ASEAN Marketing Journal ASEAN Marketing Journal
Volume 10 Number 1 Article 5
November 2021
Implementing Data Clustering to Identify Capital Allocation for Implementing Data Clustering to Identify Capital Allocation for Small and Medium Sized Enterprises (SMEs)
Small and Medium Sized Enterprises (SMEs)
Amelia Hidayah
Department of Management, Faculty of Economics and Business, Trilogi University, South Jakarta, Indonesia, [email protected]
Follow this and additional works at: https://scholarhub.ui.ac.id/amj Part of the Marketing Commons
Recommended Citation Recommended Citation
Hidayah, Amelia (2021) "Implementing Data Clustering to Identify Capital Allocation for Small and Medium Sized Enterprises (SMEs)," ASEAN Marketing Journal: Vol. 10 : No. 1 , Article 5.
DOI: 10.21002/amj.v10i1.10627
Available at: https://scholarhub.ui.ac.id/amj/vol10/iss1/5
This Research Article is brought to you for free and open access by UI Scholars Hub. It has been accepted for inclusion in ASEAN Marketing Journal by an authorized editor of UI Scholars Hub.
66
ABSTRACT
Abstract: N/A.
Manuscript type: Original Research.
Research Aims: Analyzing the investigated factors includes the firm size and the industry sector that influence capital allocation for MSMEs using the K-means clustering technique.
Design/methodology/approach: The initial step for the research is the data preparation. It is covering 20 sectors and consists of 6,666 pieces of data. The modified data are analysed using the K-means clustering technique.
Research Findings: MSMEs are divided into three clusters, with each cluster exhibiting different characteristics in terms of assets, sales, and industry sectors. However, the number of employees was not found to be significant in the analysis.
Theoretical Contribution/Originality: The size of MSMEs is defined as the total assets, sales, or number of employees. It should be considered by financial institutions when assessing the viability of awarding a loan to that MSMEs. Moreover industry characteristics affect finance institutions to grant a loan to firm.
Practitioner/Policy Implication: Financial decision makers, banks, financial institutions, and government advisors alike can determine capital allocation for MSMEs based on the clustering profile created in this study.
Research limitation/Implications: This study constructs MSMEs grouping model that relies on investigated factors, including the firm size and the industry sector in order to determine the capital allocation to MSMEs. The future research regarding the capital allocation schemes for the MSMEs clustering profile are possible to do by adding new data, such as ownership type, owner-manager characteristics, and macroeconomic factors.
Keywords: capital allocation for MSMEs, K-means clustering, MSMEs characteristics, MSMEs financing, industry sector.
IMPLEMENTING DATA CLUSTERING TO DETERMINE THE CAPITAL ALLOCATION FOR MICRO, SMALL, AND MEDIUM-SIZED
ENTERPRISES (MSMEs)
Amelia Hidayah
Department of Management, Faculty of Economics and Business, Trilogi University
South Jakarta, Indonesia [email protected]
INTRODUCTION
The Indonesian government supports the development of MSMEs, since they have become the backbone of the Indonesian economy. Such enterprises account for 99%
of all enterprises operating in Indonesia, and they have created a total of 107.6 million jobs
in the country. Moreover, Indonesia’s MSMEs are responsible for generating more than 60% of the country’s gross domestic product (GDP) (Indonesia Investments, 2016). Yet, the Indonesian government is still trying to boost the number of entrepreneurs in the country, which lags behind other Asian countries, such as Singapore, China, and Japan, in that regard.
67
Amelia Hidayah / ASEAN Marketing Journal © June (2018) Vol. X No. 1
The Indonesian government helps MSMEs by promoting their products through the creation of a digital platform equipped with an online marketing and e-payments system called E-Smart MSMEs (News Desk, The Jakarta Post, 2017). The government further aims to assist MSMEs with product certification, the purchasing of machinery and equipment as a form of industrial restructuring, and the promotion of local and international exhibitions (News Desk, The Jakarta Post, 2017).
