Price risk perceptions and management strategies in selected European food supply chains: An exploratory approach

Highlights

•	Price risk management strategies in EU food chains are diverse and well beyond traditional instruments such as futures and forward contracts.
•	Deviations of prices by more than 10 to 15% from expected levels are perceived as price volatility by food chain actors.
•	Price risk management decisions are interrelated with other business decisions.

Abstract

Agricultural prices in European food markets have become more volatile over the past decade exposing agribusinesses to risk and uncertainty. This study goes beyond the farm stage and explores through interviews the price risk perceptions and management strategies in multiple stages of the food supply chain. Respondents were farmers, wholesalers, processors, and retailers in six European food supply chains. Results show that price risk management strategies in EU food chains are diverse and well beyond traditional instruments such as futures and forward contracts. We further find that deviations of prices by more than 10–15% from expected levels were perceived as price volatility by a majority of the chain actors. This study provides new insights on price risk management, a deeper understanding of price risk perceptions and highlights the interrelation of price risk management decisions with other business decisions.

Keywords:

• Price risk
• Perceptions
• Management strategies
• European union
• Exploratory
• Interviews
• Food supply chains

1 Introduction

Prices in European agricultural markets have become increasingly volatile in the past decade [1]. The decoupling of farm income supports through successive reforms of the Common Agricultural Policy has led to a market oriented EU (European Union) farm sector that is increasingly exposed to market price volatility [2]. Price volatility implies uncertainty which may in turn lead to reduced investment in productive inputs and reduced supply by farmers [3–5^[3][4][5]]. The negative effects of price volatility also extend to actors in the downstream stages of food supply chains. According to a report by Rabobank [6], the downstream stages of the food supply chain are increasingly faced with price and supply uncertainty and are obliged to alter their sourcing strategies to mitigate the negative effects. Food security of consumers spending a large share of their income on food is also threatened by price volatility (Hernandez et al.) [7]. The above assertions suggest that managing the risk from price volatility should be done at all levels of the food supply chains.

Risk perception and management among farmers has been extensively investigated within the current literature (see for instance Martin [8], Meuwissen [9], Hall et al. [10], Bergfjord [11]). Although the literature found that farmers perceive price risk as an important source of agricultural risk, it failed to explore the price risk management strategies farmers adopt in practise. This is because the studies relied on structured questionnaires that pre-specified strategies instead of asking farmers which strategies they use in practise. The pre-specified strategies are in general limited to traditional instruments such as hedging in derivative markets, forward contracts, and diversification. To our knowledge, the studies by Heyder et al. [12] and of von Davier [13] were the only ones that addressed price risk and its management in downstream stages of the food supply chain. Similar to the farm level studies, these two studies also failed to explore actual price risk management practises as they also relied on structured questionnaires. Furthermore, the structured questionnaires generally ascribe a particular strategy to a particular source of risk (e.g. futures contracts for volatile prices) [14]. This undermines the discovery of other strategies that would not normally be considered solutions to that particular source of risk while they could in fact be solutions. A more open-ended exploratory approach to data collection could uncover price risk management strategies previously unexplored in previous quantitative studies.

Although assessing farmers’ risk perception in a categorical and quantitative manner is analytically convenient, it is unnatural for farmers to think about risk in this manner [14]. Price risk perception can be better understood with more open-ended approaches to data collection. Although it is generally agreed that price volatility implies risk, it is not clear whether it is perceived as such by food chain actors. This paper will explore through forty-two semi-structured interviews the price volatility perceptions and management strategies of farmers, wholesalers, processors and retailers in six EU food supply chains. The six chains are the Bulgarian wheat, French wheat, German pork, Dutch cheese, Dutch tomato, and Spanish tomato supply chains. Actors’ price risk perception is explored with respect to two elements. The first is the percentage price deviation from expected values which actors perceive as price volatility. The second element concerns the factors that determine whether perceived price volatility is perceived as risky. Exploring actors’ perceptions of price risk helps to better understand actors’ choices of price risk management strategies.

The benefits of qualitative exploratory research for the field of agricultural risk management are demonstrated in this study. By qualitatively exploring food chain actors’ price risk management strategies, this study informs quantitative risk management research on previously unexplored and non-traditional price risk management practises. Future studies can then investigate the adoption level of the explored strategies through large scale surveys that use an up-to-date list of price risk management strategies in their questionnaires. The usefulness of comparative methodological approaches in risk management studies are further demonstrated in this study. The comparison of perceptions and strategies across different food chains and different stages of the chains highlights the context specific nature of price risk management. Findings of this research are also informative to policy makers. Chain actors’ price risk perceptions can help define ‘excessive’ level of price volatility needed for policy intervention. On the other hand, gaps in the identified strategies inform policy makers of where in chain policy intervention is needed.

In the remainder of this paper, section 2 briefly discuss related agricultural risk management researches, section 3 details the methodological approaches used in this study, section 4 presents the results and section 5 discusses the results. Section 6 concludes the study and draws research and policy implications.

2 Previous research

Previous research on risk perceptions and management strategies in the agricultural sector has mainly focused on the farm stage. Wilson et al. [15] defined risk perception as “the awareness of the factors in the social and economic environment that create risk and the degree to which one factor is more critical than the other”. This definition is shared by most of the studies that investigated farmers’ risk perceptions. The methodological approach these studies followed is to list a set of possible sources of agricultural risks and ask farmers to rate the importance of each source of risk using Likert scales (for instance, Martin [8], Meuwissen et al. [9]; Hall et al. [10], Bergfjord [11], Wilson et al. [15], Patrick et al. [16], Knutson et al. [17], Greinier et al. [18]). A common finding of these studies is the high score that farmers assign to price risk (for instance, Meuwissen et al. [9], Bergfjord [11], Wilson [15], Patrick et al. [16], Knutson et al. [17],). The inconsistency in the terminologies the authors use to define price risk reflects, nevertheless, a lack of agreement about what price risk really is. A precise definition of price risk is not provided by the literature. Authors that use the term “price risk ” also use the terms “price variability” (Knutson et al. [17]; Patrick et al. [16]) and “declining prices” (Greinier et al. [18]) and rising “input costs” (Greinier et al. [18]) interchangeably.

The above cited studies adopted a similar approach to assess farmers’ risk management strategies. Farmers were presented with a list of pre-specified risk management strategies and asked to rate the importance or the relevance of each strategy using Likert scales. The main price risk management strategies these studies considered were forward contracts, futures and options, and off-farm and on-farm diversification. The scores assigned to the risk sources and those assigned to the risk management strategies were then compared. Surprisingly, many authors did not find a match between the score assigned to price risk and those assigned to the considered price risk management strategies (for instance, Martin [8], Meuwissen et al. [9], Hall et al. [10]; Bergfjord, [11]). Although price risk ranked at the top of the list of risk sources, the importance or relevance scores assigned to the listed price risk management strategies were unexpectedly low. This raises the question whether the strategies considered are indeed the strategies farmers adopt to deal with price risk. The generic categories of risk management strategies provided in the questionnaires and/or the limited knowledge on available risk management strategies restrict the identification of the possible set of strategies that farmers use in practise.

