Determinants of Capital Structure in Small and Medium Enterprises in Malaysia: An Exploratory Factor Analysis

It is theoretically acknowledged that decisions on capital structure remain an important goal to maximise returns. This study seeks to identify factors influencing the capital structure of Malaysia’s small and medium enterprises (SMEs) through re -assessment of the operationalisation of constructs and an evaluation of certain assumptions on capital structure choices. An analysis of 500 questionnaires from SMEs located in the central region of Peninsular Malaysia was considered in the context of reliability and validity concerning capital structure determinants. The finalised model of capital structure consists of 11 constructs and 40 items. The results of an exploratory factor analysis (EFA) confirmed that external factors such as culture and environment also influenced the capital structure choices of SMEs. Prospective research could validate the framework proposed in this study and establish the framework as a standard measurement exercise to assess SMEs’ capital structure determinants. This study also presented policy implications for strategies to manage, preserve, and enhance economic systems, legal and tax frameworks, and inflation rates that may aid SMEs in accessing different financial sources.


Introduction
Capital structure decisions are usually equated with complexity in making choices that could maximise economic returns (Nawi, 2015). Firms generally finance their activities, either individually or collectively, using personal savings, internal funds, debt, and equity (Kumar et al., 2020). The prevalence of capital structure was firstly investigated by Modigliani and Miller (1958) prompting subsequent theories to be established, for instance, pecking order theory (Donaldson, 1961;Myers, 1984;Myers & Majluf, 1984), static trade-off theory (Jensen & Meckling, 1976;Myers, 1977), and agency theory (Jensen & Meckling, 1976;Myers, 1977). A substantial share of theories as a result of these research suggested several factors that might influence firms' capital structure decisions.
Based on the suggestions raised in previous studies, this study investigated the determinants of capital structure based on firm-specific, external, and owner-related factors such as factors associated with business culture and environment as demonstrated in Malaysian-based studies. The variables found in the Malaysian-based studies included, but were not limited to potential interdependencies between significant effects of industry on Malaysian corporate financing decisions (Mohamad, 1995;Annuar & Shamsher, 1993) and financing patterns among Malaysian SMEs (Nawi, 2015;Moha & Khadijah, 2011).
In addition to developing a model on capital structure determinants of SMEs, an empirical study testing the applicability of the model in a non-Western setting was also conducted. To generate a more nuanced picture of the model, a re-assessment of the operationalisation of constructs and an evaluation of certain assumptions on capital structure choices were carried out. Therefore, the following research question demonstrated the centrality of the study: What are the factors that influenced the capital structure of Malaysia's SMEs? In particular, the central question demonstrated above was further sequenced into three subinquiries: 1. Do owner-related factors such as 'perceptions and beliefs', 'relationship,' and 'networking' influence the capital structure of SMEs? 2. Do firm-related factors such as 'objectives and goals' and 'business planning' influence the capital structure of SMEs? 3. Do external factors such as 'environment' and 'culture' influence the capital structure of SMEs? The article focusing on capital structure decisions proceeded as follows. The next section reviewed literature related to the capital structure determinants and the development of associated hypotheses. Subsequently, the methods employed in this paper were elaborated in detail, followed by a description of results and discussions of the empirical approach. Conclusions and recommendations were provided as the outcome of the article. Therefore, generating and testing the model of capital structure decisions helped develop a clearer understanding of the particular determinants of SMEs and offer practical insights to public policymakers, particularly, financial and non-financial institutions that have since provided financial facilities to SMEs.

