Selection of socks export markets for Turkey using multi-criteria decision making methods
* Author to whom correspondence should be addressed.
Sigma Journal of Engineering and Natural Sciences 2020, Vol. 38, Issue 2, pp. 795-815; doi.org/10.62051/ytu.sigma-journal-of-engineering-and-natural-sciences-selection-of-socks-export-markets-for-turkey-using-multi-criteria-decision-makin
Abstract
Keywords: Socks export; market selection; TOPSIS; GRA; ANP.
1. Introduction
Advances in communication and transportation have driven developing countries to invest heavily in the garment sector since the second half of the 20th century, which, in turn, has resulted in intense global competition (Özbek, 2018). In this intense competitive environment, clothing manufacturing countries have turned to exports to build a strong position in the market and to adapt to the competitive environment (Şarkgüneşi et al., 2017). Exports allows those countries to make the most profit out of their products, extend their service life, avoid the drawbacks of the limited demand in the domestic market, diversify the market, reduce dependence on a single market, increase productivity, gain new opportunities and reduce risks (Shahbandarzadeh and Haghighat, 2010; Shabani et al, 2013; Nunes and De Souza Lequain, *
Corresponding Author: e-mail: aytac.yildiz@btu.edu.tr, tel: (224) 300 37 26 795
2016; Sings and Yaras, 2016; Kalafsky, 2017). Research shows that countries will need to have access to raw materials and an infrastructure to communicate with their customers to be able to export their products in the long term in the future (Uzbek, 2009a; Ragland et al., 2015). The geographical location of Turkey makes it suitable for cotton production (TCAE), 2018). Using its abundant and high quality cotton, Turkey was a successful exporter of cotton and cotton fabrics in the 1970s and a cotton garment exporter in the 1980s (Özbek, 2009b). This trend indicates that it will continue to be an important cotton garment exporter in the future. Denim pants and socks contain high amounts of cotton, and therefore, are strategic products for Turkey. Table 1 shows the distribution of global socks exports across countries. Table 1. Distribution of Global Socks Exports across Countries (1000$) Exporting Countries 2014 2015 China 2.923.900 2.890.102 Germany 263.184 229.055 Italy 271.488 197.854 El Salvador 35.631 36.368 USA 130.715 130.302 Turkey 140.285 128.412 Belgium 115.933 103.273 Netherlands 52.870 52.635 Pakistan 113.914 103.203 France 44.222 37.376 Vietnam 21.582 26.114 Slovakia 49.631 36.457 Poland 49.099 39.965 Thailand 47.873 46.194 United Kingdom 35.905 36.375 Top 15 Countries Total 4.296.232 4.093.685 Top 15 Countries (%) 86,23% 87,09% Global 4.982.508 4.700.455 Source: www.trademap.org database
In 2018, global socks exports were over $ 5 billion, and the top 15 exporters accounted for about 87 percent of total exports. Of those countries, China accounted for 56.73 percent of total exports, holding more than half of the market, followed by Germany, Italy and El Salvador. Turkey is the world's sixth exporter of socks. The aim of this study is to select the best export market for the Turkish socks sector. We believe that this will help Turkey move to the top of the list. For this, criteria and alternative markets are determined, and TOPSIS, GRA and ANP methods are used. Sensitivity analysis is performed on TOPSIS, GRA and ANP data to select the best market. The second section of the study reviewed previous studies on the selection of export markets. The third section explained the materials and methods used. The fourth section addressed the results of the analysis. The final section drew conclusions and made evaluations.
