Access to Finance, Financial Development and Firm Ability to Export: Experience from Asia Pacific Countries

Transcription

1 Asian Economic Journal 2018, Vol. 32 No. 1, Access to Finance, Financial Development and Firm Ability to Export: Experience from Asia Pacific Countries Durairaj Kumarasamy and Prakash Singh Received 13 January 2016; Accepted 27 December 2017 With particular reference to Asia Pacific countries, the present study examines how access to finance and financial development affects firms ability to enter export markets. Using firm-level data from the World Bank Enterprises Survey, we found that access to finance plays a significant role in improving firms ability to export. In addition, development of the financial sector fosters export market entry. Among the financial development indicators, reach of the banking sector variable is most prominent. The present study suggests that improvements in access to finance and financial development (increases in the reach of the banking sector) enable firms operating away from capital or major cities to enter export markets easily. The present study supports policy intervention to strengthen access to the financial sector, which would encourage firms to export, and to facilitate export market entry for remotely located firms. Keywords: access to finance, exports, financial development, international trade. JEL classification codes: D22, F14, O16, O53. doi: /asej I. Introduction Exports make an indispensable contribution to economic growth. Higher exports lead to increases in income levels, more efficient allocation of resources, improved capacity utilization, greater exploitation of scale and technological Kumarasamy: ASEAN India Centre, Research and Information System for Developing Countries (RIS), India Habitat Centre, Lodhi Road, New Delhi , India. Singh (corresponding author): Centre for Regional Trade, 715, IIFT Bhawan, B-21 Qutab Institutional Area, New Delhi , India. prakash.archa@gmail.com. This research was supported by an ARTNeT post-workshop grant awarded through the project Impact of trade facilitation measures on poverty and inclusive growth funded by the Government of China and implemented by ARTNeT, UNESCAP, Bangkok. The authors are grateful to the ARTNeT Secretariat for helpful comments and guidance on an earlier draft of this paper and for technical support in disseminating this paper. We thank participants of the RIS Breakfast Seminar for their input. We thank Professor Vigneswara Swamy for his critical input. We also thank Professor Prabir De for his constant encouragement. The authors gratefully acknowledge the comments and suggestions received from an anonymous reviewer of the Journal, which have helped in improving the paper substantially East Asian Economic Association and John Wiley & Sons Australia, Ltd

2 ASIAN ECONOMIC JOURNAL 16 improvements in response to greater competition from abroad (Burney, 1996). Therefore, export promotion strategies have been adopted in several developing countries to boost economic growth, including policy support to encourage nonexporters to enter foreign markets and also to facilitate the export expansion of existing exporters. As a result, there has been a significant increase in global trade in general and trade within Asia Pacific countries in particular. This is further reflected in the trade performance of Asia Pacific countries, which has shown significant improvement over the past two decades. The contribution of trade in the global value chain and regional production network has reached 75 to 80 percent of the total trade despite the global financial crisis (ESCAP, 2015). Asia Pacific has become the largest trading region, with a 37-percent share of world trade and almost half of Asia Pacific trade occurring within the region (UN-ESCAP, 2014). In the global competitive market, exporting firms face huge challenges to match the competitive prices and keep up with technological changes. This has lead firms to invest heavily in R&D, marketing research, product development, advertising and fixed capital (Hur et al. 2006). Thus, exporting firms are more dependent on external finance to meet their liquidity needs. 1 Hence, firms with financial constraints would face difficulty entering into and remaining in the export market. Given the competitiveness of the global market, existing firms essentially require additional capital to move up in the value chain. Limited access to capital holds firms at low value-added stages of the supply chain and restricts them from exploiting profitable opportunities. In addition, firms newly entering into the export market require additional investment as a large part of this investment is sunk and upfront in nature. Access to financial instruments enables firms to boost their export competitiveness by allowing them to overcome the liquidity problems associated with export activities. Several studies discuss how financial development reduces firms borrowing costs, thereby encouraging firms to enter into the export market (e.g. Beck, 2002, 2003; Svaleryd and Vlachos, 2005; Manova, 2013). Other studies highlight that financial constraint is potentially an important cost of international trade in a financially underdeveloped country (Manova, 2013). Therefore, well-developed financial markets and strong banking institutions are crucial for firms exporting activities. Despite the empirical research emphasizing the role of access to finance in firm performance, access to finance is still a major problem at the firm level in several developing countries (Beck, 2002), especially in developing countries of Asia where the financial system remains far below the standards of developed countries. Furthermore, effect of the 2008 financial crisis transmitted to developing countries in Asia has affected their credit disbursement. The decline in trade finance has been more pronounced in countries with less developed financial 1 Up to 90 percent of world trade has been estimated to rely on some form of trade finance (Auboin, 2009).

