, Switzerland, USA, Norway and Japan. The value of EU vegetable exports to these countries increased from EUR 1767 in million in 1999 to EUR 4132 million in 2011, which is an increase of 134%. The value of vegetable exports to most of the major export destinations grew by large percentages. The exception is the value of vegetable exports to the USA, which decreased by slightly more than 6%. With respect to the import statistics, one could conclude that the value of EU vegetable imports has increased by enormous percentages too. In section 3.4.3, some major countries were highlighted. These countries include China, Morocco, Turkey, Israel and Egypt. The import value of vegetables from these countries increased from EUR 1817 million in 1999 to EUR 3627 million in 2011, which is an increase of 100%. Thus, the import value of vegetables doubled in this period. 35

36 The value of EU vegetable exports increased by 134%, and the value of vegetable imports increased by 100%. Although these numbers are similar, both percentages indicate that EU vegetable trade increased significantly in the period In only 13 years, the value of EU vegetable exports and EU vegetable imports doubled. The value of EU vegetable exports did even more than double in this period Intra-EU trade in vegetables EU vegetable production is mainly consumed within the EU. In the period , an average of 7% of the production value of fresh vegetables is exported outside the EU (European Commission, 2014). In 2015, the internal exports of fresh vegetables were worth EUR 12.9 billion, and the extra EU exports of fresh vegetables were worth EUR 1.5 billion (Eurostat, 2016b). In 2014, EU Member States exported fresh vegetables (both intra and extra) with a value of EUR 14.9 billion (CBI, 2015). 80% of the value of EU fresh vegetable exports is exported to other EU Member States. Thus, the value of Intra-EU trade was EUR 11.9 billion. Spain is the EU Member State that exports most. The vegetable exports of Spain are mainly produced within Spain. This seems to be logic, but this is not necessarily the case for so-called trade hubs. Examples of these trade hubs are The Netherlands and Belgium, which are also main EU exporters of fresh vegetables. These Member States do not produce all their vegetable exports within their Member State, but they also import and export vegetables due to their favourable logistic location. These trade hubs import and export vegetables both within the EU, as well as outside the EU. The major destinations of EU vegetable exports within the EU are Germany, United Kingdom and France. The export value of EU vegetables to these countries was worth EUR 7.1 billion in The main products within Intra-EU exports include tomatoes, lettuce and chicory, and cucumbers and gherkins (Eurostat, 2016). The main Intra-EU exporting Member States for tomatoes include The Netherlands, Spain and France. For lettuce and chicory (1 category), the top exporting Member States are Spain, Italy and the Netherlands. For cucumbers and gherkins, the top exporting Member States are Spain, the Netherlands and Belgium. 3.5 Common Market Organisation This subchapter deals with the EU agricultural policies that apply to the EU vegetables sector. First, section outlines the relevant reforms of the Common Market Organisation (CMO) of fruit and vegetables. Section provides an evaluation of the CMO. Objectives of the CMO of fruit and vegetables are presented, followed by a brief evaluation CAP Reforms The EU F&V sector is supported by the EU through the CMO. The CMO was established to remove trade barriers within the EU. The CMO provides EU Member States a unified policy related to agricultural markets and prices (Oskam et al., 2011). The CMO of fruit and vegetables has been subject to reforms in 1996, 2001 and 2007 (Agritrade, 2011). These reforms eventually led to the abandonment of price support in 2007, but there was an increase in direct payments to F&V producers. Before the 2007 CMO reform, price support was a market measure to stabilize prices and income in the F&V market (Dell Aquila and Petriccione, 2012). 36