Literature Review and Development of Hypotheses
The theories and empirical studies emphasising capital structure generally focused on large listed firms. These theories resulting from prior research on capital structure, irrespective of the companies' nature, contexts, and distinct approaches, could be applied to SMEs. Among others, the centrality of SME capital structures was evidenced in (Abor and Biekpe, 2007;Borgia and Newman, 2012;Chittenden et al., 1996;Hamilton and Fox, 1998;Michaelas et al., 1999;Reid, 1996;Sogorb and Lopez, 2003;Watson and Wilson, 2002;Lopez-Gracia and Sanchez-Andujar, 2007;Bajaj et al., 2020).
However, three consistent threads binding these research were finance access constraints, complexities, and challenges facing SMEs, particularly start-ups, despite varying financing options that were made available (Carter et al., 2003;Hood, 2000;Robb & Fairlie, 2007). One of these complexities was related to the models of capital structure that distinguished SMEs from larger companies. Smaller firms' disclosure on asset structure, business planning, and profitability was found to be restricted because smaller firms generally had lower tangible assets, higher intangible assets, and greater financial risks (Cressy & Olofsson, 1997). However, larger firms generally were allowed to release detailed information to lenders because larger firms were found to be associated with lower financial risks and existing liabilities relative to total assets, and relatively higher fixed assets to total asset ratio (Fama & Jensen, 1983;Peel & Wilson, 1996;Rajan & Zingales, 1995). Thus, the threshold of optimum capital structure that SMEs were recommended to accomplish remained uncertain.
Studies and hypotheses concerning capital structure were drawn based on the following constructs:

Business Planning
Business plans describe an organisation's plans, activities, previous statuses, existing states of conditions, and prospective goals that the organisation plans to meet. Business planning records are usually represented through four items, namely, business plans, formal and strategic long-term plans, information on formal management structure, and performance appraisals (Zunckel & Nyide, 2019;Nawi, 2015).
A great importance is emphasised to business planning because a lack of business planning will invariably lead to information opacity, resulting in difficulties in accessing and securing external financial facilities (Nawi, 2018;Nguyen & Ramachandran, 2006). However, to minimise information opacity, regular and accurate financial reports, forecasts, and statements that portrayed overall business performance reflecting firms and SMEs' ability to repay loans all needed to be released (Coleman & Carsky, 1999). However, studies showed that disclosures of information largely depended on the quality of data and the degree of privacy and confidentiality between firms and outsiders (Romano et al., 2001). Smaller firms were generally found to be reluctant to disclose firms' information to outsiders (Berger & Udell, 1998). Nonetheless, it was found that well-established firms with relatively lower default rates and lower costs of debt provided transparent financial standing records (Harris & Raviv, 1991). Therefore, because decisions made by fund-granting institutions were frequently correlated with business planning, the following hypotheses were posited: H1.1: There was a negative association between business planning and funds from family and friends. H1.2: There was a positive association between business planning and debt.

Business
Objectives SME owner-managers played two important roles, namely, in realising commercial or lifestyle goals, as part of making firms' capital structure decisions (Mat Nawi, 2015;Romano et al., 2001). Firstly, owners were concerned about maintaining a sufficient level of income so that they and their families could enjoy their chosen lifestyles (Morrison, 2006). For instance, owner-managers realised goals that accumulated wealth (Ou & Haynes, 2006), supported families by earning sufficient income from businesses (Getz & Carlsen, 2000), and, for home accommodation managers, met personal satisfaction and enjoyment in addition to receiving extra money from home-stay guests (Lynch, 1999). However, past reviews of literature indicated that there was no evidence linking commercial and lifestyle goals to capital structure in developing countries, particularly Malaysia.
Secondly, SME owner-managers helped to realise the goals of growth (i.e. expansion) and external equity. On the one hand, some owners-managers remained bullish about their businesses by choosing equity instead of debt financing (Chaganti et al., 1995). On the other hand, small productive firms relied on internal funds to materialise the long-term value of their businesses.
Thirdly, reviews of literature also pointed to stark differences among firms who employed varying sets of control in dealing with debts. Some firms were found to have relied on their own sets of beliefs, independence, and control as opposed to firms that relied on investments to grow. As a result, firms' reluctance on debt-financing was found to be negatively associated with management shareholding (Friend and Lang, 1988). Nonetheless, studies showed that businesses that relied on debt financing were found to grow steadily (Romano et al., 2001). While far fewer studies examined how firms realised their objectives, the following hypotheses were formulated: H2.1a: SMEs which focused on lifestyle goals were unlikely reliant on debt or external equity.
H2.1b: SMEs which focused on lifestyle goals were likely reliant on internal capital and retained profits.
H2.2a: SMEs which focused on social welfare goals were unlikely reliant on debt or external equity.
H2.2b: SMEs which focused on social welfare goals were likely reliant on internal capital and retained profits.
H2.3a: SMEs which focused on commercial goals were likely reliant on debt or external equity.
H2.3b: SMEs which focused on commercial goals were unlikely reliant on internal capital and retained profits.