2. Export Market Selection
The significant increase in consumer demand, recent emergence of multinational companies, exponential growth of internet use and the fact that companies see exports as a means of acquiring information about foreign markets have resulted in an increased interest in exports (Shabani and 796
Saen, 2016; Cheng and Huang, 2019). Entering a new export market has become a critical strategy for companies and countries to stay in competition and to grow their production volumes (Azar, 2014). Selecting the right market is, therefore, of paramount significance because it allows companies to make short-, medium- and long-term plans and use time, money and other resources effectively (Shabani and Saen, 2016). Export market selection also plays a key role in the success of companies because it affects target market development, effectiveness of marketing strategies, coordination of external operations and positioning strategies (Farzipoor Saen, 2011). Export market selection involves uncertainty and complexity due to such factors as marketing mix (product, price, distribution and promotion), environmental variables, market share, potential profitability and strategic importance of the market. It is, therefore, a time consuming problem that requires multi-criteria decision-making approaches (Farzipoor Saen, 2011; Shipley et al., 2013; Yavuz, 2016; Cano et al, 2017; Silva et al, 2018). Decision makers need to have sufficient knowledge of potential export markets to select the best one. However, information on markets is generally little and/or of poor quality. Methods based on classical logic or statistics fail to cope effectively with limited information. Therefore, MCDM methods taking expert decisions into consideration are used in multi-criteria decision-making cases (Shipley et al., 2013). Some studies on the problem of export market selection are described below. Wang and Le (2018) used Data Envelopment Method (DEA) to determine the ideal export market for Vietnam. Their export market selection criteria were export, total export, import tariff rates, exchange rates and ease of operation. They reported that low tariff rates, low exchange rates and high ease of operation were the most important criteria for countries and that Vietnam's ideal export markets for the 2014-2017 period were Malaysia, Singapore and the USA. Cano et al. (2017) used the market, logistics, foreign trade, socio-cultural and financialeconomic criteria and concluded that the USA was the best market for Colombian frozen beef exports. Yılmaz et al. (2017) focused on domestic target market selection for the Turkish furniture sector. They used the Analytical Hierarchy Process (AHP) method to determine the weights of sales, import-export volume, number of marriages, population growth rate and income and then used VIKOR (Vise Kriterijumska Optimizacija I Kompromisno Resenje) method to assess Ankara, Bursa, Istanbul, Eskisehir, Izmir and Kayseri to select target markets based on the weights. They concluded that the most ideal target market was Istanbul, followed by Ankara, Izmir, Bursa, Kayseri and Eskisehir. Tosun (2017) conducted a study on the decision makers of eight Turkish companies exporting fresh fruit and vegetables and used the fuzzy VIKOR method for export market selection. The market selection criteria were market competition, purchasing volume of the country, growth potential of the country's consumption, logistic opportunities and profit potential. They found that the United Arab Emirates was a better market than others. Schu and Morschett (2017) investigated the foreign market selection behavior of 140 online retailers in Europe and concluded that market size, the rule of law, information on local markets, the use of common language and logistical opportunities increased the likelihood of a country being selected as a target market. Yavuz (2016) used PROMETHEE (Preference Ranking Organization Method for Enrichment Evaluations) and Entropy methods to select a target market for Hatay furniture sector. The selection criteria were population size, number of marriages, population growth, population density, home sales rate, distance between, the number of furniture manufacturers and employees and income per capita. Koohi and Alikhani (2014) developed a questionnaire to determine factors affecting the selection of a target petrochemical export market. The factors were policies, culture, regulations, economy, infrastructure, market potential and buyers’ potentials and positions. Zhao et al. (2011) used the AHP method to determine factors affecting China's textile and apparel exports. The factors were market purchasing power, capacity of the market, the degree of
market monopoly, political stability and market requirements for product quality level. They concluded that China's target markets for textile and apparel exports were EU, USA, Japan and Hong Kong.
3. Material And Method
The aim of this study is to select the best export market for Turkish socks companies. We believe that this study will help those companies obtain maximum benefit-minimum cost advantages, have higher export shares and climb up the list of the top socks exporters. The steps in the flow chart in Figure 1 were followed for export market selection.