37 A second result of the reforms was the incorporation of the F&V sector into the Single Payment Scheme (SPS), along with sector-specific measures implemented by Producer Organisations (POs). These POs are organisations of fruit and vegetable producers. These POs group supply in order to strengthen their market position compared to demanders (European Commission, 2017d). A PO has the following three characteristics (Nilsson et al., 2012). o o o A PO is user-owned, because the users of the services of the PO also own the PO. A PO is user-controlled, because the PO owner decide the strategies and policies of the PO. A PO is for user benefit, because all the benefits of the PO is distributed to its users/ owners according to their use. One of the major outcomes of the 2007 CMO reform was that export subsidies were abandoned (Meijerink and Achterbosch, 2013). However, there is still an exception to this result. Export subsidies are still being used when trading with third countries, in order to compensate for the difference between prices in the EU and world prices. Products that are included in this exception are: apples, lemons, oranges, peaches, nectarines, table grapes, tomatoes and various nuts. Another important outcome of the 2007 CMO reform was that POs were even more supported by the EU (European Parliament, 2015). Due to this CMO reform in 2007, POs were incentivized to collaborate with other POs in so-called associations of POs (APOs), and also transnational POs. Besides the tools to prevent and manage market crises, there was an emphasis put on environmental measures Evaluation of the Common Market Organisation The CMO of fruit and vegetables apply to all fruit and vegetables produced in the EU with the exception of potatoes, wine grapes, bananas, sweet corn, beans and peas for fodder, and olives (Meijerink and Achterbosch, 2013). This CMO has four objectives (European Commission, 2017a): 1. A more competitive and market-oriented sector 2. Fewer crisis-related fluctuations in producers income 3. Greater consumption of fruit and vegetables in the EU 4. Increased use of eco-friendly cultivation and production techniques The EU supports operational programmes that will be implemented by POs in order to achieve the first objective of a more competitive and market-oriented F&V sector (European Commission, 2017b). This EU support is needed, because the F&V sector faces an increasing concentration of demand. This development of concentration of demand leads to more market power of fruit and vegetable demanders. As a result, the market power of fruit and vegetable suppliers diminishes. To counter this, POs are there to concentrate supply, which will give POs more market power. 37

38 EU funding is available for crisis prevention and management measures, which is part of the operational programmes of POs, in order to reduce crisis-related fluctuations in producers income (European Commission, 2017c). This EU funding is available for six distinctive actions, which include: 1. Market withdrawal: Taking fruit and vegetables from the market. 2. Green harvesting/ non-harvesting: Green harvesting implies that non-marketable products are grown on a certain production area, before the normal harvest takes place. Non-harvesting implies that fruit and vegetables will not be harvested during the normal production cycle. 3. Promotion and communication: Any action that relates to promotion and communication must be additional to the already existing promotion and communication actions that are already being applied. 4. Training measures: Member States that allow measures as described under 1. and 2. must adopt detailed regulation with respect to the implementation of these measures. 5. Harvest insurance: This type of EU funding is available to help safeguard producers income and cover losses as a result of disasters, climate events, diseases or pest outbreaks. One condition for this EU fund is that fruit and vegetable producers have to be member of a PO. 6. Support for the administrative costs of setting up mutual funds: This type of EU funding is available in the first three years of the mutual funds operation. A distinction is made between Member States that joined the EU after 2003, and Member States that were Member before Despite the intentions from the EU to support the EU F&V producers, there is a persistently low degree or lack of organisation in some MSs (European Commission, 2014, p.13). Especially in some Southern Member States and some Eastern Member States, there is a low degree of organisation in a PO. Figure Organisation rate F&V sector by Member State (European Commission, 2014) 38

39 The organisation rate in Figure reflects the percentage value of marketed products by POs. Eastern Member States as Estonia, Lithuania and Slovenia do not even have recognised POs. Northern Member States as The Netherlands and Belgium show high organisation rates. In 2013, the CMO for fruit and vegetables has undergone a next reform (European Parliament, 2015). This 2013 CMO reform introduced two changes. The first change was that APOs were allowed to set up an operational fund, driven by financial contributions of its member POs, as well as EU financial assistance. The second change was that the set of tools to prevent and manage market crises was extended. 39

40 4 Empirical evidence This chapter presents the empirical evidence concerning the seasonal price patterns of the 24 different time series used in this thesis. Section 4.1 first presents a descriptive analysis of these 24 time series. This section also presents the division of vegetables. That is, there is a distinction made between storable and perishable vegetables. Section 4.2 introduces the econometric model that is used to analyse the seasonal price patterns. The results of this analysis are subsequently presented in section 4.3. Section 4.4 deals with heterogeneity and stability related to the 24 time series. Results of this analysis will be given in this section. 4.1 Descriptive analysis This section provides a description of the data used in this thesis. Section first presents Table 4. 1 in which the used time series are divided into categories, which is either perishable or storable. Then, Table 4. 3 presents key statistics with respect to the time series. At the end of section 4.1.1, a brief analysis of these key statistics is presented to check whether there are clear observable patterns. Section presents a graphical impression of seasonal price patterns of both a perishable, as well as a storable vegetable. Differences between these seasonal price patterns are analysed at the end of this section. Section introduces the concept of annual price amplitudes. An example is presented to illustrate this concept. Section presents the concept of two different monthly price amplitudes. Again, examples are presented to illustrate monthly price amplitudes. Last, section presents key findings from section up to section Data description Table 4. 1 presents the division of vegetables according to their storability as used in this thesis. The first column presents the vegetables, and the second column provides the storage time interval, which serves as an indication for how long a certain vegetable can be stored. These storage time intervals are provided by Gross et al (2016). For each vegetable, the storage time intervals are chosen that belong to the best storage conditions, which are generally the temperature and humidity. Different conditions have different storage time intervals, but for the sake of simplicity it is assumed that the best storage conditions prevail for each vegetable. The third column is the maximum storage time, which is basically the upper bound of the storage time interval. The fourth category provides an indication of the FAO (2017) whether vegetables are storable or perishable. The fifth column in Table 4. 1 presents the division of vegetables according to USAID (USAID, 2008). This USAID report, which is a value chain assessment, makes also a distinction between storable and perishable vegetables. In this report, storable vegetables are labelled as chiefly root crops and perishable vegetables mainly include greenhouse vegetables. Consequently, many vegetables are not specifically mentioned. Any rationale between the distinction between storable and perishable vegetables is lacking. For instance, a definition of greenhouse vegetables is lacking. The report by Gross et al. (2016) is therefore taken as a guideline, because this report provides detailed information about every single vegetable. Column 6 provides the eventual division of vegetables which is used in this thesis. 40