MSMEs contribute to economic development in a number of key ways, including promoting industrialization and district development, job creation, more equitable income distribution, incremental resource usage, additional sources of foreign exchange income, collaboration with mature industries, and access to entrepreneurship training (Statistics Indonesia, 2016). A large number of MSMEs influence to Indonesia economic growth. According to data from the Ministry of Co-operatives and MSMEs, there are 59.3 million MSMEs in 2014: 98.75% are micro businesses, 1.15% are small businesses and 0.1% are medium-sized businesses (as cited in OECD, 2018, p. 21).
The performance of MSMEs has impressed the government due to the significant contribution they make to the country’s GDP.
Such enterprises are keen to more securely position themselves as being vital to the national economy, although they do still face a number of difficulties. The process involved in developing MSMEs is not straightforward in the context of business unit growth. Indeed, micro businesses are known to find it difficult to transition into small businesses, while small businesses are known to experience difficulty when seeking to develop into medium businesses, and so on. The difficulty stems from a fundamental misunderstanding of business profiling, since the characteristics of a micro business, for example, are different from the characteristics of a small business.
The notion of developing an MSME to the next level involves consideration of the growth factors that influence MSMEs. Hill
(2001) suggests that Indonesian MSMEs face two key problems, namely financial problems and marketing problems (as cited in Direktorat Pengembangan UKM Dan Koperasi, 2016, p.
10). The Indonesian government implements a marketing program that is intended to stimulate the market, although the program cease in 1998. In fact, the program is not actually proved effective, since there is a general assumption that MSMEs cannot compete in the market. As all MSMEs need to grow in order to be successful, their financial position has been emphasized to be an important factor during the development phase (Cook, 2001; Ou & Haynes, 2006).
Financing methods have changed in line with changes in the general business climate.
MSMEs have access to two potential sources of finance, namely internal and external financing sources. Internal financing sources include the owner-manager’s personal savings and retained profits (Wu, Song,
& Zeng, 2008), while external financing sources include financial assistance provided by family and friends (Abouzeedan, 2003).
He and Baker (2007) describe trade credit, venture capital, and angel financiers as being external financing sources, while Chittenden, Hall, and Hutchinson (1996) consider formal external sources to be represented by financial intermediaries, such as banks, financial institutions, and securities markets.
The availability of financing methods is dependent on the growth cycle of MSMEs.
During the early phase, Berger and Udell (1998) explain that the characteristics of MSMEs include informational opacity, a lack of trading history (Cassar, 2004), and a high risk of failure (Huyghebaert & Van de Gucht, 2007). They hence have to rely on internal funding sources during this phase. As MSMEs begin to advance through their business lifecycle, they change their capital structure (La Rocca, La Rocca, & Cariola, 2011). When they mature as a business, MSMEs boost their trade through commercialization and the ability to provide collateral. The objective is to improve the creditworthiness of the firm and to attract investors to inject money into
the business. This involves an automatic shift from relying on internal financing sources to relying on external financing sources, including venture capitalists, trade credit, and bank loans. Lastly, when MSMEs become more advanced in terms of the transparency of their business activities, they prefer to rely on either equity, debt, or both (Abdulsaleh &
Worthington, 2013).
The likelihood of facing global competition is a factor that should be considered by all MSMEs, especially with regard to their readiness to do so, since many international investors seek to invest in Indonesia. Although the Indonesian government has established various programs intended to strengthen the position of MSMEs, many have failed to take advantage of such programs. Based on data concerning the growth of wholesale private banking in Indonesia, of all the MSMEs in the country, only 36% have bank accounts (Irjayanti, Azis, & Sari, 2016). This highlights the lack of access on the part of Indonesian MSMEs to many financing schemes (Irjayanti, Azis, & Sari, 2016).
