44VEGETABLE PRICES IN SERBIA - TENDENCIES AND FORECASTING
Sumadinka Mihajlovic1, Natasa Vukelic2, Nebojsa Novkovic3, Beba Mutavdzic4 Corresponding author E-mail: vukelicn@polj.uns.ac.rs
A R T I C L E I N F O Original Article Received: 05 June 2019 Accepted: 15 June 2019 doi:10.5937/ekoPolj1902485S UDC 338.5.01:635.1/.8(497.11) Keywords:
vegetables, price, Serbia, forecasting
JEL: C53 Q11
A B S T R A C T
The subject of this paper is analysis of the tendencies and forecast of the prices of most significant vegetable crops in Serbia: potato, bean, tomato, pepper, onion, cabbage and watermelon. The aim of the paper is to forecast the absolute prices of the studied vegetables. Time series analysis of vegetable prices expressed in euro per ton (2002-17) was performed by means of descriptive statistics, while adequate ARIMA models were used for price forecasting (2018-22). The analysis ofthe studied vegetable crops showed that bean had the highest average annual price, while watermelon had the lowest. The price of tomato showed the highest fluctuations over the years, while the lowest were for onion and cabbage. All vegetable crops showed a tendency of absolute increase in prices expressed in euro. Based on the foregoing, it can be concluded that the market position of vegetables is generally improving, but oscillations will continue to occur.
© 2019 EA. All rights reserved.
Introduction
Vegetable production is one of the most intensive branches of plant production, and along with grain production, it is one of the most intensive branches of arable land production. This is confirmed both by the yields produced per unit of area, i.e. the amount of organic matter produced annually per unit of area, and by achieved economic
1 Sumadinka Mihajlovic, Mr.Sci., Ph.D. student, University of Novi Sad, Faculty of Agriculture, Serbia, Phone: +38163287971, E-mail:mihajlovic.sumadinka@gmail.com, https://orcid.org/0000-0003-0670-786X
2 Natasa Vukelic, Ph.D., Assistant Professor, University of Novi Sad, Faculty of Agriculture, Trg Dositeja Obradovica 8, Phone: +381214853392, E-mail: vukelicn@polj.uns.ac.rs, Serbia ORCID ID (https://orcid.org/0000-0001-7516-9204)
3 Nebojsa Novkovic, Ph.D., Full Professor, University of Novi Sad, Faculty of Agriculture, Trg D. Obradovica 8, Novi Sad, Serbia, Phone: +38162200132, E-mail: nesann@polj.uns.ac.rs, https://orcid.org/0000-0003-2419-5765
4 Beba Mutavdzic, Ph.D., Associate Professor, University of Novi Sad, Faculty of Agriculture, Trg D. Obradovica 8, Novi Sad, Serbia, Phone: +38162200133, E-mail: bebam@polj.uns.ac.rs, https://orcid.org/0000-0002-7631-0465
effects. Bearing in mind the importance that this branch of agriculture has in both the production and economic sense for producers as well as for agriculture as a whole, it is justified to expect its further development. The subject of the research in this paper is analysis of the tendencies and forecast of the prices of most significant vegetable crops in Serbia: potato, bean, tomato, pepper, onion, cabbage and watermelon. The aim of the paper is to use time series analysis of vegetable prices from the past period as a basis for forecasting the absolute prices of these vegetables expressed in euro for the future period and forecasting economic (market) conditions for production of these crops.