41 Vegetable Storage time interval i Maximum storage time (weeks) FAO g category USAID h category Category in this thesis Beetroot a days 2 weeks Storable - Perishable Carrot b 7-9 months 39 weeks Storable Storable Storable Chicory 2-4 weeks c 4 weeks - - Perishable Cucumbers Max 14 2 weeks - - Perishable days c Eggplant Max 2 2 weeks Perishable - Perishable weeks Garlic >9 months >39 weeks Storable Storable Storable Horseradish weeks - - Storable months c Lamb s Max 4 4 weeks Perishable - Perishable lettuce d weeks Leek 2-3 months c 13 weeks - - Storable Lettuce 2-3 weeks 3 weeks Perishable - Perishable Mushroom 7-9 days 1.5 weeks Perishable - Perishable Onion 1-8 months 35 weeks Storable Storable Storable Oyster 7-9 days 1.5 weeks Perishable - Perishable mushroom e Peppers 2-3 weeks 3 weeks - Perishable Perishable Tomatoes 4-7 days c 1 week - Perishable Perishable White cabbage f 3-6 weeks c 6 weeks Perishable Storable Perishable Table 4. 1 Source: author Storability of vegetables a: Gross et al. (2016) distinguish between topped beets and bunched beets. In this thesis, beetroot is assumed to be bunched beets. b: Gross et al. (2016) distinguish between bunched carrots, mature carrots and immature carrots. In this thesis, carrot is assumed to be mature carrots. c: Source: The University of Maine (2001) d: Lamb s lettuce is not explicitly mentioned by Gross et al. (2016). In this thesis, I assume that Lamb s lettuce has the same storage time and maximum storage time as lettuce. e: Oyster mushroom is not explicitly mentioned by Gross et al. (2016), in this thesis, Oyster mushroom is assumed to have the same storage time and maximum storage time as mushroom. f: White cabbage is not explicitly mentioned by Gross et al. (2016), but in the concluding table from FAO it is stated that most cabbages are highly perishable. Therefore, white cabbage is assumed to be perishable. g: Source: FAO (2017) h: Source: USAID (2008) i: Source: Gross et al. (2016) 41

42 There are vegetables included in Table 4. 1 that are not specifically mentioned by both Gross et al. (2016) and the FAO (2017). This holds for the following vegetables: chicory, horseradish and leek. The storage time intervals for these vegetables are provided by the University of Maine (2001). Then, the maximum storage time is matched to the right category in Table 4. 2 to determine whether these vegetables are storable or perishable. In case there is only a classification available from the FAO, then this classification is guiding for the determination to what category the vegetables belong. In case there is only a classification available from USAID, then the storage time interval of these vegetables is matched with the right category in Table Again, this table from the FAO is guiding in the determination whether vegetables are storable or perishable. To conclude, the FAO classification is guiding in all cases, and the classification of USAID can be considered as a confirmation if it uses the same classification as the FAO. The FAO divides the vegetables into 5 categories that indicate the relative perishability, which is presented in Table 4. 2: Relative perishability Potential storage life (weeks) Storability Very high <2 Perishable High 2-4 Perishable Moderate 4-8 Perishable Low 8-16 Storable Very low >16 Storable Table 4. 2 Relative perishability (FAO, 2017) Table 4. 2 presents 5 categories of relative perishability. The first column presents 5 distinctive relative perishability categories. The second column presents the potential storage life in weeks. In this thesis, only vegetables that belong to very high, high or moderate are considered perishable vegetables. Vegetables that belong to the remaining two categories are considered storable vegetables. The storability can be found in the third column, which is added by the author. Figure 4. 1 presents the maximum storage time per vegetable. It should be noted that there is no difference between the three horizontal lines. The second thing to be noted is that there is a black vertical fat line at week 8, which represents the boundary of perishable (<8 weeks) or storable (> 8 weeks). Figure 4. 1 Maximum storage time by vegetable in weeks (Based on Table 4. 1) 42