Gregory, Rutherford, Oswald, and Gardine (2005) suggest that MSMEs financing cannot be standardized. A study is conducted using a sample of 292 Australian firms find that the
“larger” small firms are, the more they rely on long-term debt and external financing, including bank loans (Cassar, 2004). The idea that the firm size influences the availability of finance to a firm has been tested by Petersen and Rajan (1994). They conclude that as firms grow, they develop the ability to enlarge the circle of banks from which they can borrow.
Moreover, they note that firms that deal with multiple banks and credit institutions tend to be nearly twice as large as those that deal with only one bank.
Previous research suggests that firms exhibit different characteristics. Therefore, the firm type will affect finance institutions’ decision to grant (or deny) a loan to a given firm (Hutchinson & Michaelas, 2000, as cited in Kurniawan, 2014, pp. 611-621).
Information is arguably the key driver of the modern world. Data in an unstructured form must be converted into a structured form, namely information, if it is to prove useful. Computers manipulate a massive amount of data on a daily basis to extract certain necessary information. Hence, the concept of “Big Data” has been introduced to businesses as a powerful form of technology.
The term “data clustering” refers to the process of partitioning a set of data into a set of meaningful sub-classes known as clusters.
Data clustering has numerous applications in every field of modern life. For instance, it serves to identify groups of related records that can be used as a starting point for the exploration of further relationships (Harikumar, Rajgopal, & Soman, 2010). This study uses the K-means clustering technique to analyze the factors that influence capital allocation. The investigated factors include the firm characteristics, the firm size, and the industry type.
LITERATURE REVIEW
Defining MSMEs
The Indonesian government has sought to support MSMEs by establishing Law No.
20 of 2008 on Micro, Small and Medium, Business. This defines MSMEs as follows:
1. A micro business is defined as a business owned by individuals and/or individual business entities that meets the criteria of a micro business as stipulated in the Act.
2. A small business is a standalone business operated by an individual or a business entity that is not a subsidiary or company branch that is owned, controlled, or part of a medium or large business, either directly or indirectly, and meets the small business criteria as stipulated in the Act.3. A medium business is a standalone business carried out by individuals or a business entity that is not a subsidiary or company branch that is owned, controlled, or part of a small or large business, either
69
Table 1. The Definition of MSMEs
Business Category Net Assets (IDR) Annual Revenue from Sales (IDR)
Micro <= 50,000,000 <= 300,000,000
Small 50,000,000–500,000,000 300,000,000–2,500,000,000
Medium 500,000,000–10,000,000,000 2,500,000,000–50,000,000,000
Amelia Hidayah / ASEAN Marketing Journal © June (2018) Vol. X No. 1
directly and indirectly, with the maximum assets or annual sales proceeds as stipulated in the Act.
The government also segregates the assets and sales of MSMEs as described in Law No. 20 of 2008, which is summarized in Table 1.
MSME Size
The size of MSMEs is defined as the total assets, sales, or number of employees.
Abdulsaleh (2013) argues that the idea of the firm size affecting the activities and growth potential of MSMEs is universally agreed upon by researchers. As mentioned above, Gregory, Rutherford, Oswald, and Gardine (2005) suggest that MSME financing cannot be standardized, while the financial needs and options of MSMEs are determined to a certain degree by size/
age, since only the firm size was found to be a significant predictor (in some cases) of the capital structure of MSMEs (as cited in Abdulsaleh, 2013, p. 37). The size of an MSME should be considered by financial institutions when assessing the viability of awarding a loan to that MSME. The company size greatly influences the structure of funding based on the fact that larger companies have a tendency to apply for higher loan amounts.
Industry Sector
A number of studies have found that the industry sector influences the capital allocation of MSMEs. Different industry
sectors have different characteristics, for example, the agriculture industry faces different business risks when compared to the banking industry, which in turn indirectly influence the capital structure of a business, including the asset composition. Therefore, the firm type will affect finance institutions’ decision to grant (or deny) a loan to a given firm (Hutchinson & Michaelas, 2000, as cited in Kurniawan, 2014, pp. 611-621).