There are numerous examples of applying quantitative and qualitative methods in analysing, modelling, forecasting and planning of production and economic characteristics of agricultural products and inputs in agriculture. Bannikova et al. (2018) analyzed the alternatives of development of the Russian vegetables market concerning the changes of the level and structure of production and consumption of vegetables. The main objectives of the research were to collect and analyze data of the Russian market of vegetable production, modeling and scenario forecasting vegetables market, a substantiation of directions of development of the market. Mutavdzic et al. (2007) analysed the tendencies and forecast the movements in price parities of fattening pigs and commercial maize. Novkovic et al. (2008) analysed the possibilities for future development of vegetable production in Serbia and Vojvodina using the SWOT analysis. Vukelic, Novkovic (2009) analysed the economic results of milk production on large family farms. Husemann, Novkovic, (2014) defined a quantitative model for managing a multifunctional farm. Mutavdzic et al. (2010) focused on forecasting of price parities of the main field crops based on time series analysis and the application of the ARIMA model. Mutavdzic et al., (2017) analysed quarterly movements of wheat and maize retail prices in Serbia and the Republic of Srpska in the period 2010-15. By applying the method of ratio to the overall quarterly average, the results showed that the prices of grains in the Republic of Srpska are higher. Ivanisevic et al. (2015) analysed the movements of tomato prices in Serbia using the method of descriptive statistics, followed by forecasting its value in the future period based on time series analysis. Jasinthan et al. (2015) by using a Markov chain model analyzed and predicted vegetable price movement in Jaffna. Novkovic, Mutavdzic (2016) performed the analysis of bean prices in Serbia by means of descriptive statistics. On the basis of these results, an adequate ARIMA model was applied to forecast the movements of bean prices for the following period. Mutavdzic et al.(2011) analysed the tendencies in development of vegetable production in Serbia, concluding that in the period 2001-10 the total vegetable production in Serbia significantly increased, primarily as a result of intensification of production, i.e. yield increase. The study showed the following average annual increase in production: pea 56%, pepper 26%, carrot 20%, potato 18%, cucumber 17%, cabbage and kale 13%, watermelon 12%, tomato and onion 5% and garlic 2%. Increasing trends in production were found for the following vegetables: tomato, pea, onion, pepper, bean, carrot and cucumber. Decreased trends in production were determined for potato, watermelon and garlic, while cabbage and kale showed the
general tendency of stagnation in production. Novkovic et al. (2013) focused on the analysis and tendencies of development of vegetable production in Vojvodina. In the period 2001-10 the harvested areas of the studied vegetables were reduced for almost all vegetable crops, except for pea, pepper and garlic, for which the harvested area was slightly larger compared to the previous decade. The yields of all studied vegetable crops increased (except for tomatoes) and the total vegetable production significantly increased as a result of intensification of production.
Materials and methods
The research methods applied in this paper were selected based on the described subject and aim of the research. The statistical methods included descriptive statistics and time series analysis. Descriptive statistics was used for analysis of the vegetable prices in the studied period. Forecasting of the vegetable prices was carried out using the ARIMA models, based on time series analysis. Time series analysis was conducted using the prices of seven major vegetable crops in Serbia (potato, bean, tomato, pepper, onion, cabbage and watermelon). The average annual vegetable prices in the analysis were converted into euro per tonne to enable comparison with foreign countries and to reduce the factor of domestic inflation. The absolute vegetable prices were analysed for the period 2002-17, starting thus from the year when euro entered into circulation. Conversion of the prices into euro was carried out according to the average annual exchange rate of euro based on the data of the National Bank of Serbia. Since there were shorter time series, the forecast of the vegetable prices was made for a period of five years: 2018-2022. The series of studied phenomena in this paper were either taken entirely or formed on the basis of statistical publications of the Statistical Office of the Republic of Serbia. Statistical software used for the analysis of the collected data included Statistica 10, Eviews 3.1 and SPSS.
Results
Analysis and forecast of potato prices
The average potato price in the period 2002-17 was 183.3 EUR/t. The price ranged from 84.5 EUR/t in 2005 to 249 EUR/t in 2013. The coefficient of variation was relatively high: 27%. The average annual price of potato had a rather pronounced tendency of growth, at an average annual rate of 3.83%. Such relatively high growth rate of potato price is an indicator of the improvement of its absolute position on the market. The analysis and forecasting model shows that potato price in a certain year is significantly influenced by random processes from the preceding two periods (Table 1).
On the basis of the estimated model, potato prices for the period 2018-22 were forecast (Table 2), indicating that in the following five year-period potato price will fluctuate over the years (which was also the case in the analysed period). These findings are illustrated by a graphical representation of potato price movements (Figure 1).