43 Note: The black vertical fate line at week 8 indicates the boundary of storability of vegetables. Vegetables on the left side of the fat line are considered perishable, and the vegetables on the right side are considered storable. Table 4. 3 presents key statistics related to the data of the vegetables included in the used data. All observations used to construct Table 4. 3 are measured in EUR/kg. The vegetables are ranked in such a way that all storable vegetables are in the top of the table, and the perishable ones at the bottom. Second, within each storability category, the vegetables are ranked to their maximum storage time, which can be found in Table The storable vegetables with the largest maximum storage time are ranked first. Within the category of the perishable vegetables, the ones with the highest maximum storage time are ranked first. The vegetable with the highest average price is ranked first when vegetables have the same maximum storage time. All vegetables in column 1 are labelled with an (a) or (b). These (a) and (b) indicate the data source of the time series, where (a) refers to ZMP, and (b) refers to AMI. Some vegetables are included more than once, because there are 24 time series from the two datasets, while there are only 16 different vegetables. However, all of the 24 time series are distinctive. For instance, there are five time series of tomatoes, but these 5 time series have another country of origin, which is described in the second column. There are two time series of tomatoes originating from Spain. These two time series differ from each other in the size of tomatoes. Onions (2x), Lamb s lettuce (2x), Peppers (2x) and Lettuce (2x) are also included more than once. The third column presents the number of observations. The time series from ZMP range from with 52 observations each year. In total, each time series should have (9*52)= 468 observations. The time series from AMI range from , which amounts to (17*52)= 884 observations. However, there is a limited number of missing observations, which is presented in column 4. This number of missing observations is calculated by counting the missing observations between the first and last observation. The fifth column presents the mean values per time series, which can be considered as the average prices. In the remainder of this thesis, the term average price is used. The sixth column presents the median of each time series. Then, column 7 presents the difference between mean and median. This gives an indication how skewed the observations are distributed. When this value is low, it means that the observations are not skewly distributed. Column 8 and column 9 present the minimum and maximum value respectively. Then, column 10 presents the range of every time series, that is, the difference between the maximum and minimum value. Thus, for each time series, the range is calculated by subtracting the minimum value from the maximum value. Column 11 presents a value that is calculated by dividing the range by the median. This value indicates the relative spread. Relative because the values in column 11 are now corrected for the median. Column 12 and column 13 present the standard deviation and the coefficient of variation for each time series. Column 14 presents the storability category according to Table 4. 1, and column 15 presents the ranking of the vegetables according to their maximum storage time, which is also presented in Table