Owner's Motivation, Perceptions, and Beliefs
Three reasons surrounding owners' motivation, perception, and beliefs existed. Firstly, different managers defined firms' operationalisation differently. Despite perceptions playing an important role in how individuals and firms allocate resources, managers or firms' owners may describe their concerns differently (Norton, 1990), financing attitudes, cultural norms, managerial motivation, and self-interests (Friend & Lang, 1988). Secondly, perceptions and beliefs by owner-managers about external finance determined capital structure decisions of small firms (Michaeles et al., 1998). Thirdly, analysis using pecking order hypothesis showed that capital structures' formulation, particularly among small firms, depended upon firm management's belief systems (Norton, 1990). As such, small firms generally utilised the debtaverse or 'no-debt-at-all' concept. Through previous studies and qualitative findings, this study was designed to determine the relationship between the three concepts, namely, owners' motivation', 'perceptions and beliefs', and 'capital structure'. Therefore, a hypothesis was posited: H3: Owners' 'motivation' and perceptions and beliefs' were negatively associated with debt or external equity.

Relationships and Networking
In businesses, relationships and networking are generally understood as business and social relations (Nguyen & Ramachandran, 2006). Business and social relations that form extensive relationships and networking could involve financial stakeholders (Nguyen & Ramachandran, 2006) and non-financial stakeholders (Parsons & Titman, 2007). Among the goals of business relationships and networking were generally related to increasing agency and reducing information asymmetry crises (Cole, 1998;Petersen & Rajan, 1994).
Firstly, social and business networking bolster finance access through qualitative and quantitative information (Scott, 2006). Firstly, studies showed that fund-granting institutions generally approved of loans through local banks' evaluation of records of personal relationships with clients as opposed to relying upon borrowers' financial standing evaluation (Krishnan & Moyer, 1997). Building social relationships among creditors could guarantee availability and fast decisions on funding approvals, loans at the best possible rate, and personalised loan needs (Donnelly et al., 1985). Secondly, studies also demonstrated that small firms whose accounting records were generally poor relied on their rapport with banks as they attempted to secure bank financing (Abor & Biekpe, 2007). Therefore, social relations and networking were frequently correlated with minimising asymmetric information and borrowers' liquidity constraints (Petersen & Rajan, 1994).
Secondly, it has been suggested that building 'relationships' and 'networking' could also be equated with forms of transaction lending or relationship lending (Kusi et al., 2021). Firstly, transaction lending commonly describes transparent borrowers, while relationship lending signifies opaque borrowers (Brighi & Torluccio, 2007). Secondly, the provision of relationships in this context functions as 'soft information' to support credit approval through, for instance, well-established records of relationships with financial-granting institutions to alleviate opacity risks. Thus, it could very well be anticipated that links to and relationships with banks may offer explicit or implicit guarantees of access to funds, particularly, during unforeseen financial circumstances (Nguyen & Ramachandran, 2006). Therefore, in comparison to SMEs, large firms are typically seen to have 'upper hands' in transaction lending based on 'hard information'.
Thirdly, studies suggested that 'close relationships' were key predictors to eventual lending transactions and how fund-granting institutions respond to both large firms and SMEs. Firstly, 'lending discrimination' (Petersen & Rajan, 1994) that was enforced may result in an undue limitation of access to loans or a reduction in firms' capacity to increase the marketability of products and goods (Robb & Fairlie, 2007). Secondly, well-established relationships between lenders and borrowers may help to minimise information asymmetry crises through provisions of transparent and accurate information (Nguyen & Ramachandran, 2006). Therefore, close relationships with financial-granting institutions not only promote large firms and SMEs' track records but also help to attract external finance (Scott, 2006).
Fourthly, studies suggested that networking can generate companies' information on reliability (Nguyen & Ramachandran, 2006). Because financing could be granted from informal networks or through trade credit from suppliers (Newman et al., 2011), direct and indirect transactions with firms' counterparts or counterparts' networks such as family members can satisfy creditors and lenders concerning counterparts' reliability (McMillan & Woodruff, 1999). Thus, large firms and SMEs' associations with network added value to business communities, customers, and suppliers (Holmlund, 1997). In this way, shortfalls in receiving trade credits and short-term liabilities could be traced back to managers' conduct towards networking, including trust and confidence by banks on banks' relationships with lenders (Kusi et al., 2021;Berger & Udell, 1998;Han et al., 2009;Nguyen & Ramachandran, 2006;Mat Nawi, 2015).
Finally, generating numerous channels of fund sources also depended upon large firms and SMEs' social relations (Petersen & Rajan, 2002). In China, Japan, Korea, Southeast Asian countries, and many developing countries with poor and often inadequate laws of contracts, networking, and other similar informal relationships helped facilitate business transactions without restrictions (Greif, 1993;Yeung & Tung, 1996). Following theoretical and empirical studies presented, the following hypotheses were posited: H4: A good firm-lender/supplier relationship would enhance debt levels and firms' external equity.
H5: A wide networking would increase the level of debt financing of the firm.