Abor (2007) indicates that MSMEs in the agricultural sector exhibit the highest capital structure and the highest asset structure, while MSMEs in the wholesale and retail trade sector exhibit the lowest debt ratio and the lowest asset structure.
K-Means Clustering Algorithm
According to this algorithm, the center of the cluster (or centroid) is randomly selected at an early stage from a collection of data colonies (the population). Then, the K-means are used to verify each component within the data population and to allocate the components to one of the centroids previously defined on the basis of the minimum distance between the components (data) within each centroid.
The position of the centroid will be recalculated until all the data components are grouped into every centroid, and the final step will determine a new centroid position. This iteration will be performed until a convergent condition is formed. The K-Means clustering algorithm is depicted
Start
Define number of clusters k
Define centroid
Calculatet he distance from each data/object to the center of a cluster
Group each object basedo nt he minimum distance
Is there any object to be
moved?
End
Yes
No
Figure 1. K-Means Algorithm
Table 2. Descriptions of the Variable in the Raw Data
Figure 2. K-Means Setting
Company Name Business name
Product Description Commodities that are produced by companies Total Assets Sum of all current and noncurrent assets
Total Sales Total amount of sales annually
Employee Number Total number of employees working in companies Survey Year Defined when Bank Indonesia conducted the survey to
obtain MSME-related data
71
Table 3. Iteration History
Iteration Change in Cluster Centers
1 2 3
1 154974658.363 2347081.524 187627136.228
2 0.000 155533.545 144062705.657
3 0.000 265446.569 98178370.708
4 0.000 263860.366 68829065.244
5 0.000 507216.198 72018917.094
6 0.000 580847.891 42212703.387
7 0.000 267819.066 18569511.158
8 0.000 30853.389 2237984.364
9 0.000 18490.251 1455027.860
10 0.000 0.000 0.000
Amelia Hidayah / ASEAN Marketing Journal © June (2018) Vol. X No. 1
in more detail in Figure 1 (Nugraha &
Kusumawati, 2012, p. 4).
RESEARCH METHOD
Data Preparation
This step is designed to construct a MSMEs profiling model that can be used by financial decision makers to determine the capital allocation to MSMEs. The model relies on specific variables, including the total assets, sales, number of employees, and industry sector. The MSMEs profiling analysis is conducted using the K-means clustering technique. The secondary data utilized in the analysis (covering 20 sectors) were obtained from the MSME site of Bank Indonesia. The raw data consist of eight variables and 6,715 pieces of data concerning MSMEs, as described in Table 2 below.
K-Means Clustering Process
The raw data were modified, which involved checking for missing data (empty total assets/survey year/employee number) and removing such entries from the dataset. The total modified data account for 6,666 pieces of data. Then, the modified data were analyzed using the K-Means clustering technique. The
total assets, sales, number of employees, and industry sector were used to perform the MSMEs clustering. The analysis was supported by setting the number of clusters as three, the number of maximum iterations as ten, and mark the distance of each data point from the cluster center (Figure 2).
RESULTS AND DISCUSSION
The iteration history shows the progress of the clustering process at each stage. During the early iterations, the cluster centers have settled into the general areas of their locations, while the last eight iterations make only minor adjustments.
The final cluster centers are computed as the mean of each variable within each final cluster. They reflect the characteristics of the typical case for each cluster.