44 1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12) 13) 14) 15) Vegetable Origin # obs # missing o Mean Median (5)-(6) Min Max Range (10)/(6) St. D CV Category Max st. time Horseradish (a) Germany S 52 Garlic (a) France S > 39 Carrot (a) Germany S 39 Onion (a) Italy S 35 Onion (a) Spain S 35 Leek (a) Belgium S 13 White cabbage (a) Germany P 6 Lamb's lettuce (a) Germany P 4 Lamb's lettuce (a) Belgium P 4 Chicory (a) Netherlands P 4 Peppers (b) Unknown P 3 Lettuce (b) Unknown P 3 Peppers (a) Turkey P 3 Lettuce (a) Belgium P 3 Eggplant (a) Italy P 2 Cucumbers (b) Unknown P 2 Beetroot (a) Germany P 2 Oyster mushroom (agermany P 1.5 Mushroom (a) Germany P 1.5 Tomatoes (a) Italy P 1 Tomatoes (a) Netherlands P 1 Tomatoes (b) Unknown P 1 Tomatoes (a) Spain P 1 Tomatoes (a) Spain P 1 Source: author (a) Data source: ZMP (b) Data source: AMI Note: all observations were measured in EUR/kg Note: vegetables are ranked to their maximum storage time (column 15). Table 4. 3 Descriptive statistics data (Source: own calculation) The first central insight from Table 4. 3 is that column 7, which indicates the difference between the mean and median, presents lower values for storable vegetables than for the perishable ones. The average value in this column for the storable vegetables is 0.03, whereas the average value in this column for the perishable vegetables is This means that the price observations of perishable vegetables are generally more skewly distributed than the price observations of storable vegetables. In practice this means that there is a larger price uncertainty for perishable vegetables than for storable vegetables. This makes the price behaviour of storable vegetables more predictable than for perishable vegetables. Extreme values, or even outliers, make the distribution of price observations of perishable vegetables to be more skewed. These extreme values, or outliers, can also be observed in the seasonal price patterns of perishable vegetables. These seasonal price patterns will contain more (large) peaks and lows than the seasonal price patterns of storable vegetables. The second central insight from Table 4. 3 can be found in column 10, which presents the price range. Storable vegetables have generally a smaller price range than perishable vegetables. The average price range of storable vegetables is 1.01 EUR/kg, whereas the average price range of perishable vegetables is 3.49 EUR/kg. There is a factor between these values of almost 3.5. When the range is corrected for the median (column 11), then the average value for storable vegetables is 1.33 EUR/kg, and 1.85 EUR/kg for the perishable vegetables. This second central insight can to a certain extent be linked to the first central insight. The first central insight was that price observations of perishable vegetables are generally more skewly distributed than the price observations of storable vegetables. Extreme values or outlier can cause this skewed distribution. The second central insight is that the price range of perishable vegetables is larger than the price range of storable vegetables, even if it is corrected for the median (column 11). 44

45 This insight is in line with the first central insight, because extreme values or outliers will increase the price range. The range of price observation is reflected in the seasonal price pattern of vegetables. For perishable vegetables, this means that one could observe a price pattern that has a clear range. That is, one observes a larger range for perishable vegetables than for storable vegetables. In fact, the price range is the price amplitude. Thus, price amplitudes will be larger for perishable vegetables than for storable vegetables according to the range, which is presented in Table The third central insight from Table 4. 3 can be found in column 12, which presents the standard deviations of the time series. The storable vegetables have, on average, a lower standard deviation than perishable vegetables. The storable vegetables have on average a standard deviation of 0.17 EUR/kg, whereas the perishable vegetables have a standard deviation of 0.58 EUR/kg. This means that the price observations of storable vegetables are closer to the average price than the price observations of perishable vegetables. However, when the standard deviation is corrected for the average price, which is the coefficient of variation (column 13), then this value is on average 0.21 for the storable vegetables, whereas this average value is 0.29 for the perishable vegetables. The third central insight form Table 4. 3 is in line with the first central insights. The first central insight was that price observations of perishable vegetables are generally more skewly distributed than price observations of storable vegetables. The second central insight was that the average price range of perishable vegetables is larger than the average price range of storable vegetables. The third central insight adds that the standard deviation of perishable vegetables is larger than the standard deviation of storable vegetables, even if it is corrected for the average price. Extreme values or outliers increase the standard deviation, because these values deviate a lot from the average value. The third central insight is therefore perfectly in line with the first two central insights. For a detailed data description, see Appendix III. This Appendix contains an extended version of Table 4. 3 with more descriptive statistics. This Appendix further contains a detailed explanation of how all different values in the table are calculated. The used datasets from AMI and ZMP both contain 52 observations per year, which are assigned to the months of a year. Appendix IV provides an explanation how this is done. In the remainder of section 4.1.1, selected aspects of Table 4. 3 are analysed. These aspects include one of the following three variables: mean, range or coefficient of variation (CV). This brief analysis will be conducted to check if there are clear patterns between storable and perishable vegetables according to the three variables of interest. The vegetables will be ranked by mean value (Table 4. 4), by range (Table 4. 5) and by CV (Table 4. 6). There is one thing to be noted. In this analysis, all years and all observations are taken into account to calculate the mean value, range and CV. Thus, this is no analysis by year or by month. This can implicate that the analysis may be a little bit distorted, because there could for example be a trend of increasing prices through time. This is neglected in this analysis. Table 4. 4 indicates that from the 6 storable vegetables, there are 4 listed at the bottom of the table. These 4 storable vegetables seem to have a relatively lower average price. Another insight from Table 4. 4 is that the average prices of the first three vegetables is much larger than the average prices of the last three vegetables. The average price of the first three vegetables is on average 5.32 EUR/kg, whereas the average price of the last three vegetables is on average 0.33 EUR/kg. This difference yields a factor of (5.32/0.33)=