Environment
The combination of previous research also led to a dramatic increase in research that examined the interdependencies between capital structure and inflation rate. Firstly, research determining effects of 'environment' on capital structure across developed countries generally concluded that countries' macro-economic data such as gross domestic products (GDP) growth and inflation rates had implications for debts accrued (Michaelas et al., 1999). Secondly, while studies carried out by Covin et al (1999); Michaelas et al (1999); Naman and Slevin (1993); Sener (1989); Taggart (1995) continued to generate interests in capital structure choices, some other studies found no relationship between inflation rates and capital structure decisions (see, for instance, Booth et al., 2001;Hatzinikolaou et al., 2002;Mutenheri & Green, 2002). Furthermore, studies demonstrated that irrespective of countries' economic statuses, developed or not, debt rations remained affected by macroeconomic conditions such as inflation rates and GDP growth rates in determining firms' capital structure (Booth et al., 2001). Therefore, the extent to which environment-related factors affected capital structure decisions remained unknown (Gaud et al., 2005).
Debt tenure and debt finance laws emerged as dominant factors influencing decisions on capital structure. Firstly, firms were found to remain dependent on short-term debts in response to liquidity crises despite the stringent laws banks put in place as part of banks' financial crisis mitigating measures (Deesomsak et al., 2004;Michaelas et al., 1999). Secondly, debt finance laws continued to govern firms' decisions on capital structure. Specifically, country-specific laws and context-dependent enforcement were found dominant in firms' capital structure choices. For instance, studies found that common law systems regulating equity and debt providers, that differed from one country to another, provided better protection than the provision on civil laws (La Porta et al., 1998). Thirdly, when corruption reports increased due to poor legal systems' offences on integrity, firms tended to struggle with increasing debts, particularly in dealing with short-term tenure debts (La Porta et al., 1998). Thus, firms' capital structure decisions not only depended upon inflation rates; regulations of debt tenure and debt finance laws were equally dominant in determining capital structure choices.
A dramatic increase in research in capital structure choices also suggested multinational and multidisciplinary approaches. For instance, Rajan and Zingales (1995) suggested that prospective research might better consider developments of theoretical models based on empirical findings that cut across nations. Because different countries regulated different forms of capital structure, macroeconomic circumstances (economic, legal, tax, and technological environments) determined firms' financing choices (Gleason et al., 2000;Korajczyk & Levy, 2003). By focusing on various global environment-related factors, economic crises could be mitigated particularly by taking into consideration the 1997 ASEAN economic crises. Thus, examining factors related to capital structure decisions across Malaysia's large firms and SMEs formed the central aim of the study. The following hypothesis concerning environment-related factors was posited: H6: There was a relationship between business environment and debts.

