INTRODUCTION
⌅World olive oil production has increased significantly in the last few decades (Kashiwagi et al., 2020). Traditional European supplier countries are expected to focus increasingly on quality and sustainability rather than quantity, so in parallel with a slower growth of these traditional suppliers, a substantial increase in future production in new producing countries is foreseen (Mili & Bouhaddane, 2021); in the same time, a significant increase in global demand for olive oil in non-traditional markets is expected. Among the most important reasons for this increase, the following could be cited: changing lifestyles; rising incomes of different consumer segments; higher demand from pharmaceutical firms for personal skin and hair care products; and also an increasing awareness of the positive health and sustainability attributes of the olive oil (Klonaris & Agiangkatzoglou, 2018; Kashiwagi et al., 2020; Mili & Bouhaddane, 2021).
Olive tree (Olea europaea (syn. Zaytoun, Jetun)) has been cultivated alongside Mediterranean civilizations for more than 6000 years. There are evidences of olive oil production plants cultivation since the Bronce Age in different Mediterranean countries, such as Grece, Israel, Palestine Cyprus and Syria (Vossen, 2007; Kapellakis et al., 2008). The countries illustrated in Table 1 represented more than the 85% of the total world oil production in 2019. The total production was concentrated in the following European Union (EU) countries: Spain, Italy, and Greece (with around the 36%, 10.8% and 9.7% of the total production respectively). Second-tier non-European producers include Türkiye, Tunisia, Morocco, and Syria (with around the 7.3%, 6.7%, 6.6% and 5.4% of the production respectively), suited by Algeria, Egypt, Lebanon, and Jordan. The following most significant producer countries are Argentina (with around 1% of the total) and Chile (with around 0.6% of the total). However, the production of Latin American countries is very low compared with the Mediterranean basin production countries. Although there are other producers in the Latin American region (e.g. Mexico, Peru, and Uruguay), their production percentage is even minor compared to Argentina and Chile (FAOSTAT, 2022).
Year | 2019 | 1961 | % Δ |
---|---|---|---|
Argentina | 1.0 | 0.3 | 262.5 |
Chile | 0.6 | 0.2 | 187 |
Greece | 9.7 | 16.6 | -41.4 |
Italy | 10.9 | 29.0 | -62.6 |
Morocco | 6.6 | 1.0 | 543.5 |
Spain | 36.8 | 27.1 | 36.2 |
Syrian | 5.4 | 0 | 5431.4 |
Tunisia | 6.8 | 2.8 | 140.9 |
Türkiye | 7.3 | 9.6 | -23.6 |
Rest of world % | 14.9 | 13.4 | 10.7 |
As explained by Milli & Zúñiga (2001), traditionally, the countries of the Mediterranean Basin were responsible for 95% of world production and 90% of world consumption. However, fundamental changes have occurred during the nineties, which are gradually altering these traditional patterns. Table 1 illustrates some relevant changes that have occurred in the olive oil production picture from 1961 to 2019. Among the most important EU producer countries, only Spain has increased its production from 1961 to 2019 (~ 36%), while Italy and Greece have reduced its production (~ 62% and ~ 41%, respectively). Among the second most important producers, Morocco has increased considerably its production (~ 543%) suited by Tunisia (~ 140%), while Türkiye has reduced its production in the cited period in a 23%. As regards the Latin American countries, as mentioned above, Argentina is the most important producer country, which a remarkable production increase (of 262% from 1961 to 2019) (see Table 1). As explained by Pefaur (2015), in Chile, the national olive industry was formed around 1950, but it was not until 1996 that a competitive Chilean olive growing industry began to develop. As illustrated in Table 1, its production increase has been also remarkable (~ 186%), well above the global average for the cited period (~ 10%).
The standard way to study competitiveness is to make country-level assessments for a product, product group or industry, using international trade data (exports, imports, trade balance and prices) (Laursen, 2015). One of the most commonly used analyses is the Balassa indicator of Revealed Comparative Advantage (RCA) and its adaptations based on trade flows (Bojnec, 2014; Laursen, 2015; Stellian & Danna-Buitrago, 2019). There are other techniques such as constant market share analysis (CMS), the gravity model and the intra-industry trade index. In all of them the main analysis variables are the export and import variables between countries, with some including the analysis of prices and other factors that indicate performance at the country level (Klonaris & Agiangkatzoglou, 2018; Stellian & Danna-Buitrago, 2019; Kashiwagi et al., 2020). Differences in competitiveness among exporters are attributable to changes in trade position due to comparative advantages or market circumstances, with greater differences existing between traditional and emerging producers (Klonaris & Agiangkatzoglou, 2018; Kashiwagi et al., 2020). An example of competitiveness assessment was the study on Türkiye versus the main olive oil exporting countries for the period 1990-2006 using aforesaid mentioned RCA, the Comparative Export Performance (CEP) and the Market Share Index (MSI) (Türkekul et al., 2010). Another study focused on the competitiveness of Greek olive oil with the main European competitors in the markets of Germany, Italy, the UK and the USA, using the RCA index and performing an import demand system in each market (Klonaris & Agiangkatzoglou, 2018). Other authors highlight that olive oil production is expanding to non-traditional producing countries, such as the USA, Australia, and New Zealand. The authors analyze the factors affecting olive oil exports and imports in Mediterranean countries for the period 1998-2016 using the product-specific gravity model (Kashiwagi et al., 2020). In general, studies on international olive oil trade propose an approach to the global competitiveness of exporters using trade specialization.