Figure 1. Flowchart for solution of socks export market selection problem Determining socks export market selection criteria: First, a survey containing several criteria was developed based on literature review to determine the criteria with the greatest effect on the selection of export markets. The study population consisted of export specialists of 196 socks exporters, which are members of Istanbul Apparel Exporters' Association (IAEA). Export specialists were contacted by phone and informed about the study. 50 of them agreed to participate and were asked to complete the electronic questionnaire. Table 2 shows the job positions of the participants. Table 2. Participants’ Job Positions Frequency
General Manager Assistant General Manager Export Director Marketing Director Production Manager Chairman of the Board Board Member Total
All participants were senior executives and had sufficient information on socks exports and an average of 14.08 years of work experience. 798
Table 3 shows the export market selection criteria with a frequency greater than 50% determined by participants. The frequency values of the criteria were normalized and converted to importance weights (Table 3). Table 3. Socks export market selection criteria and their importance weights Criteria Tax Rate Import Expense Population Income Per Capita Exchange Rate Distance Export Revenue Inflation Rate
Importance Weights 0.154 0.145 0.141 0.129 0.135 0.106 0.103 0.087
According to Table 3, the two most important criteria for participants were the “tax rate” (96%) and “import expense” (90%). Determining alternative socks export markets and criterion data: The world's top 15 socks importing countries were determined using the Trade Map Database (Table 4). Table 4. Distribution of global socks imports by country (1000$) Importing Countries 2014 USA 1.158.977 Japan 565.735 Germany 268.637 France 175.014 United Kingdom 182.638 Canada 135.998 Netherlands 86.920 Vietnam 2.569 Spain 75.438 Italy 96.557 Belgium 88.868 Russia 95.053 Australia 48.414 Mexican 60.899 Poland 31.646 Top 15 Countries 3.073.363 Total15 Countries (%) 78.51% Top Global Total 3.914.642 Source: www.trademap.org
In 2018, global socks imports were over $ 4.7 billion, and the top 15 exporters accounted for about 75 percent of total imports. Of those countries, the USA accounted for 32.81 percent of total imports, followed by Japan, Germany and France. Vietnam was not taken into account as an important importing country, because what placed it in the eighth position was the excessive
increase in its imports only in 2018 in the last five years. Moreover, this study took into account the export markets, which account for about 70 percent of the world's import of socks (Table 5). Table 5. Selection criteria data for socks export markets Markets
Income Export Revenue Population Inflation Tax Exchange Distance Per Capita ($) (x1000) Rate (%) Rate (%) Rate (Hour) ($)
According to Table 5, import expense, export revenue and tax rate values were derived from the International Trade Center. Exchange rate values were derived from the Central Bank of the Republic of Turkey. 2020 data should be calculated to select export markets, because it seems to be the earliest for Turkish socks exporters to enter new markets. Therefore, regression analysis was performed on 2014-2018 data (Table 5). 2020 values were estimated for each criterion (Table 6). Japan had an inflation rate of –0.1 percent in 2016, which was, therefore, not included in 2020 inflation forecast. Table 6. values estimated for 2020 Markets
USA 1647053.8 Japan 483742.2 Germany 209662.2 France 169314.6 United Kingdom 152553.4 Canada 180126.6 Netherlands 152049.8 Spain 54075.8 Italy 82162.8 Belgium 58884.2 Russia 47283.8
Determining methods to select export markets: Markets can be selected using numerous methods, some of which are MCDM methods that take both qualitative and quantitative criteria into consideration. The MCDM methods used in this study and their advantages are briefly described below.
3.1. TOPSIS Method
TOPSIS developed by Hwang and Yoon (1981) is one of the most widely used MCDM methods (Chen, 2019; de Farias Aires and Ferreira, 2019; Shukla et al., 2017; Zyoud and FuchsHanusch, 2017). TOPSIS determines the optimum alternative by calculating the distance between the positive ideal solution (PIS) and negative ideal solution (NIS) of each alternative and by ranking the alternatives for decision making problems (Baranitharan et al., 2019; Chen, 2019; de Farias Aires and Ferreira, 2019; Shukla et al. , 2017; Akgün and Erdal, 2019; Mateusz, 2018; Yayla et al., 2012). In TOPSIS, the best alternative is the closest to PIS and the farthest from NIS. PIS is a hypothetical alternative that maximizes benefit criteria and minimizes cost criteria whereas NIS minimizes benefit criteria and maximizes cost criteria (de Farias Aires and Ferreira, 2019; Shukla et al., 2017; Zyoud and Fuchs-Hanusch, 2017). Due to its comprehensibility, computational efficiency and ability to integrate other methods, TOPSIS is an effective method used in risk management, e-commerce, supplier selection, renewable energy, water resources management, climate change and sustainability assessment, risk management, logistics, energy management, design, engineering, production systems, health and safety management (Chen, 2019; Akgün and Erdal, 2019; Zyoud and Fuchs-Hanusch, 2017), location selection for solar farms, and selection of process parameters in computer networks and manufacturing industries (Baranitharan et al., 2019). The steps of TOPSIS are as follows: Step 1: Construct the decision matrix (A): Alternatives and criteria are listed on the rows and columns, respectively.