Culture
Since different forms of culture operate in different cultural environments, finance is generally understood differently in a particular society (Clugston et al., 2000). Culture, as described by Nwankwo and Lindridge (1998, p. 201), encompassed "race, religion, language group, shared history, and origin", while Hofstede et al (1991) defined culture as a "collective programming of the mind which distinguishes the members of one group or category of people from another". Schwartz's (1994) specific concepts on culture were used in this study as measurements of cultural dimensions as opposed to Hofstede's (1980), firstly because Schwartz's (1994) concepts had widely been used to test the theoretical influence of culture on capital structure decisions (Chui et al., 2002). Secondly, because Schwartz's (1994) concepts bore resemblance to individualised constructs, Schwartz's (1994) concepts were employed to understand country-specific variations on cultural dimensions, and how countryspecific variations were sufficient for hypothesis-testing. As Clugston et al (2000) argued, when independent variables were considered to measure cultural concepts, individualised measures must be used. Thus, Schwartz's (1994) cultural concepts played important roles to determine whether firms and SMEs' values, attitudes, thinking, beliefs, and behaviours influenced decisions on capital structure.
Two dimensions of culture, namely, 'conservatism' and 'mastery', were considered. On the one hand, items on 'conservatism' examined whether capital structure decisions had relationships with owners and employees who worked collectively to create harmonious relationships, preserve the public image, and tolerate uncertainty avoidance (Li et al., 2011;Shao et al., 2009;Breuer & Salzmann, 2008;Castro et al., 2007;Siegel et al., 2007;Chui et al., 2002;Licht, 2001). On the other hand, items on 'mastery', which were related to individual success, actions, and decisions, investigated whether individual satisfaction affected individuals' interests in materialising firms' success, adopting strict policies to protect firms, and choosing low-risk projects that accrued fewer debts (Hirshleifer & Thakor, 1989). Chui et al (2002) exemplified how culture determined capital structure decisions. It was reported that firms' with high scores on the construct of 'conservatism' utilised fewer debts in capital structures because owners and employees worked collaboratively to create harmonious relationships, preserve the public image, security, conformity, and tradition. As expected, firms' with high scores on the construct of 'mastery' used less debt financing, emphasised control, and celebrated individuals' success. Therefore, because a direct relationship between the impact of culture on finance was not investigated directly, the following model ( Figure 1) and hypotheses were formulated H7.1a: There was a positive association between conservatism and internal sources of finance. H7.1b: There was a negative association between conservatism and debt financing. H7.2a: There was a positive association between mastery and internal sources of finance. H7.2b: There was a negative association between mastery and debt financing.

Methodology-Procedures of Data Collection Instrument
As the study was designed to examine factors that influenced capital structure decisions across Malaysia's SMEs, Malay and English questionnaires were used. Two steps governed the process of designing the instrument. Firstly, a back-translation technique (Mullen, 1995) was used to translate Malay-to-English and English-to-Malay questionnaires. While Malay-to-English translations were carried out by the investigator, a certified bi-lingual translator wrote, scrutinised, and edited English-to-Malay questionnaires. The translator is a bi-lingual lecturer at a private university in Malaysia. Secondly, the content in the questionnaires was evaluated by five academics (i.e. three accounting and two finance lecturers from public universities in Malaysia) and three SME owners (i.e. they are the participants from qualitative study on this topic). Brislin's (1970) guide on questionnaires was used to evaluate how the questionnaires satisfied the face, content, and semantic equivalence of the translated instrument.