3 ACCESS TO FINANCE, FINANCIAL DEVELOPMENT AND FIRM EXPORTABILITY 17 sectors (Liston and McNeil, 2013). As a result, these countries have introduced extensive post-crisis policy interventions to improve the financial health and performance of their banking systems (Estrada at. al., 2010), to enhance the lending environment. In this backdrop, it is worth investigating whether access to finance and financial development promote firm entry into export markets in Asia Pacific countries. A small number of studies have examined the effect of access to finance on firm exportability for developing countries, with very few of these based on survey data. There is also a dearth of studies for the Asia Pacific region. Using World Bank Enterprises Survey (WBES) data, this study attempts to fill this gap and investigates how access to finance and financial development affects the exporting decisions of firms in Asia Pacific countries. We also examine whether access to finance and financial development could eliminate the location effect, as half of the small and medium sized firms operate away from capital or large cities in the Asia Pacific region. Several studies have measured access to finance based on balance sheet data but have not clearly indicated the productive channel of investment and have had measurement and endogeneity issues (e.g. Bellone et al., 2010). Therefore, we have considered access to finance as the proportion of investment and working capital funded by formal finance. This measure helps to map sources of finance to its uses. In addition, we have addressed the endogeneity issue by using several instruments for access to finance and tested for robustness using the instrumental variable bivariate probit regression method. Our study finds that in addition to size, age, foreign ownership and productivity, access to finance and financial development also determines firms export decision. Firms with access to finance have a greater likelihood of entering the export market. Development of the financial sector stimulates firms decision to enter the export market. Interestingly, with respect to the financial sector we found that reach of banking sector increases the probability of export decision for firms located away from capital or major cities. Thus, this suggests that, at first, access to the financial sector itself will bring huge change in enterprise exporting behavior. The remainder of the paper is organized as follows. Section II reviews the earlier studies related to the study area. Section II describes the theoretical underpinnings and Section IV provides the data source, methodology and a description of the variables. Section V presents the empirical results. Section VI includes a robustness check and, finally, Section VII concludes and discusses the policy implications of our findings. II. Previous Literature The present paper broadly relates to the theoretical literature on the effect of financial development in general and financial constraints in particular on firm

4 ASIAN ECONOMIC JOURNAL 18 investment decisions and, subsequently, their growth (Fazzari et al., 1988). More specifically, the study is linked with literature on the impact of financial factors on firm exports (Chaney 2005; Manova 2013). In one of the earliest works on the topic, Kletzer and Bardhan (1987), using the Heckscher Ohlin framework, compared two international trade models with the same factor endowments. However, in one of their models they introduced external finance for working capital and showed that credit market restrictions determine a country s specialization for sectors, which depends more on external finance. For instance, a country with better financial intermediation specializes in the sectors that rely extensively on external finance, whereas a country with a repressed financial sector specializes in the sectors which are less dependent on external finance. Hence, they argue that financial development favors industries that rely heavily on external finance and that, in turn, financial development could explain the variance in the trade structure across countries. In contrast, Baldwin (1989) argues that as financial development helps in diversifying risk, countries with developed financial markets tend to specialize in producing goods that are risky and with relatively lower risk premiums. Following Kletzer and Bardhan (1987), Beck (2002) analyzes the role of financial development in international trade and suggests that having a developed financial system increases the export share and improves the trade balance for manufactured goods. Berman and Hericourt (2010) find that financial development disproportionally increases the probability of the export decision of more productive firms. In extending Melitz s (2003) model, Chaney (2005) suggests that financial constraint has a strong bearing on whether firms decide to export. Using a dataset on export transactions at the firm level for the Belgian manufacturing sector, Muûls (2008) put forward that firms with higher productivity level and lower credit constraints are more likely to be exporters. Using panel data for French firms, Bellone et al. (2010) reveal that financial constraint reduces firm export market participation. Similarly, Manova (2013) reveals the importance of financial constraints in firms decision to export. In contrast, Feenstra et al. (2014) find that higher trade costs for exporters (measured in terms of time to export) make exporters more financially constrained compared to domestic firms. On a similar note, Greenaway et al. (2007) suggest that in the relationship between finance and firm exports, the causal direction runs from exports to finance rather than the reverse as claimed by other studies. Bridges and Guariglia (2008) present a similar result. In contrast, according to Stiebale (2011), financial constraints and firm exports are not correlated. Alvarez and Lopez (2014), in assessing the impact of access to finance on the exporting decisions of Chilean manufacturing firms, found that access to finance is more important for firms in industries that are more dependent on external finance. Thus, the existing literature does not provide a clear picture about the relationship between financial constraints and firms decision to export; hence, further verification is needed.