The only two studies that investigated the price risk perceptions and management strategies of actors downstream from the farm stage are those of Heyder et al. [12] and von Davier [13]. Heyder et al. [12], who surveyed German agribusiness firms, used actors’ expectations of price volatility developments in the next five years as a measure of perceived price risk. Similar to farm-level studies, a set of pre-defined price risk management strategies were presented to the actors who then had to evaluate the relevance of each strategy using Likert scales. The study by von Davier [13] relied on a media content analysis to identify perceptions about causes and developments of price risk and suggested management strategies. A limitation of both studies is that they failed to explore actual management strategies adopted by firms. Another limitation is that these studies did not investigate the actual levels of price volatility that chain actors perceived as risky.

In summary, previous research provides limited evidence on actual price risk perceptions and management strategies in the chain. The downstream sector of the chain has remained overlooked in previous research, as the focus has been mainly on the farm sector. These gaps in the literature are addressed in this research by following an exploratory methodological approach and by including the downstream stages of the chain in the analyses.

3 Materials and methods

3.1 Exploration through in-depth interviews

Exploration is used as a methodological approach “when a group, process, activity or situation has received little or no systematic empirical scrutiny or has been largely examined using prediction and control rather than flexibility and open-mindedness” [19]. Previous research has given little attention to the price risk perceptions and management strategies of actors in food supply chains. At the farm stage, the structured nature of the questionnaires distributed to farmers restricts the identification of the set of strategies farmers adopt in practise. An exploratory approach was therefore deemed appropriate to investigate chain actors’ price risk perceptions and management strategies.

Data was collected using in-depth interviews with semi-structured questions. Some structure was imposed on the questions to guide the interview process and keep the focus on the key topics that were the subjects of the investigation. The imposed structure also assured some consistency in the questions across respondents. The questions, nevertheless, allowed some room for probing and in-depth inquiry. Probing was facilitated by including ‘non-standardized’ or semi-structured questions. Non-standardization in interviews is “most helpful when exploring new topics, sensitive […] issues, and when the businesses are highly variable in their characteristics” [20]. The newness of the topic of price risk perceptions and management strategies, the sensitive nature of disclosing price related business strategies, and the wide ranging types of companies included in this study justified the use of semi-structured interviews.

3.2 Sample selection

A sample was constructed with representatives from five EU countries, four types of food products, four stages of the food supply chain, and different types of agribusinesses. Although the selected sample is not a statistically representative one, it accounts for the diversity in EU food chains in terms of chain structure and type of agribusinesses. The process of selecting respondents followed a series of steps. In the first step, we selected four classes of food products, namely, meat, dairy, cereals, and vegetables. Next, we selected food chains per class of food product, i.e. the fresh pork, cheese, wheat-bread, and fresh tomato supply chains. The selection of these chains further distinguishes between supply chains of fresh and processed food products. The food products, namely, fresh pork, cheese, wheat (bread) and fresh tomato were selected due to their important shares in the EU’s meat, dairy, cereal and vegetable production. For instance, the shares of cheese in the EU-28’s milk utilization was 36.2% in 2013 [21] and the share of wheat production in the EU-28’s cereal production was 46.9% in 2013 [22]. According to Eurostat [23], pork represents 9.0 % of the total EU agricultural output and is the major type of meat produced in the EU-28. The EU is also a major global producer of tomatoes [24].

In a third step, selected indicators were used to choose the EU countries for which the above food supply chains will be investigated. The final food chains selected were the Dutch cheese, Dutch tomato, German pork, French wheat, Bulgarian wheat, and Spanish tomato supply chains. The indicators used to select these chains were the share of EU total harvested production shares of area used for tomatoes in total land for fresh vegetables, shares of pig production in total livestock production, shares of cheese production in raw milk collected, and shares of wheat in total cereal production. Data compiled from the Eurostat on production and area of agricultural land during the period 2006–2013 was used to calculate the indicators. These indicators were used to rank the countries and determine the level of importance of each product in each country. The varying degrees of economic importance of the selected food products (i.e. fresh pork, cheese, wheat (bread), and fresh tomato) both at the EU and national levels give an indication of the relative economic significance of the selected products to the chain actors producing/trading these products. Chain actors are in turn expected to implement price risk management strategies to cope with the volatility of products that have important economic significance to them.

The fourth step in the sample selection process involved the selection of individual firms along each of the selected chains. Prior to selecting the firms, interviews were conducted with experts in each of the six supply chains. The expert interviews inquired about the key characteristics of the investigated supply chains. Respondent firms were then selected by taking into account these chain characteristics. The key chain characteristics that guided the selection of the respondents were membership in a cooperative/producer organization (for farms), ownership structure (private versus cooperative for the wholesale and processing sectors), size of farm/firm, and export orientation. The experts were also asked to provide contact addresses of potential interviewees. Additional addresses of interviewees were obtained through a snowball process where the initial respondents were asked if they knew other people who would be interested in participating in the interviews.

The respondents were selected along the four stages of the food supply chain, i.e. farm, wholesale, processing and retail. The initial objective was to select at least two respondents per chain stage to diversify the respondents based on the above characteristics (for example, one farmer who is member of a cooperative versus one who is not). It was not, however, possible to reach any respondents for some of the chain stages. In particular, German pig slaughterhouses/processors and retailers were often not willing to participate, most likely because price related strategies are confidential in these companies. The respondents at the farm stage were farm owners, whereas the respondents at the other stages were more diverse and included general managers, sales managers, sourcing managers, and financial directors. The sample selection process resulted in the selection of a total of 42 people for the interviews. The respondents were 15 farmers, 15 wholesalers, 9 processors, and 3 retailers. The respondents usually hold managerial and ownership positions in the companies that were selected; because it was necessary to have representatives with an overview of strategic decisions made in the respective firms. The respondents at managerial levels dealing with sourcing and sales were deemed well-suited for the interviews as price risk management strategies are expected to alter sourcing and selling strategies. Tables 1 and 2 summarize the characteristics of the participant farms and companies, respectively.

Table 1 Characteristics of participant farms.

	BW^a (n = 2)	FW (n = 2)	GP (n = 2)	DC (n = 3)	DT (n = 3)	ST (n = 3)
Farm^b size
Total land size (in ha)
<100	1	0
100–150	0	1
>150	1	1
Number of fattening pig places
<2000			1
>2000			1
Number of milking cows
<100				0
100–150				1
>150				2
Size of greenhouse area (in ha)
<10					3	1
10–15					0	1
>15					0	1
Member of a cooperative
Yes	1	2	0	3	2	1
No	1	0	2	0	1	2

a BW: Bulgarian wheat chain, FW: French wheat chain, GP: German pork chain, DC: Dutch cheese chain, DT: Dutch tomato chain, ST: Spanish tomato chain.

b Farmers source pig feed in GP and cattle feed in DC. Farmers sell wheat in BW and FW, pigs in GP, milk in DC and tomatoes in DT and ST.