46 The average price of the last three vegetables is on average 0.33 EUR/ kg, which makes these vegetables less sensitive to large price amplitudes than the first three vegetables with an average price of 5.32 EUR/kg. There is simply less room for large price amplitudes when the average price is low. A 20% price increase of the last three vegetables will result in an average price of 0.33*1.20= 0.40 EUR/kg, whereas a 20% price increase of the first three vegetables will result in an average price of 5.32*1.20= 6.38 EUR/kg. The average price of the first three vegetables increased by ( )= 1.06 EUR/kg, and the average price of the last three vegetables increased by only ( )= 0.07 EUR/kg. Vegetables with low average prices are less sensitive to large price amplitudes or seasonality. 1) 2) 5) 14) Vegetable Origin Mean Category Lamb's lettuce (a) Germany 5.73 P Lamb's lettuce (a) Belgium 5.51 P Oyster mushroom (a) Germany 4.72 P Mushroom (a) Germany 2.73 P Horseradish (a) Germany 2.64 S Tomatoes (a) Italy 2.52 P Garlic (a) France 2.32 S Peppers (b) Unknown 2.00 P Tomatoes (a) Netherlands 1.70 P Tomatoes (b) Unknown 1.60 P Lettuce (b) Unknown 1.47 P Peppers (a) Turkey 1.46 P Chicory (a) Netherlands 1.42 P Eggplant (a) Italy 1.28 P Cucumbers (b) Unknown 1.17 P Lettuce (a) Belgium 0.98 P Tomatoes (a) Spain 0.90 P Tomatoes (a) Spain 0.90 P Leek (a) Belgium 0.86 S Onion (a) Italy 0.78 S Beetroot (a) Germany 0.47 P Carrot (a) Germany 0.39 S Onion (a) Spain 0.34 S White cabbage (a) Germany 0.26 P Table 4. 4 Vegetables ranked according to descending average price (Source: own calculation) Table 4. 5 shows a similar picture as Table The storable vegetables appear in the bottom of the table, with Leek as an exception. The price range of storable vegetables seem to be relatively smaller than the price range of perishable vegetables. The average range of the first two vegetables equals 12.3, whereas the average range of the last two vegetables equals This difference yields a factor of (12.3/0.385)= 16. This factor has implications to both consumers and producers. For consumers, it means that there is only little price variation of storable vegetables over time, whereas the price variation of perishable vegetables is larger. 46

47 The consumer has more certainty about expenditures on storable vegetables compared to perishable vegetables. For producers, it means that there is large price uncertainty among vegetables, especially perishable vegetables. Large price uncertainty means that producers cannot make a good estimation of what the revenue will be for the upcoming period. As mentioned, the price ranges (or amplitudes) of the first vegetables are larger than the price amplitudes of the last vegetables. However, it should be noted that these price amplitudes are calculated by taking all years into account. Two outliers in all the available years (a minimum outlier and a maximum outlier) can result in a large price range. Nevertheless, it is assumed that these possible outliers do not make a significant difference. This can be justified by comparing Lamb s lettuce in Table 4. 5 to Lamb s lettuce in Table Lamb s lettuce has a large price range, but a low coefficient of variation. With respect to seasonality, one can conclude that perishable vegetables have generally larger price ranges than storable vegetables. This also means that the price amplitudes of perishable vegetables are generally larger than the price amplitudes of storable vegetables. Seasonal price patterns of storable vegetables therefore look flatter or more stable than seasonal price patterns of perishable vegetables. 1) 2) 10) 14) Vegetable Origin Range Category Lamb's lettuce (a) Belgium P Lamb's lettuce (a) Germany P Tomatoes (a) Italy 3.89 P Peppers (a) Turkey 3.14 P Lettuce (b) Unknown 3.13 P Oyster mushroom (a) Germany 3.07 P Tomatoes (a) Netherlands 3.01 P Eggplant (a) Italy 2.80 P Peppers (b) Unknown 2.67 P Leek (a) Belgium 2.60 S Cucumbers (b) Unknown 2.38 P Tomatoes (b) Unknown 2.22 P Chicory (a) Netherlands 2.09 P Lettuce (a) Belgium 1.92 P Tomatoes (a) Spain 1.80 P Tomatoes (a) Spain 1.80 P Garlic (a) France 1.31 S Horseradish (a) Germany 1.02 S Onion (a) Italy 0.66 S Onion (a) Spain 0.55 S Beetroot (a) Germany 0.48 P White cabbage (a) Germany 0.47 P Carrot (a) Germany 0.43 S Mushroom (a) Germany 0.34 P Table 4. 5 Vegetables ranked according to the range (descending) (Source: own calculation) 47