5 ACCESS TO FINANCE, FINANCIAL DEVELOPMENT AND FIRM EXPORTABILITY 19 III. Theoretical Underpinnings One strand of the literature suggests that financial friction affects investment decisions (Stiglitz and Weiss 1981; Fazzari et al. 1988). The decision to enter the export market requires additional investment in terms of market search, building distribution networks, additional staff, promotion and product modification. Hence, firms must incur sizable fixed costs to enter the export market and this fixed cost has to be paid upfront (Chaney 2005; Manova 2013). Thus, firms intending to enter the export market require higher liquidity and, hence, need access to additional external finance. In addition to fixed cost requirements is firms inability to provide information (i.e. about their repayment capacity) to the financer. Furthermore, it can be difficult for a financer to verify information provided by firms. This makes financing difficult. Limited access to external finance limits firms scale of operation (e.g. sales, number of products and number of export destinations) (Secchi, et al., 2012). Furthermore, in contrast to firms engaged in the domestic market, those engaged in the export market also face additional variable costs. Cross-border shipping and delivery usually take days longer to complete than domestic orders (Djankov, et al. 2010), which further intensifies the working capital requirements of exporters relative to those of non-exporters. Consequently, to meet such financial requirements, firms have to rely on external finance, which is costly in comparison to internal finance. This dependence of firms on external finance is exhibited by the size of lending for financing of international trade. 2 Thus, firms inability to access external finance, particularly formal finance, is crucial in their decision to enter the export market. Hypothesis 1: Financially constrained firms are less likely to enter the export market. As exporting firms tend to face higher liquidity constraints and require more external finance, development of the financial sector is important in improving the access to finance for firms. Financial sector development, by reducing credit constraints, will drive investment, and all firms with productivity above a threshold level will choose to become exporters (Melitz 2003). Having weak and inefficient financial institutions increases the liquidity constraint in the domestic market, and, as a result, prevents a subset of productive firms from entering the export market (Chaney 2005). Financial intermediaries are effective in promoting entrepreneurs who are either engaged in or would likely to be engaged in more productive activities. Financial intermediaries also help in the accumulation of human capital (Jacoby 1994). Financial friction due to an 2 As reported in Chor and Manova 2012 the volume of financial activities linked to trade is equal to trillion in 2008 and according to the estimates of Auboin 2009 up to 90% of world trade has been estimated to rely on some form of trade finance.

6 ASIAN ECONOMIC JOURNAL 20 underdeveloped financial sector proves to be critical for export promotion at enterprise level in particular and country level in general. Hypothesis 2: Financial development is instrumental in firms export decision. Based on the above discussion and hypothesis, this study examines the role of access to finance and financial development in promoting firms export decision. IV. Data and Methodology IV.1 Data For the analysis, we have used both firm-level and country-level information. World Bank Enterprise Survey (WBES) data are the source of firm-level information. We have combined two different waves of WBES data for a longer time span. 3 It is important to note that WBES data is pooled cross-section data and not panel data. As all the questions asked to the enterprise were not the same in the two waves, only common questions across the waves were included. The present study focuses on Asia Pacific countries. Based on data availability, we have included 16 countries for the analysis: Afghanistan, Bangladesh, Bhutan, Cambodia, China, India, Indonesia, Laos, Malaysia, Myanmar, Nepal, Pakistan, Philippines, Sri Lanka, Thailand and Vietnam. The total number of firms surveyed in these countries was around Information collected in the survey is standardized to maintain comparability of enterprise-level information across the countries. The firm-level data is merged with country-level data to gauge the effect of country-level variables. Country-level variables include measures of financial development, which are collected from the World Development Indicators (WDI). 4 IV.2 Methodology The present study examines the factors responsible for firms decisions on export market entry in general and analyzes the role of access to finance and financial development in particular. The decision of a firm to enter or not to enter the export market is unobservable directly from the data; hence, we have constructed this variable from the WBES survey data based on the percentage share of both direct and indirect exports in total sales. 3 The first wave of the survey happened during 2002 to 2006, and enterprise-level information was drawn for around 71,000 enterprises. In the second wave, which is still ongoing, data is collected for more than 140,000 enterprises. 4 The online version of the WDI has been used for data collation.

7 ACCESS TO FINANCE, FINANCIAL DEVELOPMENT AND FIRM EXPORTABILITY 21 Let P j * be the benefit accruing to a given firm j (j =1,2,3 n) located in country c (c = 1,2,3 m) from sales in the export market. The benchmark equation can be specified as: P * jsc = α 0 + α 1 X j + α 2 Y i + ε j, ð1þ where X j represents an array of firm-level factors or internal attributes, Y i is a list of factors outside the firm (e.g. industry-level characteristics, country-level attributes, location and sector). ε j is the random error term. The dependent variable P j * is not observed because it is a latent variable. Hence, the following probit model is defined as: ( P j = 1 for P* j >0 0 for P j *, ð2þ 0 where P j is a binary variable with the value 1 if the firm is engaged in export activities (export share in sales is positive) and 0 otherwise (operating in domestic market only). Let Φ() depict the cumulative standard normal distribution function. Then, the probit regression model can be represented as: EP j jx j,y i = Φ α0 + α 1 X j + α 2 Y i : ð3þ IV.3 Empirical model As firms decision to export is determined by both internal and external factors, in our empirical model, we have considered firms age, size, productivity, ownership and access to finance as internal factors, and financial development and location as external factors. The empirical model to estimate firm export decisions is as follows: Exporter i = α 0 + α 1 Age i + α 2 Size i + α 3 Productivity i + α 4 Foreign i + α 5 Access Finance i + α 6 Location i + α 7 Z i + ε i, ð4þ where Exporter is a dichotomous variable to estimate firms decision to export. A firm is said to be an exporter (exporter = 1) if the firm s direct or indirect sales are positive; otherwise an exporter assumes the value 0 if the firm s direct or indirect exports equal 0. Age of the firm is measured as the log of difference in the survey year in the country and year of firm inception. Roberts and Tybout (1997) suggest that age increases firms propensity to export. They argue that experience and familiarity with the production process comes with age and, hence, older firms are more