As regards the analysis of the olive oil international market, there are some fundamental characteristics that should be kept in mind, especially such aspects related to the origin and quality of the product, and the form in which it is traded (i.e. in bulk or bottled). As explained by Bajoub et al. (2018), over the last decades, mainly due to the increasing worldwide popularity and the trade globalization of the olive oil, quality as well as authenticity control have become a relevant issue to the different agents in the market (i.e. consumers, suppliers, retailers, and regulators), not only in traditional, but also in emerging olive oil producing countries. According to Aparicio et al. (2013), the purpose of olive oil authenticity should be focused on the labeling control as well as on protecting the genuineness of olive oils with regard to their geographical origin and botanical variety.
Main traders of olive oil have different attributes to consider within their commercial strategies. For example, in the USA, Germany and Mexico, consumer preferences show that the most valued attributes are origin and price (Chamorro-Mera et al., 2020), while Germany has a strong preference for a higher quality (Scarpa et al., 2021), and Italy favors attributes of quality that are revealed by labels stating geographical indications and organic production (Gorgitano & Sodano, 2016; Roselli et al., 2017). As regards the form in which the olive oil is traded (i.e., in bulk or bottled), main destinations for Chilean oil are the USA, Brazil and Italy. To USA and Brazil, the product is exported in bottled form (implying a higher price per unit), instead to Italy it is exported mainly in bulk, because it enters free of duty, which is not the case with bottled oil (Pefaur, 2015). As explained by Mili & Zuñiga (2001), until 2001, a high proportion of the Spanish exports still were of bulk, with a low value added. Being Italy the second most important producer and exporter (in value), as pointed out by different authors, it is also one of the world’s leading importers of bulk olive oil. Its main suppliers are Spain, Greece, Tunisia, Türkiye and Morocco (Klonaris & Agiangkatzoglou, 2018; Mili & Bouhaddane, 2021). As pointed out by Klonaris & Agiangkatzoglou (2018), the traditional strategy of different Italian olive oil processing companies consists of importing oil in bulk from different origins and qualities, thereafter they blend it producing a branded product, which is re-exported as an Italian product.
There is little research for Latin America as regards the olive oil trade. For these countries, the agricultural sector provides a vital contribution to the generation of employment and growth. To ensure their growth and sustainability, it is important to analyze the performance of these industries in international trade (Losilla et al., 2019; Paus, 2019). Accordingly, studies that analyze trade relations and the evolution of the competitiveness of this type of good can help the sector’s stakeholders develop business strategies. Considering that Argentina and Chile are emerging producers and Brazil is among their top buyers worldwide, studies that analyze the performance of these countries in the international olive oil market with international trade data are essential to assess and design trade strategies. This type of research is an instrument to promote the development of the value-added industry in Latin America (Pajares et al., 2014).
In this research, priority was given to the use of methodologies other than those traditionally used, namely CMS and RCA, to propose and validate new tools that contribute to a better understanding of trade competitiveness. Through neural networks it was possible to represent the complex flows in the olive oil trade, as well as employing competitiveness matrices to show dynamic indicators over time. Neural networks are proving to be a useful tool in various fields of science because they provide a quick, intuitive visualization to analyze complex relationships in terms of quantities of relationships and the magnitudes of these relationships.
Neural networks are quite frequently applied to financial markets and trade (Ballestra et al., 2019; Eachempati et al., 2021; Mateńczuk et al., 2021). One of the most common applications of neural networks is data prediction models, e.g., future stock market prices, calculated based on historical data (Mateńczuk et al., 2021). These methods are also increasingly applied to international trade of goods and services. Using trade data of ten countries, a recent study by Shen et al. (2021), developed a foreign trade forecasting method that relies on a neural network with long short-term memory (LSTM). The main goal in the study by Sokolov-Mladenović et al. (2016), was to predict economic growth based on trade in services, exports of goods and services, imports of goods and services, and merchandise trade based on artificial neural networks.
The objective of this study was to analyze the performance of emerging Latin American countries in the world olive oil market for the period 2010-2019.
MATERIAL AND METHODS
⌅Data
⌅For the general characterization of the olive oil industry, trade data (COMTRADE, 2022) for category 1509 “Olive oil and its fractions, whether or not refined, but not chemically modified” was used. Specific information relating to tariff code 150910 Olive oil, virgin (including extra virgin, virgin and lampante virgin, unfit for human consumption), and code 150990 “Olive oil (excl. crude & virgin) & fractions thereof, whether or not refined, but not chemically modified”, i.e., mostly refined oil, was also used. Olive oil is a very heterogeneous product in terms of its qualities, including extraction method, organoleptic defects (aroma and flavor), and degree of free acidity (expressed as oleic acid). The harmonized commodity description and coding system (HS) prior to the 2017 and 2022 updates does not offer a disaggregation that allows the precise identification of the oil marketed according to its characteristics. Therefore, category 150910 does not allow differentiation between lampante olive oil, virgin olive oil and extra virgin olive oil. In order to improve the understanding of the results obtained, data from the customs offices of those countries that offer a more detailed classification were consulted.