Units of Analysis
Units of analysis were defined in terms of firms' contexts. Firstly, only SMEs that consisted of a total manufacturing workforce of less than 200 or an annual sale turnover of RM 50 million were selected (SME Corporation Malaysia, 2020). For SMEs involved in the service industry, only SMEs with less than 75 employees or a sale turnover of RM 20 million were considered and chosen (SME Corporation Malaysia, 2020). Secondly, a total of 560 questionnaires were distributed to SMEs' owners from a variety of business backgrounds. 500 questionnaires were returned, indicating a high response rate of 89%.

Scales of Measurement
The scales of measurement adopted in this study were constructed in two ways. Firstly, the reviews of literature guided the selection of eight constructs. Secondly, the constructs identified were integrated for conciseness. Table 1 illustrates the examples of constructs and constructs' corresponding items:  (2015) 2. Standard operating procedures were helpful to employees 3. Harmonious working relationships and social harmony were important for the company 4. Instructions for operations were important for employees 5. Preserving public images was one of the main policies for the company

Content Validation
The process of content validation was carried out in two ways. Firstly, the full list of potential items was reviewed. Secondly, items that best captured the dimensions of capital structure determinants were carefully selected. As a result of a three-panel expert review, a final list of items was generated. In total, 68 items were generated across the following constructs: four items on 'business planning', five items on 'lifestyle goals', four items on 'commercial goals', three items on 'social welfare goals', four items on 'perceptions and beliefs', 16 items on 'relationship', 10 items on 'networking', seven items on 'stable environment', two items on 'benign environment', three items on 'external environment', five items on 'mastery', and five items on 'conservatism'.

Items Purification
68 items measuring 12 constructs of capital structure determinants were finalised. Specifically, 12 variables made up factors related to 'objectives' and 'goals', 30 variables built factors related to 'perceptions and beliefs', 'relationships', and 'networking', four variables made up factors related to 'business planning', ten variables built factors related to 'cultural dimensions', and 12 variables made up factors related to 'environment'. The items were grouped to ensure that the number of observations per item for all analyses was at least 5:1 (five participants per variable) (Cavusgil & Zou, 1994;Hair et al., 2010). As argued by Menon et al (1996), assessing fewer measurement models from various constructs would yield more reliable results.
Subsequently, a five-step purification process of measurement scales was carried out using scale reliability and EFA. SPSS 28.0 was used to measure coefficient alpha and item-tototal correlation values before EFA was applied. Firstly, EFA was applied by dividing the constructs into five groups based on theoretical constru-cts. Secondly, an eigenvalue greater than 1.0 was considered to determine factors to extract (Pallant, 2001;Hair et al., 2010). Thirdly, an oblique rotation was applied to initially extract factors due to inherent correlations among constructs. Fourthly, results from Kaiser-Meyer-Olin (KMO) Measure of Sampling Adequacy (above 0.5) and Bartlett's Test of Sphericity (p-values below 0.05) were analysed to assess the factorability of items. Only communalities with more than 0.50 were considered (Hair et al., 2010).
Finally, reliability analysis was carried out to test the reliability and internal consistency of all factors. Specifically, the coefficient alpha and item-to-total correlation for provisional dimensions were considered to evaluate the internal consistency of all variables. The reliability analysis revealed that the statistical criteria for item retention were higher than 0.35 for item-to-total correlation (Bearden et al., 2001;Hair et al., 2010). Pearson correlation's value was reported lower than 0.30, while coefficient alpha value was recorded higher than 0.5 (Nunnally, 1967).