where aij is the real value of the alternative i according to the criteria j. Step 2: Construct the normalized decision matrix using Equation (1). m
Step 3: Calculate the weighted normalized decision matrix. The weighted normalized value 𝑣𝑖𝑗 is calculated as follows: 𝑣𝑖𝑗 = 𝑟𝑖𝑗 . 𝑤𝑗
where 𝑤𝑗 is the weight of the 𝑗𝑡ℎ criterion or attribute and
Step 4: Determine the positive ideal 𝐴∗ and negative ideal 𝐴− solutions.
Where B and C are the benefit and cost criteria, respectively. Step 5: Calculate the separation measures using the m-dimensional Euclidean distance. The separation measures of each alternative from the positive ideal solution (S*i ) and the negative ideal -
Step 6: Calculate the relative closeness to the ideal solution.
Step 7: Sort alternatives by the relative proximity ( C i ) to the ideal solution.
3.2. Grey Relational Analysis Method
The grey system theory was developed by Julong Deng in 1982 (Li et al., 2019; Yaser and Shunmugesh, 2019; Baranitharan et al., 2019). It can model and solve problems that stochastic or fuzzy decision making methods cannot, and therefore, is widely used for cross-system analysis, modeling, and forecasting and decision-making problems. Not all factors of real life problems can be characterized as full positive or full negative. Similarly, not all factors can be characterized as completely specific or completely uncertain (Kursun et al., 2016). A system with no uncertainty is defined as white, a system whose all factors are completely uncertain is defined as black and partially definite and partially indeterminate systems are defined as grey (Li et al., 2019; Yaser and Shunmugesh, 2019; Baranitharan et al., 2019). The grey relational analysis (GRA), which is a sub-part of grey system, is a decision-making method (Uzun, 2019). GRA is used in problems with multivariate statistical and inadequate data that do not fit into any distribution and cannot be modeled due to uncertainty (Uzun, 2019; Yaser and Shunmugesh, 2019, Anand et al., 2019). Due to its simplicity and evaluation capacity, GRA is used in various engineering applications (Baranitharan et al., 2019). GRA is a multi-response optimization method used to transform single-objective problems into single-response optimization problems (Baranitharan et al., 2019; Yazdani et al., 2019; Anand et al., 2019). It can even compute the basic relationship by using small and weak information. It examines the geometric proximity of different curves to evaluate the relationship between them (Li et al., 2019). GRA first converts the performance of all alternatives into a comparable sequence and then defines a reference sequence (ideal target sequence). It then calculates the grey relational coefficient between all comparability sequences and the reference sequence. Lastly, it calculates the grey relational grade between the reference sequence and every comparability sequences based on grey relational coefficients (Kuo et al., 2008). The steps are summarized as follows (Uzun, 2019):
Step 1: Prepare the dataset and construct the decision matrix Step 2: Construct the reference sequence and comparison matrix
Step 3: Normalize the data and construct the normalized matrix If larger sequence values contribute positively, then normalization for “the larger the better” attributes is as follows:
where xi (k ) is the original value in the row k. in the sequence i. xi (k ) is the value in the 0
row k. in the sequence i after normalization, min xi (k ) is the minimum value in the sequence 0
i and max xi (k ) is the maximum value in the sequence i. If larger sequence values contribute positively, then normalization for “the larger the better” attributes is as follows:
where x is the nominal value. Step 4: Create the absolute value table and grey relational coefficient matrix
Let k be the row k. on a n-length sequence and x0 (k ), xi (k ) be the grey relational
coefficient at the point k. calculated using Equations (12), (13), (14) and (15).
where (0,1) is a coefficient between 0 and 1. The function sets the difference between 0 i and max .