Table 2 Characteristics of participant wholesalers, processors and retailers.

	Wholesale^a (n = 15)						Processing^b (n = 9)			Retail^c (n = 3)
	BW n = 3	FW n = 1	GP n = 2	DC n = 3	DT n = 2	ST n = 4	BW n = 3	FW n = 1	DC n = 5	DC n = 1	ST n = 2
Cooperative structure
Yes	0	1	1	0	0	3	1	0	4	0	0
No	3	0	1	3	2	1	2	1	1	1	2
Number of employees
<50	0	1	2	0	0	1	0	0	0	0	1
50–250	0	0	0	2	2	0	2	0	3	0	0
<250	3	0	0	1	0	3	1	1	2	1	1
Involved in exports
Yes	3	0	0	3	2	3	0	1	4	0	0
No	0	1	2	0	0	1	3	0	1	1	2

a BW: Bulgarian wheat chain, FW: French wheat chain, GP: German pork chain, DC: Dutch cheese chain, DT: Dutch tomato chain, ST: Spanish tomato chain. Wholesalers source wheat grain in BW and FW, pigs in GP, cheese in DC, and tomatoes in DT and ST. Wholesalers sell wheat grain in BW and FW, pigs in GP, cheese in DC, and tomatoes in DT and ST. One of the wholesalers is a large Dutch cooperative buying tomatoes from Spain. The suppliers in Spain are not a member of the Dutch cooperative.

b Processors source wheat grain (n = 2) in BW and in FW, and milk in DC. Processors sell wheat flour (n = 2) and bread (n = 1) in BW, wheat flour in FW, and cheese in DC.

c Retailers source cheese in DC and tomatoes in ST. Retailers sell cheese in DC and tomatoes in ST. One of the retailers is a British supermarket buying tomatoes from Spain.

The results of this study cannot be easily generalized per country/product/chain stage due to the small sample size used. Nevertheless, the fact that chain actors were selected from different countries, produce different products and operate in different chain stages can help to see if the differences in country, product and chain stages result in differences in price risk perceptions and price risk management strategies among chain actors. Differentiating across storable (i.e. cheese, wheat) versus non storable products (i.e. tomatoes, pigs) can affect the range of price risk management strategies that can be implemented (for instance, storage and futures markets as price risk management strategies are mainly implemented for storable products). The level of (perceived) volatility can also differ for storable and non-storable products, with the volatility of the latter type of product expected to exhibit stronger price volatility. Representing farms that are members and non-members of farmer cooperatives can highlight differences in the type of price risk management strategies implemented by these two types of farms. As an example, membership in a cooperative can provide member farmers the opportunity to pool the financial resources necessary to implement price risk management strategies that would otherwise require substantial financial outlays of individual farmers. Finally, differentiating respondents across countries can highlight differences in the infrastructure supporting the implementation of certain price risk management strategies (e.g. hedging in futures markets is only possible in countries where such markets exist).

3.3 Interview questions

The interview questions are attached as an Appendix A to this paper. The questions in the Appendix A correspond to questions asked to tomato traders. The questions asked to farmers (tomato, dairy, wheat, pig), traders (cheese, wheat, pigs), processors (cheese, wheat) and retailers (tomato, cheese) are slightly different than those in the Appendix A to this paper. As an example, while the question ‘In your opinion, when does tomato price volatility become risky? applies to the tomato supply chain, the question “In your opinion, when does wheat price volatility become risky?” is for the wheat supply chain. As another example, the question ‘What is the size of your company in total number of employees?’ is not asked to farmers, but only to traders, processors and retailers.

The interview questions consisted of four major blocks, with two structured and two semi-structured blocks of questions. The first block was an introductory block that used structured questions to inquire about characteristics such as farm size, company size, farmer cooperative membership, and legal form (cooperative/non-cooperative) of companies. In the second block of the interview, structured questions were used to evaluate how challenging the respondents have found various sources of business risks in the past 5–7 years. Likert scales from 1 (=extremely challenging) to 7 (=not challenging at all) were used for this purpose. The aim of this second block was to determine how challenging price volatility was relative to other business risks. The list of business risks presented was related to the sourcing and selling activities of the farms and companies.

The third block of the interview consisted of questions about the actors’ perceptions of price risk and the fourth block consisted of questions about the strategies used by the actors to manage the risk from price volatility. Both blocks of questions were made semi-structured to get a deep understanding of actors’ perceptions and management strategies. The semi-structured nature of the questions further allowed the questions to be refined as the interviews progressed. To gain an understanding of their perception of price risk, the respondents were asked to provide the percentage price deviation from an expected price level which they perceived as price volatility. Actors were asked to give the percentage price deviations with respect to the periods that prices are mostly set in the respective chains (i.e. daily, weekly or monthly). They were then asked to indicate the strategies they would use if faced with the indicated or a higher level of price volatility.

The interview responses revealed that actors’ strategic responses to price volatility depended not only on the level of price volatility faced but also on whether such level of price volatility is perceived as risky. The interview questions were thus refined to include questions on the factors that determine the riskiness of a certain level of price volatility. Actors were not always able to think of strategies they would use to specifically manage the risk from price volatility. Further probing questions on actors’ past experiences with price volatility were therefore needed to get a deeper understanding of actors’ strategic choices. The chain actors were probed on whether they have faced price volatility in the past and on the strategies they have used to manage the risk from price volatility.

The first few interviews revealed that actors’ strategic responses to price volatility depended not only on the level of price volatility faced but also on whether such level of price volatility is perceived as risky. Therefore, the question on the factors that determine the riskiness of a certain level of price volatility (Q.8) was added as a structured question in the subsequent interviews. Similarly, when the interviews were first started, some interviewees were not always able to think about the strategies they would use to manage the risk from future price volatility. Therefore, their past experiences with price volatility (Q.7 and Q.9) were first put as probing questions and then included as structured questions in the subsequent interviews. Although the nature of the above interview questions changed as the interviews progressed (i.e. first as probing questions and then as structured questions), all interviewees were exposed to the same questions, therefore having no implication for the results obtained.

The interviews were conducted between January and July 2014. The questions were sent one day in advance to the interviewees to allow them to prepare prior to the interviews. Each interview lasted between forty-five minutes and one hour and a half. The interviews were conducted by the authors with accompanying translators in some cases. The responses were audio-recorded and transcribed on the same day the interviews took place.