Table 1. Model parameters for forecasting potato prices
Paramet. Input: Krompiret: Transformations: D(1) Model:(1,1,2) MS Residual= 2943,9
Param. Asympt. Std.Err. Asympt. t( 12) p Lower 95% Conf Upper 95% Conf
p(1) -0,694871 -0,266954 0,552352 0,379337 -1,83180 0,091904 -1,52138 0,131634 0,460227 1,046666
q(1) 0,333751 -0,79986 0,439339 -0,99414
q(2) 0,226873 2,43463 0,031465 0,05804
Table 2. Forecast of potato prices (2018-22)
CaseNo. Forecasts; Model:(1,1,2) Input: Krompiret: Start of origin: 1 End of origin: 16
Forecast Lower 95,0000% Upper 95,0000% Std.Err.
17 200,3884 82,17182 318,6050 54,25735
18 188,5011 52,30655 324,6956 62,50860
19 196,7612 55,50266 338,0198 64,83283
20 191,0215 38,15485 343,8881 70,16053
21 195,0099 35,97128 354,0485 72,99324
Figure 1. Changes in potato prices
Forecasts; Model:(1,1,2) Input: Krompiret
-Observed - Forecast -± 95,0000%
Analysis and forecast of bean prices
In the analysed period, the average annual price of bean was 1,333.6 EUR/t, ranging from 948.4 EUR/t in 2004 to 2,213.2 EUR/t in 2014. The coefficient of variation was, similarly to potato, moderately high amounting to 28.5%. The absolute average annual price of bean also showed a tendency of increase, but it was slightly lower compared to potato. The average annual growth rate of bean was 1.33%, which means that bean showed the tendencies of slight improvement of its absolute price (economic) position on the market. Bean production is characterised by oscillations, which is reflected also in the prices of this crop. Bean price in a certain year is influenced by the price from the preceding year, and it is statistically significantly influenced by a random process from the preceding two years (Table 3).
Table 3. Model parameters for forecasting bean prices
Input: PASULJET: Transformations: D(1) Model:(1,1,2) MS Residual = 79536,
Paramet. Param. Asympt. Std.Err. Asympt. t( 11) p Lower 95% Conf Upper 95% Conf
Constant 17,18598 18,88401 9,100810E-01 0,382289 -24,3774 58,74939
p(1) 0,27152 0,32447 8,368215E-01 0,420501 -0,4426 0,98567
q(1) 0,32940 0,00000 2,011084E+16 0,000000 0,3294 0,32940
q(2) 0,67050 0,00000 5,039653E+32 0,000000 0,6705 0,67050
The estimated model provided the forecast values of bean prices for the five-year period (Table 4), showing that bean price will alternatively decline and grow over the years. The graphical representation of changes in bean prices is given in Figure 2.
Table 4. Forecast of bean prices (2018-22)
CaseNo. Forecasts; Model:(1,1,2) Input: PASULJET: Start of origin: 1 End of origin: 16
Forecast Lower 95,0000% Upper 95,0000% Std.Err.
17 1596,261 975,5357 2216,986 282,0215
18 1487,382 634,5705 2340,194 387,4682
19 1470,339 602,8603 2337,818 394,1321
20 1478,231 609,6779 2346,785 394,6203
21 1492,894 624,2604 2361,527 394,6566
Figure 2. Changes in bean prices
Forecasts; Model:(1,1,2) Input: PASULJET
0 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 - Observed - Forecast - ± 95,0000%
Analysis and forecast of tomato prices
The average annual price of tomato in the analysed period was 298.2 EUR/t. The price varied within the range of 106.6 EUR/t in the first year of the analysed period (2002) to 564.9 EUR/t in 2014. The coefficient of variation of the average annual tomato price was extremely high amounting to 44.6%. The average annual rate of price change was the highest for tomato compared to other analysed vegetable crops, amounting to 9.1%. This means that tomato had the most pronounced tendency of price growth of all analysed vegetable crops, i.e. it had the tendency of the greatest improvement of economic (price) conditions for its production. Unlike the potato prices, the forecast price of tomato in the following five years showed a tendency of increase. The forecast values were obtained on the basis of the estimated model (Table 5), which shows that tomato price in a current year is significantly influenced by its price in the preceding year. The tendency of increasing prices in the following period is illustrated graphically (Figure 3).
Table 5. Forecast of tomato prices (2018-22)
CaseNo. Forecasts; Model:(1,1,0) Input: Paradajzet Start of origin: 1 End of origin: 16
Forecast Lower 95,0000% Upper 95,0000% Std.Err.