8 ASIAN ECONOMIC JOURNAL 22 efficient than younger firms. Efficient firms tend to survive in the market for longer than inefficient firms and are, therefore, more likely to enter the export market. In contrast, it is also possible that younger firms are more competitive and can adopt efficient production processes, which could further increase their likelihood of exporting. Several studies have shown mixed results. Ottaviano and Martincus (2011) find an insignificant relationship between older and younger firms, Alvarez and Lopez (2005) report a negative relationship, although the effect of age is very small, and Duefias-Caparas (2006) presents mixed results, with some sectors (e.g. clothing) having a positive relationship and some sectors (e.g. electronics and food processing sectors) having a negative relationship. Therefore, the coefficient of age can have both signs in the estimated empirical model. Size of the firm is a categorical variable (with small-0, medium-1 and large-2 categories). Firms with fewer than 20 workers including permanent and temporary workers (number of temporary workers is adjusted for number of days a worker worked in the year to total number of days in the year) are classified as small firms (0). Firms with workers are classified as medium-sized firms (1) and large firms (2) are those with 100 and more workers. The base category is small firms. Several theoretical and empirical studies reveal that firm size is one of the important predictors of their export decision and also increases the likelihood of exporting. Large firms can absorb sunk costs due to economies of scale, which facilitates entry to the export market (Roberts and Tybout, 1997). Some studies argue that an increase in firm size makes firms more productive and reduces the marginal cost of production, which, in turn, would improve firms exportability (Bernard and Jensen, 2004; Ottaviano and Martincus, 2011). Therefore, we expect the coefficient of firm size to be positive. Firm productivity is measured as the log of capacity utilization of the firm. The capacity utilization of the firm is defined as the percentage by which a firm can exploit their input capacity. The literature pertaining to what determines whether a firm decides to export suggests that firms that are highly profitable tend to export. As with higher productivity, a firm is likely to enjoy more profitability; hence, productivity will have positive effects on firm exportability (Roberts and Tybout, 1997). Thus, we expect productivity will assume a positive coefficient. To capture the effect of ownership (foreign vs domestic) we have included the dummy variable Foreign. Foreign equals 1 if the foreign holding in the firm is equal to or more than 10 percent; otherwise it is 0. The reference category is domestic firms (Foreign = 0). The existing literature considers ownership as one of the instrumental variables which governs firms export market entry decision (Bernard and Jensen, 2004; Greenaway et al., 2007). Foreignowned firms usually have better access to information related to the foreign market. In addition, foreign-owned firms tend to have superior resources (e.g. human and technological) and stronger business relationships, coupled

9 ACCESS TO FINANCE, FINANCIAL DEVELOPMENT AND FIRM EXPORTABILITY 23 with the advantage of distribution networks, which facilitates their exporting decision. Location, which is also a categorical variable, is included in the model to capture the effect of location (geographical location of firms). It takes the value 0 if the firm is located in a capital city or a city with a population of 1 million plus and 1 if the firm operates from a city with fewer than 1 million people. Firms located in capital cities and cities with 1 million plus people are considered as the base category. The literature suggests that firms located in capital cities or in major cities have easy access to markets for both inputs and outputs, which reduces their marginal costs. In addition, superior infrastructure availability reduces costs, which, in turn, increases their probability of exporting (Elbadawi et al., 2001). Access to finance (Access finance) is measured in terms of access to formal finance and access to bank finance. Both the variables are categorical in nature. Access to finance-formal has a value of 1 if more than 50 percent of a firm s working capital or fixed asset is financed through formal sources (formal sources include government banks, private banks and non-banking financial companies); otherwise it is 0. Similarly, Access to finance-bank has a value of 1 if more than 50 percent of a firm s working capital or fixed assets are financed through banks (government or private); otherwise the value is 0. For both the access to finance variables, the reference category is 0. Given the high sunk costs involved, exporting requires a higher level of investment. Access to finance leverages firms to meet this additional investment requirement and positively affects firms decision to enter the export market (Chaney, 2005; Bellone et al., 2010). Following the argument, we expect that the coefficient for access to finance would be positive. Z represents year, industry and country fixed effects. Year dummies are included to control for time-specific effects or shocks that are common across all firms (such as financial crises, exchange rate and trade policy agreements at world level). Again, industry dummies are included to control industryspecific effects. It may be possible that some industries are, by design, more oriented towards the export market and, at the same time, it may be possible that some industries are more oriented towards the domestic market. Furthermore, some industries are more dependent on external finance. Thus, it is necessary to control these industries biases towards export market entry. Similarly, country fixed effects are included to absorb country-specific effects on firms decision to enter the export market. Country-specific bias arises as some countries promote exports by formulating policy supporting exporting firms. Furthermore, the present study analyses the role of financial development in firms decision to export. Equation (4) is extended after inclusion of financial development indicators as follows:

10 ASIAN ECONOMIC JOURNAL 24 Exporter i = α 0 + α 1 Age i + α 2 Size i + α 3 Productivity i + α 4 Foreign i + α 5 Access Finance i + α 6 Location i + α 7 FD i + α 8 Z i + ε i, ð5þ where, Age, Size, Productivity, Foreign, Access Finance and Location are the same as in the previous empirical model. FD i represents an array of financial development indicators. The theoretical literature advocates that financial development is skewed towards industries that are more dependent on external finance (Kletzer and Bardhan 1987). Drawing from the published literature, we expect that financial development, by reducing information asymmetry, improves the overall liquidity position of a country, which, in turn, improves the investment environment of enterprises. Furthermore, financial development could improve the exportability of firms by putting pressure on managers investment decision. Thus, the combined effect would increase the likelihood of firms deciding to export. Hence, the coefficient of financial development is expected to have a positive sign in the model. We have used the market capitalization to GDP ratio, the credit to GDP ratio and bank branches per thousand population (BBptp) as a proxy to measure financial development. V. Results and Discussion V.1 Descriptive statistics Table 1 presents the summary statistics of the variables for exporting and nonexporting firms. The mean of age for exporting and non-exporting firms suggests that exporting firms are older. Similarly, the mean of size indicates that exporting firms are larger than non-exporting firms. In comparing the mean value of productivity, we found that exporting firms are more productive as compared to their non-exporting counterparts. In terms of location, there is not much difference between the two location categories. Foreign-owned firms are found to be more export-oriented. With regard to both access to formal and bank finance, non-exporting firms have less access to finance vis-à-vis exporting firms. A high mean value of financial development for exporting firms also suggests that financial development improves firms exportability. V.2 Impact of access to finance on firm exportability We first analyze hypothesis 1; that is, the effect of access to finance on firm exportability using Equation (4). The estimation results based on the probit model are presented in Table 2. We use two sets of dependent variables to test the sensitivity of the estimation results: first with the percentage of total exports (including direct and indirect) to total sales (column 1 and 2) and second with the share of total export to total sales greater than 10 percent (columns 3 and 4 of Table 2). We also carry out the same set of analysis for two different

11 ACCESS TO FINANCE, FINANCIAL DEVELOPMENT AND FIRM EXPORTABILITY 25 Table 1 Summary statistics N Mean Standard deviation Minimum Maximum Firm age Exporter Non-exporter Firm size Exporter Non-exporter Capacity utilization Exporter Non-exporter Location of the firm Exporter Non-exporter Foreign ownership Exporter Non-exporter Access to finance-formal Exporter Non-exporter Access to finance-bank Exporter Non-exporter Market capitalization Exporter to GDP ratio Non-exporter Credit to GDP ratio Exporter Non-exporter Bank branches per thousand populations (BBptp) Exporter Non-exporter Source: Authors calculation based on WBES and WDI database. measures of access to finance variables: access to finance-formal and access to finance-bank. We start with the effect of firms age on their export decision, captured using log of age. The coefficient of the variable age is positive and statistically significant, which implies that older firms are more likely to enter the export market. The results for age are consistent with the finding of Roberts and Tybout (1997) that with age firms acquire technical know-how and, thus, older firms are more likely to engage in export activity. We observe a positive and significant coefficient of productivity (capacity utilization), which suggests that firms with higher productivity find the export market more attractive. The result is consistent with Roberts and Tybout s (1997) findings but contrary to the results of Greenaway et al. (2007) and Bellone et al. (2010), who find that productivity is not significantly related to firms decision to export. Does the size of a firm influence their decision to export? Given that the coefficient of the medium and large firm size dummy is positive and increasing, it can be argued that size does influence firms decision to export and the result is in coherence with the findings of Levenson and Willard (2000). Furthermore, a significant positive sign with an increasing coefficient size for medium and large firms suggests that with an increase in the size of a firm, the chances of entry in the export market increase.

12 ASIAN ECONOMIC JOURNAL 26 Table 2 Access to finance and firm exportability (1) (2) (3) (4) Coefficient Marginal effect Coefficient Marginal effect Coefficient Marginal effect Coefficient Marginal effect Log of age 0.108*** 0.033*** 0.122*** 0.042*** 0.054*** 0.015*** 0.063*** 0.019*** (0.012) (0.014) (0.013) (0.015) Log of capacity utilization 0.113*** 0.035*** 0.129*** 0.044*** 0.128*** 0.035*** 0.144*** 0.044*** (0.028) (0.034) (0.030) (0.036) Size (medium) 0.610*** 0.188*** 0.606*** 0.208*** 0.593*** 0.161*** 0.577*** 0.175*** (0.027) (0.034) (0.029) (0.037) Size (large) 1.343*** 0.414*** 1.358*** 0.466*** 1.246*** 0.338*** 1.243*** 0.377*** (0.028) (0.036) (0.030) (0.038) Access to formal finance 0.169*** 0.052*** 0.115*** 0.031*** (0.020) (0.021) Location (fewer than 1m people city) 0.089*** 0.027*** ** ** ** ** (0.021) (0.026) (0.022) (0.027) Foreign firm 0.880*** 0.271*** 0.862*** 0.296*** 0.813*** 0.220*** 0.793*** 0.241*** (0.034) (0.041) (0.033) (0.040) Access to bank finance 0.059* 0.020* (0.023) (0.024) Constant 6.178*** 4.962*** ** (0.349) (0.636) (0.669) (0.985) Pseudo-R Observation Notes: Robust standard error in parentheses. Significance level: ***p < 0.01, **p < 0.05, *p < 0.1. Models (1) (4) includes country, industry and year fixed effects.