3.4 Analysis

We used content analysis of the interview transcripts to describe the price risk perceptions and management strategies of the food chain actors. This type of research design is usually appropriate when existing theory or research literature on a phenomenon is limited [25]. The process of content analysis includes open coding, creating categories and abstraction [26]. Elo and Kyngas [26] define open coding as the writing of notes and headings to describe the content of the interview transcripts. They define abstraction as the naming of the categories and sub-categories of notes and headings generated in the open coding process. The nature of content analysis is therefore such that a pre-defined conceptual framework is not used to guide the research. In this study, content analysis is used to identify the factors that determine whether a certain level of price volatility is perceived as risky and to identify the categories of strategies actors use to manage the risk from price volatility.

4 Results

4.1 Price volatility perceptions

The degree to which various business risks have been a challenge to each actor in the past 5–7 years is shown in Table 3. The risks relate to business-to-business input sourcing and output selling activities. The ‘challenging’ or ‘C’ columns of Table 3 reveal that price volatility has been a prime challenge for a majority of the actors. This table further emphasizes that price volatility is not the sole concern of farmers as is often believed. An interesting finding is the comparable scores the majority of the actors assigned to high input/low output prices and input/output price volatility. The following subsections describe and compare actors’ perceptions of price volatility.

Table 3 Number of actors who rated each business risk as challenging (C), moderately challenging (MC), or not challenging (NC).

	Farm^a (n  = 15)			Wholesale^b (n = 15)			Processing^c (n = 9)			Retail^c (n = 3)
	C	MC	NC	C	MC	NC	C	MC	NC	C	MC	NC
Input sourcing related risks
High input prices	3	2	0	6	4	5	4	2	3	2	1	0
Instability in volume of input supply	0	0	5	6	2	7	4	1	4	1	0	2
Low volume of input supply	0	0	5	4	1	10	4	0	5	1	0	2
Low quality of input supply	0	1	4	3	2	10	1	1	7	1	0	2
Inconsistent quality of input supply	0	1	4	6	1	8	2	1	6	1	0	2
Poor on-time delivery of inputs purchased	0	0	5	3	2	10	1	0	8	1	0	2
Tracing and tracking challenges	0	0	5	1	0	14	2	0	7	1	0	2
Detection of diseases in inputs	0	0	5	3	1	11	1	0	8	1	0	2
Input price volatility	1	0	4	8	2	5	6	0	3	1	1	1
Imbalance of market power with input suppliers	1	0	4	5	5	5	1	0	8	1	0	2
Output selling related risks
Low output prices	9	2	4	10	1	3	6	1	1	Ni^d	Ni	Ni
Instability in demand volume	3	1	11	5	3	6	3	1	4	Ni	Ni	Ni
Low demand volume	3	0	12	6	1	7	5	0	3	Ni	Ni	Ni
Lack of capacity to fill high demand volume	0	2	13	2	1	11	1	1	6	Ni	Ni	Ni
Lack of capacity to fill high quality specifications of customers	1	1	13	2	1	11	1	0	7	Ni	Ni	Ni
Instability in quality specifications demanded by customers	3	0	12	4	0	10	0	1	7	Ni	Ni	Ni
Output price volatility	8	2	5	7	1	6	6	2	0	Ni	Ni	Ni
Imbalance of market power with customers	6	1	8	6	1	7	3	0	5	Ni	Ni	Ni
Poor on-time payment for deliveries	1	0	14	5	1	8	4	1	3	Ni	Ni	Ni

a Highlighted in bold are the number of farmers, wholesalers, processors and retailers who found high input/low output prices and input/output price volatility as challenging. C: Challenging (=1, 2, 3 in Likert scale), MC: Moderately challenging (=4 in Likert scale), NC: Not challenging (=5, 6, 7 in Likert scale). Input sourcing related challenges were not inquired in the case of wheat and tomato farmers.

b One of the wholesalers was not familiar with the selling related challenges of the company as the company sold its output to the parent company.

c The selling related challenges were not inquired for companies (bakery and retailers) selling directly to final consumers.

d Not inquired.

4.1.1 Percentage price deviations perceived as price volatility

Given that this paper is an explorative research with a small sample of respondents, statistical measures of distribution (such as standard deviations, skewness, etc) of the percentage price deviations given by the respondents were not calculated. Instead, simple averages of the percentage price deviations which chain actors perceived as price volatility were calculated per chain and chain stage. Table 4 provides a summary of the average percentages per chain and chain stage. Deviations of prices by more than 10–15% from their expected levels were perceived as price volatility by a majority of respondents. Actors in the Dutch and Spanish tomato chains were an exception to this, as the majority perceived a price deviation higher than 20% as price volatility. Similarly, Dutch dairy farmers argued that feed prices (maize in particular) are volatile if prices deviate by more than 20% from their expected values. Recurring and large changes in the prices of fresh tomatoes and cattle feed explain the price volatility perceptions of the tomato and dairy farmers.

Table 4 Percentage deviations in prices perceived as price volatility.

	Farm	Wholesale	Processing	Retail
Percentage deviation^a	• Wheat grain:10% • Pig feed:10% • Pigs:10% • Cattle feed: 15% • Milk: 10% • Tomatoes: 20%	• Wheat grain: 10% • Pigs: 10% • Cheese:10% • Tomatoes: 20%	• Wheat grain and flour: 10% • Milk and Cheese^b: 10%	• Cheese: 3% • Tomatoes: 20%

a Percentage deviations from an expected price level that exceed the specified percentages are perceived as price volatility by the chain actors. The percentages are averages across respondents, rounded to the nearest decimal. Percentages are expressed relative to the price settlement period in each chain, namely monthly in the Dutch cheese chain and weekly in the rest of the chains.

b Note that only one cheese processor perceived a price deviation greater than 20% as volatility, the rest of the processors perceived a price deviation less than 5% as volatility.

Actors specified percentage price deviations with respect to the periods that prices are set in the respective chains. In the Dutch cheese supply chain this period is monthly. Although cheese prices are set for a longer time period (exceeding one month), Dutch cheese processors, wholesalers, and retailers form cheese price expectations on a monthly basis because milk prices serve as reference prices for cheese. A practise in the Dutch cheese chain is that farmers deliver their milk to dairy cooperatives (which are fairly common in the Netherlands) during the month only receiving payment for their milk at the end of the month. Since cheese prices are (partly) based on milk prices and only the month to month fluctuation of milk prices is relevant, then the month to month cheese price fluctuations are mainly relevant in cheese pricing. In the tomato and pork chains prices are mostly set on a weekly basis. In the wheat supply chains, although high frequency trading can take place at the wholesale stage, weekly price expectations seem to be the norm.