1. The MSMEs in cluster 1 have more assets and highest sales revenues when those in compared to clusters 2 and 3.
2. The MSMEs in cluster 2 have fewer assets and lower sales revenues when compared to those in clusters 1 and 3.
3. The MSMEs in cluster 3 have the highest assets of all the clusters as well as moderate sales revenues.
Table 4. Final Cluster Centers
Table 5. Distances Between the Final Cluster Centers
Table 6. Number of Cases in Each Cluster
Cluster
1 2 3
Industry Sector 15 12 6
Total Assets in Million
IDR 44,737,700 257,404 185,966,205
Sales Revenues in Million
IDR 520,000,331 85,921 37,167,668
Number of Employees 3 3 3
Cluster 1 2 3
1 N/A 521,813,654.510 503,063,486.148
2 521,813,654.510 N/A 189,374,799.135
3 503,063,486.148 189,374,799.135 N/A
Cluster 1 3.000
2 6580.000
3 83.000
Valid 6666.000
Missing 0.000
However, the number of employees does not have a significant impact on the clusters, since it remains the same for all three clusters.
Table 5 shows the distances between the final cluster centers. Greater distances between the clusters correspond to greater dissimilarities between them. Clusters 1 and 2 are the most different.
A large number of cases are assigned to the second cluster, which unfortunately involves the least assets and the lowest sales revenues.
Cluster 3 has a low number of cases, the lowest number of industry sectors, the highest total assets, and moderate sales revenues.
Cluster 1 has the lowest number of cases, the highest number of industry sectors, more total assets, and the highest sales revenues.
Financial decision makers, banks, financial institutions, and government advisors alike can determine capital allocation for MSMEs based on the clustering profile created in this study. Another important consideration during capital allocation concerns the industry sector, for example, agriculture, forestry, and fishery businesses could potentially support food self-sufficiency programs, and they require equipment and technology to optimize their operational activities.
Indonesian national fishery production stands at 12.29 million tons. Although fishpond cultivation has contributed 1.1 million tons to this total figure, demand has increased by 11%; therefore, the potential of the fishery business should not be underestimated (Kerjasama LPPI Dengan Bank Indonesia, 2015).
73
Amelia Hidayah / ASEAN Marketing Journal © June (2018) Vol. X No. 1
CONCLUSION
This paper demonstrates how MSMEs’
characteristics, total assets, sales, and industry sector influence financial decision makers’ approach to capital allocation.
Interestingly, the number of employees, as an aspect of MSMEs’ characteristics, is not a factor that influences the capital allocation.
In conclusion, MSMEs can be divided into three clusters. The firms in cluster 1 have more assets and the highest sales revenues when compared to those in clusters 2 and 3. The firms in cluster 2 have fewer assets
and lower sales revenues when compared to those in clusters 2 and 3. Finally, cluster 3 features a low number of cases, less industry sectors, the highest total assets, and moderate sales revenues. The future research to determine capital allocation schemes for the MSMEs clustering profile are possible to do by adding new data, such as ownership type, owner-manager characteristics, and macroeconomic factors, is both desirable and feasible. Such research should serve to enrich the information available concerning the determination of capital allocation for MSMEs.
REFERENCES
Abdulsaleh, A. A., & Worthington, W. A. (2013). Small and Medium-Sized Enterprises Financing: A Review of Literature. International Journal of Business and Management, 8, 36–54.
Abor, J. (2007). Industry Classification and the Capital Structure of Ghanaian SMEs. Studies in Economics and Finance, 24(3), 207–219.
Abouzeedan, A. (2003). Financing Swedish Small and Medium-Sized Enterprises (SMEs):
Methods, Problems and Impact. Paper presented at the 43rd European Congress of the Regional Science Association, Jyväskylä, Finland.
Berger, A. N., & Udell, G. F. (1998). The Economics of Small Business Finance: The Roles of Private Equity and Debt Markets in the Financial Growth Cycle. Journal of Banking and Finance, 22, 613–673.
Cassar, G. (2004). The Financing of Business Start-Ups. Journal of Business Venturing, 19(2), 261–283.
Chittenden, F., Hall, G. C., & Hutchinson, P. J. (1996). Small Firm Growth, Access to Capital Markets and Financial Structure: Review of Issues and an Empirical Investigation. Small Business Economics, 8(1), 59–67.