13 ACCESS TO FINANCE, FINANCIAL DEVELOPMENT AND FIRM EXPORTABILITY 27 Looking at the coefficient of the variable location, we find that it is consistently negative and statistically significant. A negative coefficient of the variable implies that firms operating in small cities find it difficult to access the export market. Operating away from capital cities or big cities probably increases the production cost for firms as they do not have ready access to markets, technology and infrastructure and, thus, it is difficult for these firms to break into the export market. In line with expectations, the coefficient of foreign ownership is positive and significant at 1 percent. This shows that foreign ownership and exporting are positively correlated. Our results concur with those of Greenaway et al., (2007) and suggest that multinational firms have superior technological capabilities and are better placed against domestic firms in terms of networking, which increases their odds in favor of exporting. Having established the relationship between age, size, productivity, location and foreign ownership and firm exportability, we now move on to analyze the effect of the impact of access to finance on firms decision to export. Columns 1 and 3 of Table 2 display the results with formal finance as the measure of access to finance, whereas columns 2 and 4 exhibit the results for access to bank finance. As expected, the coefficients of both the access to finance variable (formal and bank) have positive signs. This indicates that firms with access to finance (formal or bank) have a higher probability of entering the export market. It is also clearly depicted in Figure 1 that access to finance increases the number of firms in the export market. Comparing the same across the different size groups we found that the effect is clearly visible for medium and large firms, whereas for small firms the relationship is not strongly evident. Access to finance would not only reduce the fixed costs involved in the export decision but also smooth the production friction due to delays in payment from buyers and, thus, support the results of Chaney (2005), Bellone et al. (2010) and other studies. We have also calculated the marginal effects (see Table 2). The marginal effect of the access to finance (formal) variable suggests that keeping everything constant, firms with access to finance (formal) have a 3 5 percent higher likelihood of entering the export market compared to their counterparts and also the change is statistically significant in nature. The strongest effect we see is for the size variable (large), which suggests that after controlling other variables, a large firm has a 40-percent greater probability of entering the export market entry compared to a small firm. Again, the effect is significant in a statistical sense. V.3 Financial development and firm exportability After having examined Hypothesis 1 we now analyze the role of financial development in firm exporting decisions (Hypothesis 2) using the specification in Equation (5). The results of the estimated model are presented in Table 3. The coefficient of the variables indicates that even after including indicators of financial development, age, size and location continue to have the expected signs and

14 ASIAN ECONOMIC JOURNAL 28 Figure 1 Small firm Access to finance and firms export decision Medium firm (mean) access_formal (mean) access_formal (mean) export_dind Fitted values (mean) export_dind Fitted values Large firm All firms (mean) access_formal (mean) access_formal (mean) export_dind Fitted values (mean) export_dind Fitted values Source: Based on authors calculations using WBES data statistical properties. Similarly, the coefficient of foreign ownership is positive and significant at the 1-percent level and it is consistent across the models. With regard to the effect of financial development indicators, the results reveal that the coefficient of the market capitalization to GDP ratio is positive and significant at the 1-percent level. Similarly, the coefficient of other indicators of financial development is also positive and statistically significant. Thus, the results confirm the indispensable role of financial development in firms decision to export. The coefficient size of the stock market to GDP and the credit market to GDP is small compared to that of bank branches. This may be because, except for China and India, Asia Pacific countries in the sample do not have well developed capital markets. Therefore, the impact of the development of capital markets on firm exportability is not truly captured. It is also illustrated in Figure 2 that the firms operating in financially developed countries are better off when they enter the export market in contrast to firms operating in less developed financial market. Besides, small firms are not able to make use of financial development compared to medium and large firms due to inaccessibility to the financial markets. To understand how the export decision is responding to the rate of change in the explanatory variable we have also calculated the marginal effect at the mean (see Table 4). The calculated average marginal effect suggests that except banking sector reach, other indicators of financial development are not very strong drivers of firms decision to export. The marginal effect of