A comparison across chain stages shows some similarities and differences in the specified percentage price deviations. On average, the percentage price deviations (Table 4) are rather comparable across chain stages and chains. An examination of the percentages specified at the level of the individual respondent reveals some differences in perceptions. Processors and retailers seem to perceive lower percentages of price deviations as price volatility compared to farmers and wholesalers. For example, a deviation greater than 5% in grain and flour prices was perceived as price volatility by a Bulgarian and a French wheat miller. Similarly, the Dutch retailer indicated that a deviation higher than 3% in cheese prices is perceived as price volatility.

4.1.2 Factors determining the riskiness of price volatility

4.1.2.1 Persistence of price volatility

Price volatility, defined as a percentage price deviation from the expected level, is not perceived as risky by all interviewed actors. The persistence of price deviation was found to be one of the factors that determine whether price volatility is perceived as risky. The degree of persistence perceived as risky is summarized in Table 5 per chain and chain stage. A finding common to most of the interviewed farmers is that a high input price or low output price persisting for at least one year or production cycle (i.e. a year for wheat and dairy farmers, and one production cycle for pig and tomato farmers) is perceived to be more risky than weekly or monthly changes in prices. A situation perceived to be even more risky is when a persistent high output price level (or low input price level) unexpectedly changes to a persistent low output price level (or high input price level) between years or production cycles. When such reversals in price levels occur, it becomes challenging for farmers to reverse major investments made during good price years. Though undesirable, more frequent price changes (within the year or production cycle) were seen as less risky because price drops can cancel out price increases during the year yielding a fair average annual price. Similar to farmers, retailers tended to be more concerned about changes in yearly prices. Their reasoning is also that higher frequency price changes can compensate each other during the year.

Table 5 Persistence of price volatility perceived as risky.

Farm^a	Wholesale	Processing	Retail
• A year or a production cycle^b (DC, DT, BW, GP, ST) • Depending on stock level and position in forward contract^c (FW)	• Depending on stock level and position in forward contract (BW, DC, FW) • One week and longer^b(GP, DT, ST)	• Depending on stock level and position in forward contract (BW, DC)	• A year (DC, ST)

a BW: Bulgarian wheat chain, FW: French wheat chain, GP: German pork chain, DC: Dutch cheese chain, DT: Dutch tomato chain, ST: Spanish tomato chain.

b Price volatility is perceived as risky if a high input or low output price level persists for the specified period or longer.

c Price volatility is perceived as risky when output prices drop (input prices rise) and stay low (high) during the period that input prices (output prices) are fixed through contracts at a high (low) level. It is perceived as risky when output prices drop while there are goods in stock.

Price changes occurring during the year were found to be more of a challenge for the wholesale and processing stages. This is particularly true for the wheat and cheese wholesalers and processors. Fixed-price sales contracts that are not matched with fixed-price purchase contracts (and vice versa) and storage are the main causes of this challenge. For instance, it is risky when output prices drop and stay low during the period that an input price is fixed at a high level through a contract (and vice versa for input prices). Sudden drops in output price are also risky for goods in stock. Cooperative German pig, Spanish tomato, and Dutch tomato wholesalers were concerned about both weekly changes and persistent changes in the pig and tomato prices received by their member farmers.

4.1.2.2 Reason of price volatility

According to the interviewees, the reasons why prices deviate from their expected levels also determined whether price volatility is perceived as risky. Price changes caused by sudden and major changes in local weather conditions and changes in global demand and supply conditions (caused for instance by conflicts in major producing countries or by border restrictions of major importing countries) were seen as worrying by actors in the cheese and wheat supply chains. Actors in the tomato supply chains mainly considered price changes caused by sudden and major changes in local weather conditions as risky ones. In the pork supply chain, the most challenging price changes were those caused by animal health related crises. Predictable seasonal price changes and price changes believed to have arisen from speculation were not considered as risky by most of the actors.

4.1.2.3 Stability in margins

In addition to the above factors, the direction of the price deviation determined the riskiness of price volatility. The interviews revealed that actors are more concerned about downside price changes (increase in input price or decrease in output price) than price volatility in the sense of fluctuations (both upside and downside) in prices. Moreover, stability in margins was found to be more important than stability in prices. All interviewed actors argued that a sudden and large decrease in an output price is not a concern if it is matched by a proportionate and immediate decrease in the input price (and vice versa). In practise, this rarely happens due to a number of factors, such as time lags in production, contracts (either on the buying or selling side), and the influence of retailers on prices.

4.2 Price volatility management strategies

Strategies actors use to manage risk can be classified in several ways. Waters [27] for instance provides eight categories of risk management strategies used by managers. The eight categories are ignoring risk, reducing the probability of risk, reducing the consequence of risk, transferring risk, making contingency plans, adapting to risk, opposing a change and moving to another environment. Hardaker et al. [28] classified farm level risk management strategies into two broad categories and further sub-categories. The two broad categories are on-farm strategies and strategies to share risks with others. In this paper, we provide our own classification of price risk management strategies. The categories provided emerged from the content analysis of the interview transcripts. Four categories of price risk management strategies were identified: Survival, adaptive, control and hedging strategies. Each category of strategy is briefly described below. Table 6 lists the management strategies classified in each strategy category. The number of actors that adopted each category of strategy is presented in Table 7.

Table 6 Price volatility management strategies.