17 444,9676 223,4167 666,5184 102,5523
18 441,5752 213,7784 669,3720 105,4435
19 479,8995 188,7984 771,0006 134,7460
20 486,4824 183,6015 789,3633 140,1987
21 517,2167 175,6456 858,7877 158,1078
Figure 3. Changes in tomato prices
Forecasts; Model:(1,1,0) Input: Paradajzet
0 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 -Observed - Forecast -± 95,0000%
Analysis and forecasting of pepper prices
In the analysed period, the average annual price of pepper was 310.5 EUR/t. The price varied within the range from 188.3 EUR/t in 2006 to 447.8 EUR/t in the last year of the analysed period (2017). The coefficient of variation of the average annual price of pepper was high, amounting to 31.3%. The average annual rate of change in pepper prices was very high and amounted to 5.87%. This means that pepper also had a significant tendency of absolute improvement of its economic position. Pepper price in a certain period was, as was the case with most other analysed vegetable crops, influenced by its
prices in the preceding two years. The estimated model forecast pepper prices for the following five years (Table 6), showing that pepper price is expected to continuously increase over the years to the end of the forecast period (by 2022). The movements of pepper prices are given in Figure 4.
Table 6. Forecast of pepper prices (2018-22)
CaseNo. Forecasts; Model:(2,1,0) Input: Paprikaet: Start of origin: 1 End of origin: 16
Forecast Lower 95,0000% Upper 95,0000% Std.Err.
17 479,5019 359,3420 599,6619 55,14926
18 470,8765 339,5894 602,1637 60,25628
19 494,4590 361,6839 627,2341 60,93921
20 523,7226 369,7675 677,6776 70,66005
21 530,8480 363,7849 697,9110 76,67618
Figure 4. Changes in pepper prices
800 700 600 500 400 Forecasts; Model:(2,1,0) Input: Paprikaet 800 700 600 500 400
300 200 100 0 300 200 100
2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 - Observed - Forecast -± 95,0000%
Analysis and forecast of onion prices
The average annual price of onion in the observed period amounted to 178.2 EUR/t, ranging from 119 EUR/t in 2004 to 270.7 EUR/t in 2011. The variation coefficient of the average annual price of onion was moderately high (but the lowest compared to
other analysed vegetables) and it amounted to 22%. The average annual rate of change in onion price was positive and amounted to 1.63%. This means that onion had a slight tendency of price increase in the analysed period, i.e. there was a tendency of slight improvement in the economic (price) conditions for its production.
Based on the prices in the period 2002-17, the estimated model showed that onion prices in an observed year were significantly influenced by the prices from the preceding two years. On the basis of the model, forecasting of onion prices was made for the period 2018-22 (Table 7), indicating that onion prices will continuously increase over the years during the forecast period. This is illustrated by graphical representation of price movements in the analysed and forecast period (Figure 5).
Table 7. Forecast of onion prices (2018-22)
CaseNo. Forecasts; Model:(2,1,0) Input: Crniluket: Start of origin: 2 End of origin: 15
Forecast Lower 95,0000% Upper 95,0000% Std.Err.
16 189,2947 154,1689 224,4205 15,76465
17 193,8105 135,6749 251,9461 26,09157
18 195,5434 131,2175 259,8693 28,86979
19 196,0136 130,0927 261,9345 29,58564
20 197,8529 128,2964 267,4094 31,21731
Analysis and forecast of cabbage prices
In the analysed period, the average annual price of cabbage was 158.8 EUR/t. The price ranged from 80.5 EUR/t in 2004 to 212.5 EUR/t in 2007. The coefficient of variation of the average annual price of cabbage was moderately high and amounted to 22.3%. The average annual rate of change in cabbage price was slightly positive and amounted to 1.48%. This means that cabbage had a slight tendency of absolute improvement of its economic (price) position. Cabbage price in a certain year was statistically significantly influenced by prices from the preceding period. Based on the estimated model, price movements were forecast for the following five years (Table 8), showing that the price of cabbage will fluctuate over the years, i.e. it will alternatively decrease and increase. These tendencies are confirmed by the graphical representation of price movements in the analysed and forecast period (Figure 6).