15 ACCESS TO FINANCE, FINANCIAL DEVELOPMENT AND FIRM EXPORTABILITY 29 Table 3 Access to finance, financial development and exportability Dependent variable-exporter (1) (2) (3) (4) (5) Log of age 0.126*** 0.083*** 0.129*** 0.129*** 0.130*** (0.013) (0.017) (0.015) (0.015) (0.015) Log of capacity utilization 0.115*** * 0.079* 0.080* (0.029) (0.035) (0.031) (0.031) (0.031) Size (medium) 0.634*** 0.559*** 0.619*** 0.618*** 0.619*** (0.029) (0.037) (0.030) (0.030) (0.030) Size (large) 1.350*** 1.383*** 1.327*** 1.326*** 1.329*** (0.031) (0.038) (0.032) (0.032) (0.032) Location (fewer than 1m people in city) 0.051* *** 0.231*** 0.188*** (0.021) (0.028) (0.024) (0.041) (0.030) Foreign firm 0.915*** 0.892*** 0.863*** 0.867*** 0.861*** (0.037) (0.042) (0.042) (0.024) (0.035) Access to finance-formal (AF) 0.184*** *** 0.209*** 0.132*** (0.021) (0.029) (0.024) (0.042) (0.042) Market Capitalization to GDP 0.004*** (0.001) Credit to GDP 0.005*** (0.001) Bank Branch per thousand 0.365*** 0.362*** 0.376*** population (BBptp) (0.102) (0.102) (0.102) Location*BBptp 0.007** (0.002) Location*AF 0.139** (0.045) Constant 2.737*** 1.969*** 3.081*** 3.061*** 3.102*** (0.146) (0.167) (0.341) (0.341) (0.341) Pseudo-R Observation Notes: Robust standard error in parentheses. Significance level: ***p < 0.01, **p < 0.05, *p < 0.1. All specification includes country, industry and year fixed effect. market capitalization and the bank credit variable is only 1 percentage point, whereas in contrast, for the variable which measures the reach of the banking sector, the marginal effect is 10 to 11 percentage points. Similarly, the access to finance (formal) variable has a significant and strong effect in driving firms decision to export (1 6 percentage points). Furthermore, given that half of the small and medium sized firms are located away from capital cities or cites with more than 1 million people, we tried to understand the effect of access to finance and financial development on firm exports through the channel of location. For this we have estimated the model by interacting Location with BBptp and the access to finance variable (see columns 4 and 5 of Table 3). The result reveals that the coefficient of location and its interaction with the reach of financial sector variable is significant with the opposing sign, which suggest that as the reach of the financial sector increases,

16 ASIAN ECONOMIC JOURNAL 30 the negative effect of location diminishes. Thus, with financial sector development (reach of banking sector) and access to finance, regardless of the location of the firm, the likelihood of firms entering the export market increases. Therefore, the result reveals that financial sector development in general and access to finance in particular are instrumental in firms decision to export, and, hence, the results suggest the need for policy intervention to strengthen the financial market to promote export activity. In sum, the results suggest that the internal factors of age, size, productivity, location and foreign ownership are strongly related with firms decision to export. These results confirm the findings of previous studies (e.g. Robert and Taibout, 1997). Access to finance and financial development are also important determinants of firm exportability. Furthermore, the results on the interaction of location with access to finance and BBptp suggest that access to finance and financial sector reach would enhance export market entry chances of firms located away from capital or big cites. These results support previous empirical work on the relationship between finance and trade (Kletzer and Bardhan 1987; Beck 2002; Bellone et al., 2010; Berman and Hericourt, 2010). VI. Robustness Checks The existing literature suggests that firms decision to export and financial constraint could be endogenous. Financially constrained firms find it difficult to enter the export market, and firms who enter the export market obtain better access to finance. Thus, there is a possibility of endogeneity or reverse causality between access to finance and firms decision to export. To control for potential endogeneity and to check the robustness of the results, we have used the instrumental variable (IV) bivariate probit regression technique. Both banking and non-banking financial institutions scrutinize loan applications based on the proposal of the investment, past performance, asset holding and reliability of the firm. Therefore, for instrumenting access to finance, we include ISO certification, overdraft or line of credit, and whether or not the firm paid collateral for the loan. ISO certification captures the goodwill of the firm, as it reflects firms pursuit in following the international standard and upgrading technology to meet the competitive requirements. Collateral for loans acts as an assurance for the lenders in granting loans. Overdraft or loan by a firm indicates the loan history of the firm and how effectively the firm managed to pay the loan over the loan repayment period, thus overdraft or loan improves the information availability and credit score of the enterprise 5. The results of the estimated IV bivariate probit model are presented in Table 5. The IV test for underidentification and weak instruments suggest that 5 We are constrained to the available questionnaire information, although there may be a better proxy for access to finance to use as an instrumental variable.