SC^a	Farm	Wholesale	Processing	Retail
Survive	• Substitute or cut production (DC, ST, BW) • Substitute expensive ingredients (DC) • Increase production efficiency, reduce costs and increase productivity (DC, BW, DT) • Avoid major investments (DC) • Wait a bit and sell at whatever price (ST, GP, BW) • Diversify production (BW, FW) • Promote product by producer organizations in times of sudden price drop caused by excess production (DT)	• Ask farmers to wait a week or two weeks more before harvesting the plant^b (ST) • Agree with competitors to throw away excess production and raise back prices (ST^b, DT) • Increase production efficiency of member farmers^b(ST) • Diversify suppliers (ST) and buyers (DT)^b • Sell quickly at whatever price^b (GP, DC) • Sell excess production through retail promotion (ST, DT^b) • Cut purchases during overproduction as prices are too low to sell back (ST)	• Diversify production (DC) • Cut production (BW)	• Diversify suppliers (ST)
Adapt	• Shorter contracts with small quantities per contract (FW) • Closely follow market development, improve price predictions and concentrate production in high price weeks (ST, DT)	• Shorter contracts (DC) • Renegotiate fixed price contracts (DC) • Long-term contracts with flexible output prices (DC) • Buy aggressively in case of price spikes (due to shortages) and wait when prices are too low (BW) • Take risk by taking a long or short position to profit from cheese price volatility (DC) • Secure supply at whatever price (ST, BW) • Agree on buying and selling price on same day (BW) • Sell majority of grain right after purchasing it during harvest (FW) • Closely follow market development and improve price predictions (DT, BW, DC)	• Use milk pools to set output prices^c (DC) • Shorter sales contracts (DC) • Contracts with flexible output prices (flexible with milk prices; use output price bands to share price risk with retailer; cost-plus pricing) (DC) • Switch production among alternative products (DC, BW) • Adjust production volume (BW) • Agree on input and output price on same day (FW, BW) • Avoid storage/buy only for daily needs (BW) • Transmit price changes (BW) • Buy spot milk to take advantage of volatility (sudden drop in milk prices) instead of buying from own cooperative member farmers (DC) • Ø Renegotiate fixed price contracts (DC)	• Secure supply at whatever price (DC) • Transmit price changes (price decreases in particular due to competition) (DC, ST)
Control	• Fixed price forward contract for inputs (DC, GP) • Fixed price forward contract for outputs (FW, DT^d) • Forward integration to process own milk^d (DC) • Backward integration to produce own maize (DC) • Improve output quality (BW^d, DT, ST) • Closer relationship with retailers for improved product development and with long-term fixed price contracts^d (DT) • Better marketing/promotion of produce by producer organization to add value to the produce (DT)	• Trade quality produce (ST, BW) • Fixed price forward sales contract with 100% advance payment^d (BW) • Fixed price forward contract for outputs (ST) • Pay farmers an average of 2 weeks’ prices (GP) • Merger among wholesalers to gain more market power and secure higher output price^d (DT) • Closer relationship with retailers for better marketing/promotion of produce to add value to the produce (DT)	• Produce quality product (DC, BW) • Do not overreact: fix milk price at moderate level (DC) • Store (BW)	• Secure quality product (DC, ST) • Fixed price purchase contract (ST)
Hedge	• Hedge in futures market (GP^d, FW) • Use average seasonal price offered by cooperatives (FW)	• Hedge in futures market^d (DC, GP, BW) • Use options (FW)	• Hedge in future market^d (DC) • Over-the-counter contracts for milk^d (DC)	None

a SC: Strategy category. Note that BW: Bulgarian wheat chain, FW: French wheat chain, GP: German pork chain, DC: Dutch cheese chain, DT: Dutch tomato chain, ST: Spanish tomato chain.

b Strategies used by cooperative wholesalers, whose main objectives are to minimize the losses that member farmers face in times of sudden drop in prices.

c A cooperative producing only cheese can pay farmers a competitive milk price that is based on a ‘weighted-average’ of final dairy prices of competitors.

d Strategies not yet implemented, but planned for the future.

Table 7 Number of actors per strategy category and chain stage.

Strategy^a category	Farm (n = 15)	Wholesale (n = 15)	Processing (n = 9)	Retail (n = 3)	Total
Survival	13	7	2	0	12
Adaptive	2	10	9	3	24
Control	5	4	3	1	13
Hedging	3	3	1	0	7

a Note that one actor can use a combination of strategies.

The interview questions ‘What strategies did you use to manage the risk from the observed price volatility described in q.10? and “What strategies would you use to manage the risk from future price volatility?” shown in the Appendix A were used to explore the strategies that chain actors have used and would use to cope with price volatility. In order to have a concise reporting of the strategies identified with the two interview questions and arrive at the strategies reported in Table 6, the original strategies indicated by the chain actors were slightly adjusted. For instance, in Table 6, the strategy ‘Wait a bit and sell at whatever price’ categorized as a ‘survival strategy’ was formulated differently by the Spanish tomato, German pig and the Bulgarian wheat farmers. In case of the Spanish tomato farmer, this strategy involved keeping the tomatoes on the plants for one or two more weeks during weeks with significant tomato price drops. In case of the German pig farmer, the strategy involved keeping pigs in their barns (i.e. while fattening them) for a few weeks during which pig prices have dropped significantly. Finally, in case of the Bulgarian wheat farmer, the strategy involves keeping in store the wheat harvested in July until the month of October to see if wheat prices improve. If prices do not improve until then, the farmer sells all of the harvest in October. A similarity across the strategies used by these three farmers is that all involve waiting a bit before selling at the prevailing market price.

After reformulating the strategies obtained from the interviews, they were categorized into the four categories shown in Table 6. The categorization results in groups of strategies that have similarities in the ways the strategies were implemented. Categorization provides a means to describe the phenomenon under study (here the strategies) and to increase understanding (Elo & Kynga). In forming the categories, the strategies were first categorized as those that involve the transfer of risk to a third party and those that do not. This resulted in the ‘hedging’ category. Then, those strategy that do not involve the transfer of risk are categorized based on whether the objective of the chain actor is to minimize losses from adverse price changes or to keep profits/earn superior profits while managing the risk from price volatility. This results in the ‘survival’ category where the aim is to minimize losses. Finally, whether chain actors adopted flexible strategies (e.g. flexible timing of sale, flexible prices, etc) or attempted to control market prices (e.g. with contracts, with improved quality of product, etc) to keep stable profits resulted in the ‘adapt’ and ‘control’ strategies, respectively. It should be noted though that these four categories may not be mutually exclusive as some strategies may fit in more than one of the identified categories. Moreover, one can end up with a different set of categories using the same set of identified strategies depending on the angles from which the strategies are seen. Therefore, the categories provided in this paper should be seen more as a means to briefly describe the list of the strategies listed in Table 6 than as an end goal of the paper.

4.2.1 Survival strategies

Survival strategies are strategies aimed at minimizing losses, such as reducing physical production and major investments, improving efficiency, and diversification. These strategies, which are mainly long-term strategies, were mostly adopted by farmers. As farmers cannot easily respond to short-term price changes, a majority of them concentrated their strategies on price changes that persist for at least one year or production cycle. Producers of storable products, as French wheat farmers in our sample attested, can be considered an exception to this as their ability to store wheat gives them the flexibility to decide when and how to sell. The interviewed Bulgarian wheat farmers, on the other hand, indicated a limited capacity to store wheat during the year. Survival strategies were also adopted by cooperative pig and tomato wholesalers. The only strategy these wholesalers can adopt to minimize the losses of member farmers in times of sudden price drops is to wait one more week before selling farmers’ pigs and tomatoes to processors and retailers, respectively. Keeping pigs and tomatoes for a longer period of time can result in further losses in the values of the produce.

4.2.2 Adaptive strategies

The focus of strategies in this category is on flexibility, following the market, and securing a stable margin regardless of price movements Most of the interviewed wholesalers and processors, except for cooperative pig and tomato wholesalers, adopted adaptive strategies. Setting buying and selling prices on the same day, linking output prices to input prices, and avoiding open long-term fixed price forward contracts are some of the major adaptive strategies these actors adopted. The focus is on flexibility achieved through quick adaptation to market price movements. Not only was there an interest for flexible prices but also for flexible production. For instance, a Bulgarian wheat baker argued that switching from flour to bread production can be a solution in times of big drops in grain prices, and from bread to flour in case of big rises in grain prices. The aim of this strategy is to avoid the drop in bread sales during bread price increases as bread is a staple product in Bulgaria. A specialty cheese processor argued that switching from processing milk to processing more volumes of cheese can be a solution to manage the risk from milk price volatility. In case of a large drop in milk prices, it becomes profitable to process more cheese than processing and selling milk because of the value that cheese adds to the low priced milk.