Step 5: The grey relational degree is calculated using Equation (16). n
x0 , xi is a measure of the geometric similarity between the sequence xi and the reference sequence x 0 in a grey system. If the weights of criteria are given in advance, grey correlation coefficients are calculated by the multiplying grey relationship coefficients and weights of the criteria.
3.3. ANP Method
AHP and ANP are the most commonly used MCDM methods (Nedjati and Izbirak, 2013). Saaty (1980) developed AHP method to solve complex decision-making problems, especially multi-criteria decision-making problems. AHP is used in many areas due to its ability to form hierarchical structures of decision-making problems easily and its computational efficiency. It has, however, some disadvantages such as not taking into account dependencies between criteria (Rad et al., 2018; Chou, 2018; Liang et al., 2013) because most decision-making problems cannot be hierarchically structured. It sometimes involves the interaction of various factors with highlevel factors, depending on low-level factors. Structuring a problem with functional dependencies allows for feedback between clusters and is regarded as a network system (Liang et al., 2013). Saaty, therefore, developed ANP, which is a revised decision-making tool used to overcome the disadvantage of the traditional AHP (1996). The main difference between AHP and ANP is that the latter takes into account the dependence and feedback among criteria, and thus, approaches the problem area from a more holistic perspective (Mahmoudkelaye et al., 2018; Rad et al., 2018; Nedjati and Izbirak, 2013; Chou, 2018; Liang et al., 2013; Kilic et al., 2015). AHP represents a framework with a one-way hierarchy relationship while ANP establishes complex relationships among decision levels and qualifications (Liang et al., 2013). The hierarchical structure of AHP, which falls within one-objective criteria, sub-criteria and alternatives, becomes a network in ANP. All clustered elements of the network can be associated with feedback and dependency relationships among all clustered elements of the network (Mahmoudkelaye et al., 2018). In general, ANP involves four main steps: (1) Model construction, (2) making pairwise comparisons and constructing a priority vector, (3) supermatrix formation and transformation, and (4) making the final ranking/determining priorities (Yıldız, 2014; Mahmoudkelaye et al., 2018; Kilic et al., 2015; Yildiz and Ergul, 2015; Chou, 2018). ANP is used to evaluate R&D projects, the performance of spa hotels, green suppliers and green supplier development programs, to select maintenance performance indicators, machine tools, product development and logistics service providers, to determine basic technologies, to measure sectoral competition level and performance and to perform SWOT analysis for the airline industry (Mahmoudkelaye et al., 2018; Kilic et al., 2015; Chou, 2018).
4.1. TOPSIS Results
In this section, it is aimed to select the best market place by using the above steps of TOPSIS method. Firstly, the estimated criterion data of the alternative export markets in Table 6 for the 805
year 2020 are used as the decision matrix. Then, the data in the decision matrix were normalized using the Equation (1) (Table 7). Table 7. Normalized Matrix Markets
The values in the matrix normalized using Equation (2) were multiplied by the criterion weights (Table 3) obtained using the survey results in order to obtain a weighted normalized matrix given in Table 8. Table 8. Weighted Normalized Matrix Markets
USA Japan Germany France United Kingdom Canada Netherlands Spain Italy Belgium Russia
Import expense, population, income per capita and exchange rate were the benefit criteria while export revenue, inflation rate, tax rate and distance were the cost criteria. The Equation (3) was used to obtain PIS (𝐴∗ ). The Equation (4) was used to obtain NIS (𝐴− ). A* ={0.135, 0.001, 0.114, 0.060, 0.011, 0.000, 0.045, 0.002} A- ={0.004, 0.064, 0.004, 0.008, 0.036, 0.105, 0.001, 0.082} In the next step, the separation measures of each alternative from the positive ideal solution (S*i ) and the negative ideal solution (Si− ) were calculated using Equations (5) and (6). The relative closeness of the alternatives to the ideal solution was calculated using Equation (7). The alternatives were ranked according to the closeness values shown in Table 9. 806
USA Japan Germany France United Kingdom Canada Netherlands Spain Italy Belgium Russia
Of the export market alternatives, the USA, France and the United Kingdom ranked 1 st (0.633), 2nd (0.495) and 3rd (0.494), respectively while Canada ranked last (0.304).