17 ACCESS TO FINANCE, FINANCIAL DEVELOPMENT AND FIRM EXPORTABILITY 31 Figure 2 Small firm Financial development and firms export decision Medium firm (mean) credit_gdp (mean) credit_gdp (mean) export_dind Fitted values (mean) export_dind Fitted values Large firm All firm (mean) credit_gdp (mean) credit_gdp (mean) export_dind Fitted values (mean) export_dind Fitted values Source: Based on authors calculations using WBES data among all the instruments tried, overdraft or line of credit and collateral requirement are statistically better instruments, whereas ISO certificate is a relatively weak instrument (see table 5). The IV biprobit regression result suggests that even after controlling for endogeneity issues, results are, by and large, the same as for the probit model. The signs of the coefficients of the variables for age, productivity, size, location and foreign ownership have not changed and are they are also statistically significant. The measure of access to finance continues to have a positive and significant coefficient. The positive and significant coefficient of the access to finance variable suggests that even after correcting the problem of endogeneity, access to finance plays an important role in firms decision of whether to enter the export market. The results were consistent with all the instruments used. Having confirmed the positive role played by access to finance in firms export decision, we next examined the effect of financial development on firm exportability using instruments for the access to finance variable. As we have used an IV probit model, it is necessary to test the properties of the instrument statistically. Various diagnostic tests were used to identify the statistical properties of the instruments. Test statistics suggest that overdraft or line of credit and collateral requirement as instruments passed the test and, thus, the instruments are neither weak nor overidentified (see Table 6). The results of the instrument variable biprobit model are presented in Table 6. As expected, the coefficients

18 ASIAN ECONOMIC JOURNAL 32 Table 4 Access to finance, financial development and exportability (marginal effect) Dependent variable-exporter (1) (2) (3) (4) (5) Log of age *** *** *** *** *** Log of capacity utilization *** ** * ** Size (medium) *** *** *** *** *** Size (large) *** *** *** *** *** Location (fewer than 1m people in city) ** *** *** *** Foreign firm *** *** *** *** *** Access to finance-formal (AF) *** * *** *** *** Market Capitalization to GDP *** Credit to GDP *** Bank Branch per thousand population (BBptp) *** *** *** Location * BBptp ** Location * AF ** of variables are consistent and are in line with expectations. The access to finance variable is also positive and statistically meaningful. With regard to the interaction of location and access to finance, the results again confirm the importance of financial development and access to finance in mitigating the disadvantage of location on firms decision to export. Overall, the results confirm the positive effect of firm s age, size, foreign ownership and productivity on their decision to export. Furthermore, access to finance and financial development are found to be instrumental in increasing firms likelihood of entering the export market. Results are also robust to different measures of firm export and access to finance. Given the significance of access to finance and financial development, the study supports policy intervention to unleash the financial sector and help the banking sector to reach and operate in remote locations. VII. Conclusion The present study has explored two main questions: first, how does access to finance affect firm export decisions and, second, whether financial development boosts firms exportability or not. In addition, this study has explored the effect of access to finance and financial development on the decision to export by firms with locational disadvantage. Firms in the Asia Pacific region are appropriate for testing the hypothesis as these countries are following intensive export-led growth models. In addition, firms in this region are more financially constrained compared to those in developed countries and their financial sectors tend to be underdeveloped. The present study has used WBES and World Development Indicators data for testing the hypothesis. Preliminary findings suggest that exporters are old, large, owned by foreign firms and tend to be more productive. However, in terms of location, there is not

19 ACCESS TO FINANCE, FINANCIAL DEVELOPMENT AND FIRM EXPORTABILITY 33 Table 5 Regression results of IV estimation: Access to finance and exportability (1) (2) (3) Access to finance-formal 0.247*** 0.168* 0.256*** (0.045) (0.089) (0.028) Log of age 0.109*** 0.108*** 0.109*** (0.012) (0.012) (0.012) Log of capacity utilization 0.115*** 0.113*** 0.115*** (0.028) (0.028) (0.028) Size (medium) 0.605*** 0.610*** 0.604*** (0.027) (0.028) (0.027) Size (large) 1.331*** 1.343*** 1.330*** (0.029) (0.031) (0.028) Location (fewer than 0.096*** 0.089*** 0.096*** 1m people in city) (0.022) (0.022) (0.021) Foreign firm 0.883*** 0.880*** 0.883*** (0.034) (0.034) (0.034) Constant 5.862*** 7.772*** (0.369) (0.242) (6.896) Instrumental variable Overdraft or line of credit ISO certificate Collateral requirement First-stage coefficient 0.864*** 0.067* 0.282*** (0.022) (0.028) (0.031) Identification test Under identification test: Kleibergen Paap rk *** *** *** Lagrange multiplier statistic Weak identification test: Cragg Donald Wald *** $ *** F-statistic Kleibergen Paap rk *** $ *** Wald F-statistic Number of observations Wald χ Notes: Reported results are second-stage regression analysis with the exporter as the dependent variable. Robust standard errors are in parentheses. Significance level: ***p < 0.01, **p < 0.05, *p < 0.1. $ close to p < All specification includes industry and year fixed effect. much difference between exporting and non-exporting firms. Exporting firms are found to have better access to finance and to come from more financially developed countries. In the next step, we used probit and IV-probit estimation techniques to empirically test the hypothesis. Overdraft or line of credit, ISO certificate and collateral requirement are used as instruments for access to finance. The estimation results suggest that firms age, size, productivity and foreign ownership are important internal attributes affecting firms decision to export. The results regarding location indicate a shift from the preliminary investigation