Results and Discussions
The following sections highlight discussions pertinent to the results generated from the EFA. The sections describing the results of the current study are drawn based on the order of the questionnaires' content.

a. Business Planning
The EFA result across the four-item group on 'business planning' revealed that items loaded clearly on one factor. With a value of 0.861, the KMO measures for the items indicated meritorious sample adequacy (Kaiser, 1974). Bartlett's Test of Sphericity also showed significantly higher levels of associations with the factorability of the correlation matrix. In addition, the assessment of internal consistency reliability was positively associated with factors' coefficient alphas (α=0.96). The reliability of this factor was confirmed by Pearson inter-correlation for subsequent investigation on internal consistency. While the results revealed a significant value of 0.001, item-to-total correlations were found to be slightly higher than the threshold value (0.35).

b. Objectives and Goals
All variables concerning 'lifestyle', 'commercial', and 'social welfare' goals demonstrated relatively positive relations. With a value of 0.678, the combined KMO measurement on 'objectives' showed a mediocre sample adequacy (Kaiser, 1974). Moreover, Bartlett's Test of Sphericity revealed a higher level of factorability of correlation matrix because the test reached statistical significance (0.000). A three-factor solution was produced after the rotation. Two items were eliminated because they were cross-loaded. All remaining loadings were above 0.7 and the Cronbach's alpha value recorded for 'lifestyle goals' was at 0.898. Higher item-to-total-correlation values for all variables demonstrated values higher than 0.5. However, Cronbach alpha values of variables on 'commercial goals' and 'social welfare goals' were 0.634 and 0.514, respectively. It should be noted here that although the coefficient value was slightly higher (below 0.70), the coefficient value was still considered acceptable because the value was higher than 0.50 (Nunnally, 1967).

c. Relationships, Networking, Perceptions, and Beliefs
Based on KMO measurement, 30 items that measured 'perceptions and beliefs', 'relationships', and 'networking' demonstrated sampling adequacy of 0.745. The Barlett's Test of Sphericity was significant at p < 0.001, revealing that the R-matrix was negatively related to the identity matrix. The result suggested slightly higher levels of relationships between the variables. As a result of a three-factor solution, 16 items were eliminated because they were cross-loaded and that they produced communality values of less than 0.50. However, the second run of factor analysis demonstrated that all factor loadings indicated slightly above 0.7. While the Cronbach alpha values generated for items on 'networking' and 'perceptions and beliefs' were 0.85 and 0.817, respectively, the item-tototal-correlation values were slightly higher than 0.5 (Hair et al., 2010).
Analysis on the construct of 'relationship' revealed interesting results. Firstly, because a value of 0.30 was recorded for these items, namely, 'provided data to lenders', 'provided data to suppliers', 'reviewed services of lenders regularly', and 'reviewed services of suppliers regularly', they were eliminated to increase the reliability. Secondly, the subsequent reliability alpha tests which were re-analysed demonstrated a significantly higher value of Cronbach alpha (0.894) and an item-to-total-correlation value that was acceptably higher than 0.3 (Field, 2005). It should be noted here that none of the items would increase the reliability if they were deleted.

d. Business Culture Orientations
The factor analysis was conducted on ten items that measured 'business and culture'. One variable ('details of job requirements and instructions were important') was eliminated from the analysis because the communality value showed slightly lower than 0.5 (communality= 0.417). As demonstrated earlier, reviews on literature concerning finance and management were sourced to identify the variables on 'business and culture'.
In the second run of factor analysis, all factor loadings showed a significantly higher value than 0.7. The KMO measures for the items showed a value of 0.833, a 'meritorious' value (Kaiser, 1974). Barlett's test of sphericity χ ²(36) = 2003.108, p<0.001 indicated that the correlation between items was significantly high for PCA. For the dimensionality of the remaining items, the EFA result demonstrated that they loaded clearly (significantly higher than 0.5) on two factors; the first factor ('conservatism') consisted of five variables, while the second factor ('mastery') was found to load with four variables. The communalities values for all items were relatively higher than 0.5. In addition, the Cronbach's alpha for 'conservatism' and 'mastery' illustrated values of 0.893 and 0.897 respectively. Item-to-total-correlations reported for both constructs were significantly higher than 0.5 (Hair et al., 2010).