4.2. GRA Results
In GRA, Table 6 was the decision matrix, and the data were normalized using Equations (9) and (10). The import expense, population, income per capita and exchange rate were the benefit criteria and normalized for “the larger the better” attribute while the export revenue, inflation rate, tax rate and distance were the cost criteria and normalized for “the smaller the better” attribute. Table 10 shows the normalized matrix. Table 10. Normalized Matrix
Reference sequence Markets USA Japan Germany France United Kingdom Canada Netherlands Spain Italy Belgium Russia
In the next step, absolute values were calculated using Equation (13). The grey relational coefficient matrix (Table 11) was obtained using Equations (14), (15) and (12).
USA Japan Germany France United Kingdom Canada Netherlands Spain Italy Belgium Russia Weights
Lastly, Equation (17) was used to multiply the criterion weights (Table 4) by the grey relational coefficients using Equation (17) to calculate the weighted grey relational degrees (Table 12). The alternative export markets were ranked according to the total grey relational degrees. Table 12. Weighted Grey Relational Degrees and Ranking Markets
Import Export Population Income Inflation Tax Expense Revenue Per Rate Rate Capita
Of the export market alternatives, the USA, France and the Netherlands ranked 1 st (0.0891), 2nd (0.0847) and 3rd (0.0841), respectively whereas Canada ranked last (0.0621).
4.3. ANP Results
First, Super Decisions was used to generate a network structure in given Figure 2 based on the export market selection criteria and alternative markets.
Figure 2. Export Market Selection Problem Network Structure Then, the relationships among the criteria were considered, and the criteria affecting each other (Table 13) were determined by two export experts (Textile engineer and industrial engineer). Table 13. Criteria Interactions Criteria Population Export revenue Import expense Inflation rate Tax rate Distance
Criteria that affect import expense, export revenue, income per capita import expense, income per capita, inflation rate export revenue, income per capita, inflation rate, exchange rate import expense, export revenue, exchange rate import expense, inflation rate import expense, export revenue
The criterion weights in Table 3 and the criterion values of the alternatives were entered in the Super decision program and the ranking in Table 14 was obtained from the result screen of the program.
Of the export market alternatives, the USA, the Netherlands and Germany ranked 1 st (0.1229), 2 (0.1106) and 3rd (0.1065), respectively, whereas Russia ranked last (0.0109). nd
4.4. Sensitivity Analysis
The TOPSIS, GRA and ANP rankings were compared in Table 15. The USA ranked 1 st in all of them. France ranked 2nd in TOPSIS and GRA and 5th in ANP. Table 15. Alternative Rankings Markets USA Japan Germany France United Kingdom Canada Netherlands Spain Italy Belgium Russia
Topsis
Sensitivity analysis was performed to determine and evaluate the effect of the criterion weights on the rankings in Table 15. In the sensitivity analysis, the first criterion was assigned the lowest and highest criteria weight while the weights of the others were kept constant, and the criteria were ranked again using 16 scenarios. In this way, all three methods were used for analysis. Figure 3 shows the results.
TOPSıS 11 10 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
USA Japan Germany France United Kingdom Canada Netherlands Spain Italy Belgium Russia
Scenarios
GRA 11 10 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
USA Japan Germany France United Kingdom Canada Netherlands Spain Italy Belgium Russia
Scenarios
ANP 11 10 9 8 7 6 5 4 3 2 1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
USA Japan Germany France United Kingdom Canada Netherlands Spain Italy Belgium Russia
Scenarios
Figure 3. Sensitivity Analysis The USA ranked 1st in all scenarios according to the sensitivity analysis results for TOPSIS method (Figure 3). France and England ranked 2nd and 3rd, respectively, except in scenario 10 where “export revenue” was assigned the highest criterion weight. Canada ranked last in all scenarios. According to the sensitivity analysis results for GRA, the USA ranked 1 st, except in scenario 1 where “import expense” was assigned the lowest criterion weight. France ranked 2 nd, expect in scenarios 1 and 10. The Netherlands ranked 3 rd, except in scenario 4 where “income per capita” was assigned the lowest criterion weight and in scenario 10. According to the sensitivity analysis results for ANP, The USA ranked 1st, expect in scenarios 7 and 8 where “exchange rate”
and “distance” were assigned the lowest criterion weight. The Netherlands ranked 2nd, expect in scenarios 7 and 8 while Germany ranked 3rd, expect in scenario 7. Russia ranked last. These results are the same as those of the three methods, which means that even though criterion weights change, rankings remain the same, indicating that the results of the methods are sensitive. According to the sensitivity analysis results, the USA ranked 1 st in all three methods and is therefore the best export market for Turkish socks enterprises. France ranked 2 nd in TOPSIS and GRA, which do not take into account interactions among criteria. However, it was the Netherlands that ranked 2nd in ANP, which takes into account interactions among criteria. In other words, France and the Netherlands are alternative markets to the USA.