The data used for exports and imports, expressed in tons (volume) and thousands of dollars (value), as well as the unit or average value (dollars per ton), were obtained from the United Nations Comtrade online database (COMTRADE, 2022). The data sample covers the period 2010-2019; however, in some indicators only data up to 2018 are used, due to the unavailability of information for 2019 or a high variability of the information for this year with respect to the rest of the series.
To assess competitiveness, the markets in which Argentine and Chilean exports compete were chosen. These markets were also considered to be representative of the main geographical areas and which play a leading role in global imports of this good (top 10). This is the case of USA, Brazil, Japan, and Spain.
METHODS
⌅Firstly, an analysis of the relationship between the main exporters and importers of olive oil was carried out, using a mesh graph made with UCINET software. This is an application that can represent networks and obtain indicators for analysis (Song et al., 2020). This case will provide a simplified way to show the main trade flows for the analyzed good.
Export efficiency
⌅To analyze the overall export performance of Chile and Argentina with respect to the major exporting countries, a decomposition of export value growth was applied. This method consists in explaining the growth rate of the exported value by decomposing the growth rate of the exported quantity and the growth rate of the average unit value exported (hereinafter “price”) (Guevara et al., 2021). Eq. 1 is proposed for a better understanding:
where GRvj: growth rate of the value exported by country j; n: number of periods analyzed; Xkj: olive oil exports in period k (initial) by country j; X(k+nj): exports of olive oil in the period k+n (present) by country j; Pkj: olive oil price in period k (initial) by country j; and P(k+nj): olive oil price in the period k+n (present) by country j
Import dependency
⌅On the other hand, an index that evaluates the dependence of exporting countries on imports was applied. This was calculated using Eq. 2, which relates the imports of the good (Mi) divided by the exports of the good itself (Xi) for the country analyzed.
where IDij: dependence on imports to exports of good i in country j.
In general, this indicator evaluates what proportion of exports might hypothetically come from imports. Although these imports could also go to domestic consumption, they would still allow a greater quantity of the good to be available for export. If the value of the indicator is close to zero it means that the country does not depend on imports, whereas a value close to 100 indicates heavy dependence, and a higher value indicates that all exports and a part of domestic consumption are conditioned by imports.
To understand the importing behavior of exporting countries, a price analysis was carried out consisting of the difference between the export price (Pije) and the import price (Piji). If the result is negative, it indicates that the imports would be destined for domestic consumption for a market segment willing to pay for a better-quality product. On the contrary, if the result is positive, it indicates that the country could obtain a benefit by re-exporting the good, or by substituting domestic consumption with imported oil of lower value, devoting more of its domestic production to exports. In practice, however, it is impossible to establish the exact traceability of countries’ import-export behavior. Therefore, in order to qualify the most significant results, specific information on the characteristics of imports and exports in these countries is analyzed.
where ωij: net price of product i in country j.
The export-re-export strategy was analyzed by correlating the annual averages of the variables calculated in Eqs. 2 and 3.
Combined indicators to measure competitiveness
⌅Finally, the calculation of import share indicators related to the market share indicator made it possible to assess competitiveness in the main markets selected for code 1509 (Guevara & Morales, 2018). Performance by grade (code 150910 vs. 150990) was also analyzed for the world market.
The share of olive oil imports (SI) was calculated by using Eq. 4:
where SIIZ: share of imports of product i by the importing country z; Xij: exports of good i from country j to the importing country concerned; and Miz: total imports of good i, from importing country z.
On the other hand, the market share (MS) indicator was calculated. One of the most used parameters is the competitiveness of companies or countries with respect to the share of participation in a specific market for a product or service (Damijan et al., 2020):
where MSijz: market share of product i by exporter j by importer z; Xij: exports of good i from country j to the importing country concerned; and Miz: total imports of good i, from importing country z.
The trends of the indicators SI and MS were evaluated by applying a linear regression with respect to time: the slopes of the functions ᶋ(SIiz) and ᶋ(MSijZ). Using the sign of the coefficients, it was possible to determine whether the evolution of the specific MS, and the share of the evaluated product were statistically significant, for a confidence level of 95%, indicating the positive or negative sign if it is increasing or decreasing, respectively.
The competitiveness results evaluated in the markets analyzed allow us to understand the scenario that Chile and Argentina are facing in terms of reformulating their commercial strategies. The competitiveness was obtained from the calculation of the SI and MS indicators, the context of competitiveness used was not based on the magnitude of their results, but rather on the evolution of these over time. Therefore, it was possible to have both a high MS and negative competitiveness (Guevara & Morales, 2018).
The results obtained were represented in a nine-quadrant competitiveness matrix (Guevara & Morales, 2018). This allows representing the evolution of the MS of the exporting country in the importing country and the evolution of the share of the analyzed product in the importing country itself (Fig. S1 [suppl]).