Cook, P. (2001). Finance and Small and Medium-Sized Enterprise in Developing Countries.
Journal of Developmental Entrepreneurship, 6(1), 17.
Direktorat Pengembangan UKM Dan Koperasi (2016). Strengthening MSMEs For Economic Growth. Warta KUMKM, 5(1), 4–35.
He, W., & Baker, H. K. (2007). Small Business Financing: Survey Evidence in West Texas. The Journal of Entrepreneurial Finance and Business Ventures, 12(1), 27–54.
Huyghebaert, N., & Van de Gucht, L. M. (2007). The Determinants of Financial Structure: New Insights from Business Start-Ups. European Financial Management, 13(1), 101–133.
Indonesia. (2008). Law Of The Republic of Indonesia Number 20 Year 2008 Regarding Micro, Small and Medium-Sized Enterprises. Retrieved from https://www.bi.go.id/id/tentang-bi/
uu-bi/Documents/UU20Tahun2008UMKM.pdf.
Harikumar, A.H., Rajgopal, C.R., & Soman, K.P., (2010). A Simple Approach to Clustering in Excel. International Journal of Computer Applications (0975 – 8887) Volume 11– No. 7, 19-25.
Indrawati, S. M. (2016). Indonesian Economy: Micro, Small & Medium Sized Enterprises.
Retrieved from https://www.indonesia-investments.com/id/news/todays-headlines/
indonesian-economy-micro-small-medium-sized-enterprises/item7068.htm.
Irjayanti, M. A., Azis, A. M., & Sari, P. A. (2016). Indonesian SMEs Readiness For Asean Economic Community. Actual Problems of Economics, 3, 31–38.
Kerjasama LPPI Dengan Bank Indonesia. (2015). Profil Bisnis Usaha Mikro, Kecil Dan Menengah (UMKM). Jakarta: Author.
Kurniawan, K. (2014). Factors Affecting Financial Access in Micro, Small and Medium-Sized Enterprises (MSMEs) Loans in Brebes Regency. Journal & Proceeding Fakultas Ekonomi
& Bisnis UNSOED, 4(1), 611–621.
La Rocca, M., La Rocca, T., & Cariola, A. (2011). Capital Structure Decisions During a Firm’s Life Cycle. Small Business Economics, 37(1), 107–130.
News Desk, The Jakarta Post. (2017, March 4). Industry Ministry aims to see 20,000 new SMEs set up by 2019. The Jakarta Post. Retrieved from https://www.thejakartapost.com/
news/2017/03/04/industry-ministry-aims-to-see-20000-new-smes-set-up-by-2019.html.
Nugraha, J. A. M., & Kusumawati, Y. (2012). Information Processing of Drug Supplies In The Pandanaran Health Center Semarang By Using Data Mining. Retrieved from http://eprints.
dinus.ac.id/13153/1/jurnal_13589.pdf.
OECD. (2018). SME and Entrepreneurship Policy in Indonesia 2018. Retrieved from https://
read.oecd-ilibrary.org/employment/sme-and-entrepreneurship-policy-in-indonesia- 2018_9789264306264-en#page23
Ou, C., & Haynes, G. W. (2006). Acquisition of Additional Equity Capital by Small Firms – Findings from the National Survey of Small Business Finances. Small Business Economics, 27(2), 157–168.
Petersen, M. A., & Rajan, R. G. (1994). The Benefits of Lending Relationships: Evidence from Small Business Data. Journal of Finance, 49(1), 3–37.
Statistics Indonesia. (2016). Perkembangan Indeks Produksi TriwulananIndustri Mikro Dan Kecil 2014-2016. Jakarta: Author.
Wu, J., Song, J., & Zeng, C. (2008). An Empirical Evidence of Small Business Financing in China. Management Research News, 31(12), 959–975.