48 The storable vegetables also appear in the bottom of Table However, this pattern is less clear than in Table 4. 4 and Table In Table 4. 6, two storable vegetables are ranked quite high. There is also a difference between the coefficient of variation of the first two and last two vegetables. The average CV of the first two vegetables is 0.39, whereas the average CV of the last two vegetables is This difference yields a factor of (0.39/0.045)= 9. The CV presented in Table 4. 6 can be considered as a measure of relative variation, because the standard deviation is corrected for the mean. Storable vegetables tend to have relatively lower CV s than perishable vegetables, based on Table However, it should be noted that two storable vegetables are ranked very high in Table 4. 6, which makes it complicated to justify that storable vegetables tend to have relatively lower CV s than perishable vegetables. Based on these findings, one could conclude that storable vegetables tend to have lower CV s, which means that the weekly price observations did not deviate that much from the average price. In contrast, perishable vegetables tend to have relatively larger CV s, which means that the weekly price observations did deviate more from the average price than the storable vegetables did. This leads to the result that storable vegetables are less sensitive to seasonality than perishable vegetables. In graphical terms, the price amplitudes of storable vegetables tend to be smaller than the price amplitudes of perishable vegetables. 1) 2) 13) 14) Vegetable Origin CV Category Tomatoes (a) Netherlands 0.40 P Leek (a) Belgium 0.38 S Peppers (a) Turkey 0.38 P Onion (a) Spain 0.37 S Lettuce (a) Belgium 0.36 P Lamb's lettuce (a) Belgium 0.36 P Lettuce (b) Unknown 0.36 P Cucumbers (b) Unknown 0.35 P Eggplant (a) Italy 0.34 P Tomatoes (a) Spain 0.32 P Tomatoes (a) Spain 0.32 P White cabbage (a) Germany 0.30 P Tomatoes (a) Italy 0.28 P Peppers (b) Unknown 0.28 P Lamb's lettuce (a) Germany 0.28 P Chicory (a) Netherlands 0.27 P Tomatoes (b) Unknown 0.25 P Carrot (a) Germany 0.22 S Beetroot (a) Germany 0.17 P Onion (a) Italy 0.17 S Oyster mushroom (a) Germany 0.13 P Garlic (a) France 0.12 S Horseradish (a) Germany 0.06 S Mushroom (a) Germany 0.03 P Table 4. 6 Vegetables ranked according to the coefficient of variation (descending) (Source: own calculation) 48

49 The similarity between Table 4. 4, Table 4. 5 and Table 4. 6 is that the storable vegetables appear at the bottom of the tables, and the perishable vegetables are ranked at the top of the tables. With respect to the price range, storable vegetables have low price ranges, since prices of storable vegetables can be smoothed by storage, and selling products at a chosen time (see 2.2.4). Another similarity of the abovementioned three tables is the conclusion that the storable vegetables are less sensitive to seasonality than perishable vegetables. Storable vegetables had lower average prices (Table 4. 4), which makes these vegetables less sensitive to seasonality compared to perishable vegetables. The lower the average price, the smaller the room for large price amplitudes (or seasonality). Storable vegetables had lower price ranges than perishable vegetables, which also means that price amplitudes of storable vegetables tend to be smaller than the price amplitudes of perishable vegetables. Again, storable vegetables are less sensitive to seasonality based on Table Storable vegetables also tend to have lower CV s, compared to the CV s of perishable vegetables. Price observations of perishable vegetables deviate more from the average price than the price observations of storable vegetables. The overall conclusion from Table 4. 4, Table 4. 5 and Table 4. 6 is that storable vegetables are less sensitive to seasonality than perishable vegetables. The difference between Table 4. 4, Table 4. 5 and Table 4. 6 is that the first two tables have a high a factor of 16 between the highest and lowest values, whereas the factor between the highest and lowest value in Table 4. 6 is 9. However, the factors are still high. The main difference is that Table 4. 4 and Table 4. 5 had a clearer difference between storable and perishable vegetables. The difference between storable and perishable vegetables in Table 4. 6 was less clear. Although there are differences between these three tables, the overall conclusion in the previous paragraph is straightforward Seasonal price patterns The concept of the price amplitude was discussed in section It reflects the maximum price minus the minimum price within a certain time period, which is set in this analysis to 1 month and 12 months, respectively. For each of the time series, three price amplitudes are calculated: I. The annual price amplitude is calculated by subtracting the lowest (weekly observation from the highest (weekly) observation within a given year. II. The first monthly price amplitude is calculated by subtracting the lowest weekly price observation from the highest weekly price observation within a single month of a given year. III. The second monthly price amplitude is calculated by subtracting the lowest weekly price observation from the highest weekly price observation within a single month, without taking different years into account. The annual price amplitude (I) is analysed in section Examples and graphical illustrations are presented in this section. The two monthly price amplitudes, (II) and (III), are discussed in section The difference between the monthly price amplitudes is that the first one (II) takes all weekly price observations within a single month of a given year into account when calculating the monthly price amplitude. The second monthly price amplitude (III) takes all weekly price observations within a single month into account, but now for all available years. Thus, the second monthly price amplitude (III) is computed out of many more weekly price observations than the first monthly price amplitude (II). All these price amplitudes are calculated for each of the 24 time series. For a detailed overview, see Appendix I. 49