4.2.3 Control strategies

Strategies in this category focus on achieving price stability by taking control over prices. The expected interest in control strategies through price-fixing contracts and vertical integration was not found among most of the chain actors. This is particularly true for wholesalers and processors. Interest in contracts and vertical integration was however found among farmers (for instance, Dutch dairy, German pig, and French wheat farmers). Producing and trading premium products is another way of exercising control over prices. Examples include the production of specialty cheese by cheese processors, and the production and trading of tomatoes with no pesticide residues and of tomatoes of specialty varieties by tomato farmers, wholesalers, and retailers. Many of the interviewed actors argued that prices of premium products are not as volatile and low as standard products. This argument prevailed in particular among actors downstream from the farm stage. For retailers, transmitting sudden increases in input prices is easier when the product is a premium product. Improved marketing of produce through promotion and better services to fill customers’ needs is another method to add value to the product and command higher and more stable prices. In the Dutch tomato chain, product value addition is achieved through closer collaboration between growers’ associations and the retail sector.

4.2.4 Hedging strategies

Although hedging through futures and option contracts is a widely accepted price volatility management strategy, its use was limited among the interviewed actors. Interest to use these instruments in the future was, nevertheless, expressed by a German pig farmer, a French wheat farmer, a Bulgarian wheat wholesaler, a pig wholesaler, a Dutch cheese wholesaler, and a Dutch cheese processor. The absence of active futures markets for these products in the respective countries was mentioned as the main reason for the current non-use of these instruments. Except for one German pig farmer and one French wheat farmer who currently use options, no mention of interest in hedging with futures and options was made by any of the remaining interviewed farmers.

5 Discussion

Most of the previous quantitative risk management studies have considered traditional instruments such as hedging in derivative markets, forward contracts, and diversification as the main instruments farmers use to manage price risk (for instance in Martin [8]. Meuwissen et al. [9], Hall et al. [10], Bergfjord [11]). This study showed that a notable development in farmers’ strategies is to create added value through selection of better varieties to plant, production with less pesticide residues, product promotion, and collaboration with the retail sector to develop improved products. Prices of premium products are perceived to be more stable than standard quality products. Another development is the avoidance of long-term fixed price contracts particularly by wholesalers and processors. Although long-term contracts are often argued to minimize risk and uncertainty (for instance in Heyder et al. [12]), this study finds the contrary. In this study we find that chain actors perceive fixed price long-term contracts as risky if one is not able to secure such contracts both on the input and output sides.

Besides uncovering developments in price risk management practises, this study also showed that price risk management strategies are diverse and interrelated with other business decisions. At the farm stage, price risk management affects investment decisions (i.e. increasing production efficiency, reducing costs and increasing productivity), product development decisions (i.e. improving output quality) and decisions on vertical and horizontal collaborations (i.e. closer relationship with retailers for improved product development). Production decisions of processors are also affected by price risk management strategies (i.e. cut production, adjust production). At the wholesale stage, it was found that collaboration decisions are affected by price risk management strategies (i.e. Sell excess production through retail promotion, merger among wholesalers to gain more market power and secure higher output price, closer relationship with retailers for better marketing/promotion of produce to add value to the produce).

This paper achieved a deeper understanding of chain actors’ price risk perceptions in contrast to previous quantitative risk management studies which measured risk perceptions by rating the relative importance of different agricultural risks through Likert scales (for instance, Martin [8], Meuwissen et al. [9], Hall et al. [10], Bergfjord [11], Wilson et al. [15], Patrick et al. [16], Knutson et al. [17], Greinier et al. [18],). This study went beyond showing the relative importance of price risk and identified the factors that determined whether price volatility is perceived as risky. The study also explored the percentage deviation of prices from expected levels perceived as price volatility. Future research could explore whether the range of identified percentages apply to a wider sample of chain actors. Through a closer investigation of the identified strategies, one can further detect a link between price risk perceptions and the adopted management strategies. The fact that farmers are concerned more about long-term price changes than short-term price changes can explain why farmers adopt long-term strategies. Long-term strategies include diversification, achieving cost efficiency and quality product development. Wholesalers and processors, on the other hand, worry about monthly or weekly price changes from expected prices, and therefore choose flexibility to manage the risk from these price changes.

This study showed that the choice of price risk management strategies is context specific. While farmers’ strategies tend to fit more within the ‘survival’ category, those of wholesalers and processors fit more into the category of ‘adaptive’ strategies. Strategies also differed across chains. While Dutch tomato farmers are inclined towards strategies that require closer horizontal and vertical collaborations, Spanish tomato farmers tend to adopt more individual strategies. In the wheat chain, Bulgarian wheat farmers’ strategies are limited to long-term strategies implemented on an annual basis (i.e. crop diversification, cutting production, planting new varieties). French wheat farmers on the other hand use also short-term strategies such as short-term contracts. Although this study did not quantify the impact of chain characteristics on the choice of price risk management strategies, it provides the basis on which such studies can be conducted in the future.

This study used semi-structured questions to inquire about chain actors’ price risk perceptions and price risk management strategies. That is both structured questions that did not vary from one interviewee to another as well as probing questions were used. Since the answers to the structured questions were open-ended, they sometimes revealed important information that required adjustments in the questions. The structured questions therefore had to be adjusted during the interviews and these adjustments were kept for the next interview. The questions in the Appendix A show the final adjusted questions. Question 7 (i.e. on past experience with price volatility) and 9 (i.e. strategies to deal with past volatility) were added because it was found that some interviewees were not always able to think about the strategies they would use to manage the risk from future price volatility. Therefore, their past experiences had to be asked. Question 8 was added because some interviewees indicated that they did not implement price risk management strategies just because prices deviated from their expected values (or just because price volatility was observed). The implication of adjusting the structured questions during and between interviews brings the risk of questions’ inconsistency between the interviewees. This in turn can reduce the generalizability of the interview responses to all interviewees. Such limitation can be addressed in future research through pilot testing of interview questions using a fairly large sample of respondents.

A major limitation of this study is the small sample size used. In particular, the limited number of respondents per country/product/chain stage limits the generalizability of the results. Another limitation is the limited coverage of the retail sector. In light of the increasingly concentrated European retail sector, the sector can have significant influence on prices in the chain through its pricing strategies. Weber and Anders [29] for instance assert that the retail sector uses market power to keep consumer prices rigid and irresponsive to upstream price changes. Retailers’ strategies have in turn implications on the prices received by the rest of the chain. Future research aimed at exploring and understanding the price risk management strategies implemented by the retail sector can improve the understanding of price risk management strategies in upstream stages of the chain.