Figure 5. Changes in onion prices
300 280 260 240 220 200 180 160 140 120 100 0 Forecasts; Model:(2,1,0) Input: Crniluket 300 280 260 240 220 200 180 160 140 120 100
2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 -Observed - Forecast -± 95,0000%
Table 8. Forecast of cabbage prices (2018-22)
CaseNo. Forecasts; Model:(1,1,0) Input: KIKET Start of origin: 1 End of origin: 16
Forecast Lower 95,0000% Upper 95,0000% Std.Err.
17 156,3881 62,31970 250,4565 43,54276
18 169,6683 67,96312 271,3735 47,07769
19 164,9611 40,75553 289,1666 57,49276
20 170,8477 36,16592 305,5294 62,34202
21 170,4950 21,63835 319,3516 68,90334
Figure 6. Changes in cabbage prices
Forecasts; Model:(1,1,0) Input: KIKET
Analysis and forecast of watermelon prices
The average annual price of watermelon was 118.1 EUR/t. The price of watermelon varied from 68.4 EUR/t in 2004 to 213 EUR/t in 2010. The variation coefficient of the average annual price of watermelon was moderately high, amounting to 31.1%. The average annual rate of change in watermelon price was moderately high, amounting to 2.84%. This means that watermelon had a pronounced tendency of absolute improvement of its economic (price) position. The estimated model for analysing and forecasting watermelon prices showed that the price in a certain year was influenced by prices from the preceding two years, while the influence of the price from the previous year was statistically significant. On the basis of the estimated model, price movements of watermelon were forecast for the following five-year period (Table 9), showing that in the initial years the price will have a decreasing tendency, i.e. it will decrease in the first three years of the forecast period, while in the last two years of this period it is expected that watermelon price will increase. The indicated characteristics of watermelon price movements, especially the forecast values, are confirmed by the graphical representation of these movements (Figure 7).
Table 9. Forecast of watermelon prices (2018-22)
CaseNo. Forecasts; Model:(2,1,0) Input: DILET : Start of origin: 1 End of origin: 16
Forecast Lower 95,0000% Upper 95,0000% Std.Err.
17 138,1975 48,98740 227,4075 40,94434
18 137,7790 43,75934 231,7986 43,15177
19 132,1873 33,50217 230,8725 45,29308
20 143,3726 29,86298 256,8823 52,09703
21 145,6496 25,77610 265,5231 55,01780
Figure 7. Changes in watermelon prices
300 250 200 150 Forecasts; Model:(2,1,0) Input: DILET 300 250 200 150
100 50 0 0 100 50 0
2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 -Observed - Forecast -± 95,0000%
Discussion and Conclusion
The analysis which included seven vegetable crops indicated that in the period 2002-17 in Serbia bean had the highest average annual price, while watermelon had the lowest average annual price. The highest price fluctuations on the annual basis were found for tomatoes, while onion and cabbage had the lowest fluctuations (twice as low). All vegetable crops showed the tendency of absolute increase in prices expressed in euro. By far the highest average annual growth rate was found for tomato, whereas it was
the lowest for beans. Considering the prices of individual vegetable crops in the period 2018-22, the results of the forecast are as follows:
The price of potato will fluctuate over the years (which was also the case during the observed period). The price of this crop will range from 200 EUR/t in 2018 to 191 EUR/t in 2021.
The average price of bean will alternatively decline and grow over the years, within the range from 1.596 EUR/t in 2018 to 1.470 EUR/t in 2020.
Unlike the prices of potatoes and beans, the forecast price of tomato shows a tendency of increase from 441 EUR/t in 2019 to 517 EUR/t in 2022.
Continuous price growth was forecast for pepper throughout the whole forecast period: from 470 EUR/t to 530 EUR/t in 2022.
Continuous price growth was also forecast for onion: from 189 EUR/t in 2018 to 198 EUR/t in 2022.
The forecast values of cabbage price show that there will be minor oscillations over the years, i.e. the price will alternatively increase and decrease, but with a positive tendency. The price of cabbage will range from 156 EUR/t in 2018 to 170 EUR/t in 2022.