50 4.1.3 Annual price amplitudes The first part of this section presents two figures that provide a first descriptive analysis of the seasonal price patterns of both a perishable vegetable (peppers), as well as a storable vegetable (garlic). The differences in price patterns will be briefly analysed. The price patterns are based on the used datasets from AMI and ZMP. The seasonal price patterns are determined by weekly prices. The second part of this section presents a figure with the seasonal price pattern of tomatoes. This is done to illustrate the difference between this seasonal pattern with the seasonal pattern of peppers. The difference is that the seasonal pattern of tomatoes contains more variation between years than the seasonal pattern of peppers through years. Figure 4. 2 Seasonal price pattern of peppers from Turkey (Source: own calculation) The vertical axis represents the price in EUR/ kg, and the horizontal axis represents the week numbers. The yearly price patterns of peppers turn out to be quite similar in the period Within the respective years, there is a price peak in week Figure 4. 2 indicates a large annual price amplitude for most of the years. For instance, the price amplitude in 2004 was 3.14 EUR/ kg. This means that there is large seasonality in the pepper prices. Figure 4. 2 indicates that pepper prices from this time series is subject to large seasonality, which is in line with both the theory as well as the empirics from section

51 Figure 4. 3 Seasonal price pattern of garlic from France (Source: own calculation) The vertical axis in Figure 4. 3 represents the price in EUR/ kg, and the horizontal axis represents the week numbers. The price of garlic indicates quite a stable price pattern throughout the year. This stable price pattern is found for every year. The largest amplitude was measured in 1999, which was 0.94 EUR/ kg. This is a much smaller annual price amplitude than the largest price amplitude found in Figure Figure 4. 2 and Figure 4. 3 differ from each other in annual price amplitudes. The annual price amplitudes of peppers are generally higher than the annual price amplitudes of garlic. However, there is also a similarity. The seasonal price patterns of peppers in Figure 4. 2 are quite similar in the period There is little variance, but there is a general pattern. The same holds for the seasonal price patterns in Figure The annual seasonal price patterns in Figure 4. 3 also show a general pattern. That is, prices are quite stable throughout the year. This holds for all years. Thus, in both figures there is a clear seasonal price pattern. Pepper prices in Figure 4. 2 are subject to seasonality because of the large price amplitudes. The seasonality in garlic prices (Figure 4. 3) was much smaller compared to the pepper prices. Given the fact that peppers are considered perishable and garlic storable, these findings are in line with both the theory in chapter 2, as well as the findings in The seasonal price patterns in Figure 4. 4 also indicate that these tomato prices have large price amplitudes. The largest price amplitude was in 2003, when it was 3.17 EUR/kg. Figure 4. 4 indicates that there is quite some variation between years. The lines are more spread compared to Figure 4. 2 and Figure But there is a certain pattern observable. The tomato prices reach the lowest price around week 29, and this holds for most years. 51

52 Figure 4. 4 Seasonal price pattern of tomatoes from Italy (Source: own calculation) With respect to seasonality, one can conclude that these tomato prices have large price amplitudes (see Figure 4. 5), which is in fact seasonality. Concluding, there is large seasonality in these tomato prices. Given the fact that tomatoes are perishable vegetables, this finding is in line with previous findings in this section, as well as the findings in Figure 4. 5 presents the annual price amplitudes corresponding to Figure That is, both figures are based on the same data. It could easily be observed that there is a large price amplitude for every year. The lowest price amplitude was in 2001 with a value of 1.55 EUR/kg. This figure already reveals that there is large seasonality in the tomato prices, because of the large price amplitudes. Figure 4. 5 Annual price amplitudes of tomatoes from Italy (Source: own calculation) 52