Despite the small sample used, this study has provided new insights in risk management practises, a deeper understanding of risk perceptions and highlighted the interrelation of risk management decisions with other business decisions. Such benefits could also extend to the wider field of agricultural research. A natural extension of this research could be to test whether the strategies identified in this study apply to a wider sample of food chain actors. Future research could also test the effectiveness of the identified strategies in reducing the risk from price volatility. A number of hypotheses to be tested in future research can be derived from the results of this paper. These are, 1) Price volatility is not perceived as price risk if the percentage deviation of prices from expected levels is less than 10%, if margins are stable, if the deviation of prices is short-lived and if the price deviation is caused by speculation, 2) A forward contract fixing prices during the contract period is perceived to be more risky than a contract with flexible prices, 3) The decision to manage the risk from price volatility alters other business decisions such as investment and vertical collaboration decisions, 4) Price risk is the most challenging business risk for all actors in the food supply chain, 5) Farmers of non-storable products have a more limited set of alternative price risk management strategies than farmers of storable products.

6 Conclusions and implications

Food and agricultural commodity prices have been increasingly volatile both at the global and EU levels since the last decade. Although the current literature proposes alternative strategies to deal with price risk, the proposed strategies have often targeted the farm stage. The scope of the proposed strategies has also been limited to few strategies with forward contracts, futures, options and diversification being the main ones. In this study, we took a broader approach and explored the strategies used both at the farm and beyond the farm stages of the food supply chain. The strategies were explored by conducting forty-two in-depth interviews with farmers, wholesalers, processors, and retailers in six EU food-supply chains. The chain actors’ perceptions of price risk were also explored during the interviews to gain a better understanding of actors’ choices of price risk management strategies. The two key perception elements explored were the percentage price deviation which actors perceived as price volatility and the factors that determined whether price volatility is perceived as risky.

Results show that a deviation in prices by more than 10–15% from their expected levels is perceived as price volatility by a majority of the respondents. Price expectations were set with respect to the periods prices are mostly set in the chain (i.e. weekly, monthly). Three main factors determined whether chain actors perceived price volatility as risky: the persistence, the reason and the stability of margins. Whereas farmers and retailers perceive persistent price deviations as more risky, wholesalers and processors perceive short-term price changes occurring during the year or production cycle as more risky. The interviewed farmers and retailers argued that higher frequency price increases can cancel out price drops occurring during the year and yield an overall fair average price for the year.

Farmers’ strategies are mostly survival strategies through output and cost reduction in response to adverse price movements. Wholesalers and processors focus on adaptive strategies that allow them to secure stable margins regardless of price movements. Retailers’ main focus is to secure a continuous supply of quality produce for their customers rather than to reduce price volatility. The results stressed that wholesalers, processors and retailers are in a better position to deal with the risk from price volatility than farmers. Results also showed that price risk management strategies differed depending on chain characteristics such as the nature of the product (e.g. storable versus non-storable), cooperative membership of farmers (e.g. Dutch tomato farmers versus Spanish tomato farmers) and the country investigated (e.g. French wheat farmers versus Bulgarian wheat farmers). Overall, this study highlighted the diversity in perceptions and strategies along EU food chains and challenged current assumptions that price risk management strategies are limited to few traditional instruments.

This study has several implications for agricultural risk management research. First, it demonstrated the benefits of qualitative exploratory research to discover current and new practises in agricultural risk management. Second, it showed that a qualitative approach can highlight the interrelation of price risk management practises with other business decisions. Third, the study showed that a comparative methodological approach can be used to stress the context specific nature of agricultural risk management decisions. This study has also policy implications. The 10 to 15% price deviation perceived as price volatility provides an important signal for price stabilizing policy interventions in agricultural markets. Some strategy gaps could be filled with policy interventions. Areas for policy support include the further encouragement of cooperation among farmers and along the chain (inter-professional organizations), the establishment of futures markets where such markets are missing, and the timely dissemination of improved and accessible market price data and predictions. Such price data could be used to support chain actors’ production and sales decisions, as well as contract decisions.

Acknowledgement

We are grateful for the European Union funding obtained under the FP7 project ULYSSES.

Appendix A

Interview questions

Aim of the interview

The aim of this semi-structured interview is to investigate the strategies tomato traders use to cope with tomato price volatility.

Part I – Introduction

1	What is your position in the company?

2	What is the legal form of your company?

3	What is the size of your company in total number of employees?

4	Who are the key buyers of your tomatoes?

Part II – Rating of business challenges

5	What are the most significant challenges that you have faced in the last 5–7 years in relation to tomato trading? Please evaluate each of the following challenges from 1 to 7 (1 = Extremely challenging, 7 = Not challenging at all).

Table

Challenges related to tomato sourcing
High tomato purchase prices	1	2	3	4	5	6	7
Instability of tomato supply	1	2	3	4	5	6	7
Poor on-time delivery of tomatoes purchased	1	2	3	4	5	6	7
Low volume of tomato supply	1	2	3	4	5	6	7
Low quality of tomato supply	1	2	3	4	5	6	7
Inconsistent quality of tomato supply	1	2	3	4	5	6	7
Poor on-time delivery of tomato supplied
Tracking and tracing challenges of tomatoes	1	2	3	4	5	6	7
Detection of diseases in tomato supplied	1	2	3	4	5	6	7
Tomato purchase price volatility	1	2	3	4	5	6	7
Imbalance of power with tomato suppliers	1	2	3	4	5	6	7

Table

Challenges related to tomato sales
Low tomato sales prices	1	2	3	4	5	6	7
Instability in volume of tomato demand	1	2	3	4	5	6	7
Low tomato demand	1	2	3	4	5	6	7
Lack of capacity to meet the high volume of tomato demand	1	2	3	4	5	6	7
Inability to meet the high quality specification demanded by customers	1	2	3	4	5	6	7
Instability in quality specifications demanded by customers	1	2	3	4	5	6	7
Tomato sales price volatility	1	2	3	4	5	6	7
Imbalance of power with customers	1	2	3	4	5	6	7
Poor on-time payment for tomato deliveries	1	2	3	4	5	6	7

Part III – Price risk perceptions

6	What is the percentage price deviation from an expected tomato price level that you perceive as price volatility? Please provide the percentage price deviation with respect to the periods that prices are mostly set in the tomato supply chain.

7	Please describe the periods in which you experienced tomato price volatility in the past 5–7 years. Provide examples of how tomato prices were behaving in those periods.

8	In your opinion, when does tomato price volatility become risky?

Part IV – Price risk management strategies

9	What strategies did you use to manage the risk from the observed price volatility described in q.7?

10	What strategies would you use to manage the risk from future tomato price volatility?

Appendix A

Supplementary data

Supplementary data associated with this article can be found, in the online version, at http://doi:10.1016/j.njas.2016.11.002.