e. Environment
A combined KMO measurement score of items on 'environment' indicated a value of 0.675, a value equated with 'mediocre' sample adequacy (Kaiser, 1974). Bartlett's Test of Sphericity also reached statistical significance (0.000), supporting the positive and significant factorability of the correlation matrix. Five items were eliminated from the first run of EFA due to cross-loading and low communality (a value lower than 0.5). The outcome of the second run of EFA was considered appropriate because most variables loaded significantly high on two factors and all communalities' showed values higher than 0.5. Four items were found to load onto the first factor ('stable environment') and three items were found to load onto the second factor ('external environment'). While the reliability alpha of the first factor was 0.604, Nunnally (1967) argued that this value was acceptable because it recorded a value relatively higher than 0.50. However, the first factor's alpha value was significantly higher than the standard estimation of 0.70 (Nunnally and Bernstein, 1994). Subsequent internal consistency investigation depicted confirmed reliability; Pearson inter-correlation test indicated a significance at 0.001 levels for both factors. Item-to-total-correlations for all items were slightly higher than 0.3, a relatively good value (Field, 2005). It should be noted here that none of the items would increase the reliability if they were deleted.

Conclusions and Implications
Factors that influenced capital structure were identified and confirmed in this research, offering new insights on SMEs' capital structure decisions. Firstly, this study shed light on business planning, strategic plans, and management structure. In particular, ownersmanagers had a higher likelihood to materialise objectives and goals that they see fit in choosing their financing capitals. Secondly, SMEs agreed that establishing relationships and networking more intensively with lenders and suppliers were important. Thirdly, SME decision-makers exhibited a higher emphasis on perceptions and beliefs while dealing with financial decisions. Finally, the study emphasising factors that influenced SMEs' capital structure decisions equally recognised the interdependencies of internal and external factors, particularly culture and environment. Given the levels of factors influencing SMEs' decisions on capital structure, policymakers might better control, maintain, or improve economic systems, legal issues, tax environment, and inflation rates that complement SMEs to access various financial sources.
All theoretical, operationalisation of constructs, and methodological underpinnings as demonstrated in this study contributed to prospective research. Firstly, a more nuanced understanding concerning capital structure decision factored in business planning, goals, relationship, networking, cultural orientations, environment, and owners' perceptions and beliefs. Relevant attitudinal factors that were commonly associated in Western-imposed settings were all accounted for concerning Malaysia's SMEs.
Secondly, potential interdependencies among perceptions, culture, and environment shed light on the feasibility of how concepts were operationalised. Specifically, while nearly all constructs employed in this research had been investigated previously, their operationalisations, in academic practice, were rare as a likely result of inadequate findings on measurement scales' validity and reliability. Hence, an analysis of questionnaires to Malaysian SMEs presented in this study not only revealed significant constructs but also illustrated evidence concerning construct validity and reliability of previous scales.
Thirdly, the findings presented in this study on factors that determined capital structure complemented recent empirical research. For instance, while studies carried out by Michaelas et al (1998); Romano et al (2001) suggested that owners' perceptions, business environment, and culture determined capital structure decisions, unfortunately, these factors were not tested in their studies. Moreover, while far fewer studies focused on owners' perceptions, business environments, and culture, this research emphasising capital structure decisions replicated and extended previous studies by (Chui et al., 2002;Gaud et al., 2003;Gleason et al., 2000;Norton, 1990). Finally, given the prevalence of factors influencing capital structure decisions, the framework as presented above can be introduced as a standard measurement among SMEs. By focusing on financing preferences and factors influencing financing choices, prospective development of a model could be facilitated.