The price of watermelon will have a tendency of decrease in the initial years (the first three years of the forecast period), while in the last two years the price is expected to increase. The price of this crop will vary from 132 EUR/t in 2020 to 145 EUR/t in 2022.
Based on presented analysis, it can be concluded that the market position of vegetables is generally improving, but there will still be fluctuations and variability of their prices over the years.
Conflict of interests
The authors declare no conflict of interest.
References
1. Bannikova V. N, Onezhkina N. O., Agalarova G. E., Tenishchev V. A. (2018): Forecasting the tendencies of the Russian vegetables market development, Journal of Business and Retail Management Research (JBRMR), 13 (1), 148 - 155.
2. Husemann, Ch., Novkovic, N. (2014): Farm management information systems: A case study of on a German multifunctional farm, Economics of Agriculture 61 (2), (269-552), 441-453.
3. Ivanisevic, A., Mutavdzic, B., Novkovic, N., Vukelic, N. (2015): Analysis and prediction of tomato price in Serbia, Economics of Agriculture 62 (4) (899-1178), 951-961.
4. Jasinthan, P., Laheetharan, A., Satkunanathan, N. (2015): A Markov Chain Model for Vegetable Price Movement in Jaffna , Sri Lankan Journal of Applied Statistics, Vol (16-2), 93-105.
5. Mutavdzic B., Drinic Lj., Novakovic , T., Vasko Z., Novkovic, N. (2017): The Comparative Analysis of Grain Prices in Serbia and Republic of Srpska, Book of Abstracts, 6th International Symposium on Agricultural Sciences, University of Banja Luka, February 27 - March 2, 2017 Banja Luka, Bosnia and Herzegovina, ISBN 978-99938-93-41-7, 54.
6. Mutavdzic B., Novkovic, N., Nikolic-Doric E., Radojevic, V. (2007): Analysis and prediction of pig price parity - maize, Conteporary Agriculture 1-2, 177-181. [in Serbian: МутавциЙ Б., НовковиЙ, Н., НиколиЙ-ЪориЙ Е., РадоjевиЙ, В. (2007): Анализа и предви^аае паритета цена свиае - кукуруз].
7. Mutavdzic Beba, Novkovic N., Ivanisevic D. (2011): Tendencies of vegetable development in Serbia, Agroznanje 12 (1), University of Banja Luka, Faculty of Agriculture, 23-31. [in Serbian: МутавциЙ Беба, НовковиЙ Н., ИванишевиЙ Д. (2011): Тенденцще развоjа повртарства у Срби]и].
8. Mutavdzic, B., Novkovic, N., Ivanisevic, D. (2010): Prediction of parity of prices of basic agricultural products, AroSym, Faculty of Agriculture East Sarajevo, Jahorina, 176-182. [in Serbian: МутавциЙ, Б., НовковиЙ, Н., ИванишевиЙ, Д. (2010): Предви^аае паритета цена основних ратарских производа].
9. Novkovic N., Ilin Z, Janosevic M., Mutavdzic B. (2008): Importance of vegetable production for the multifunctional rural development, Multifunctional agriculture and rural development, IEP, Book I, Belgrade, 141-148. [in Serbian: НовковиЙ Н., лин Ж, JаношевиЙ М., МутавциЙ Б. (2008), Значаj производае поврЙа за мултифункционални рурални развод зборник радова ме^ународног научног скупа, Мултифункционална полопривреда и рурални развоj III"].
10. Novkovic, N., Mutavdzic B. (2016): Analysis and forecasting of bean prices in Serbia, Proceedings of papers: Policy and Economics for Sustainable Agricultural and Rural Development, AAEM 10th International Conference, 12-14 May, Ohrid, Association of Agricultural Economists of the Republic of Macedonia, 195-203.
11. Novkovic, N., Mutavdzic B., Ilin, Z., Ivanisevic, D. (2013): Potato Production Forecasting, Agroznanje 14 (3); Banja Luka, 345-355.
12. Vukelic N., Novkovic, N. (2009): Economic results of milk production on large households, Economics of Agriculture, 56, 99-110 [in Serbian: ВукелиЙ Н., НовковиЙ, Н. (2009), Економски резултати производае млека на крупним селачким газдинствима].