5. Conclusion And Suggestions
Export provides foreign exchange inflow and job opportunities, and therefore, is an important field of economic activity promoted by every country. Turkey has a 500 billion export target and a relatively high unemployment rate. Turkey should increase the volume of value added exports to achieve the export target and reduce the unemployment rate. Cotton is a very important raw material of Turkey. Turkey converts cotton into a value-added product by first converting it into yarn, then into fabric and finally into ready-made garments. 100 percent cotton denim jeans and socks are strategic products for Turkey. If Turkey achieves to select the right export market for those products, it can create short-, medium- and long-term plans and increase productivity and profit by using time, money and other resources efficiently. It can also increase its knowledge of the target market and develop effective marketing strategies. However, Turkey has to choose from a large number of markets with different capacities and potentials. However, this time-consuming problem requires multicriteria decision-making approaches that involve uncertainties and complexities due to conflicting objectives and various factors. This study addressed the problem of export market selection for the Turkish socks sector. First, criteria for the selection of export markets were determined. A survey was developed using those criteria. The survey was completed by export specialists of 50 socks exporting companies, which are members of IAEA. According to frequency analysis, the most important criteria for the socks export market were import expense, export revenue, population, income per capita, inflation rate, tax rate, exchange rate and distance. Alternative export markets were determined and analyzed using TOPSIS, GRA and ANP. In the sensitivity analysis of all three methods, the USA ranked 1st as the best stocking export market. According to the data of www.trademap.org, the USA is Turkey's seventh largest stocking export market. It ranked 1 st in all three methods due to high import expenditures depending on population and high income per capita. Turkey's socks export volume to the USA is small due to high taxes and distance. However, our results show that Turkey can increase its export volumes if it selects the USA as the target market in the coming years. The high potential purchasing power of American consumers will compensate for the reduced profitability arising from logistic costs and tax rates due to distance. Although France ranked 2nd in TOPSIS and GRA, which do not take into account interactions among criteria, it ranked 5th in ANP, which takes into account interactions among criteria. Proximity, lack of taxation, relatively low inflation and high population were the reasons why France ranked 2 nd in TOPSIS and GRA. The Netherlands ranked 6th in TOPSIS, 3rd in GRA and 2nd in ANP. ANP results show that the Netherlands and France are the best markets after the USA. Despite its small population, the Netherlands has high socks import expenditures and a high GDP and does not impose tax on Turkey. It is, therefore, a potential and attractive alternative market. Russia and Canada ranked last as alternative markets due to high tax rates, low exchange rate and distance (Canada). This study is a guide for Turkey in particular and for all countries that would like to increase their export volume in general. Different aspects of TOPSIS, GRA and ANP can be considered to
select different markets. The selection criteria are generally quantitative criteria. However, further research should take into account qualitative criteria and use methods that consider fuzzy sets.
Data Availability
All data, models, and code generated or used during the study appear in the submitted article.
Share and Cite
YILDIZ, A.; ÖZBEK, A. Selection of socks export markets for Turkey using multi-criteria decision making methods. Sigma Journal of Engineering and Natural Sciences 2020, Vol. 38, pp. 795-815. https://doi.org/10.62051/ytu.sigma-journal-of-engineering-and-natural-sciences-selection-of-socks-export-markets-for-turkey-using-multi-criteria-decision-makin