A. Introduction

1. The small- and medium-sized enterprise (SME) sector in Hungary has yet to recover from the global financial crisis. This sector is relatively large, accounting for 71 percent of the business sector’s employment.² However, employment and real value added are estimated to remain well below their pre-crisis level (about 9 percent and 7 percent, respectively), suggesting that the SMEs’ weak activity continues to weigh on economic recovery (chart). While the poor SMEs’ performance is in part a reflection of the ongoing corporate sector’s deleveraging, and the tight credit conditions, which limited the access to finance for more risky SMEs, it also resulted from the sharp contraction of domestic demand—the SMEs’ main market (Figure 1A in the Appendix).

Source: European Commission and IMF staff calculation.* Adjusted to prices and exchange rate effects.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Source: European Commission and IMF staff calculation.* Adjusted to prices and exchange rate effects.

• Download Figure
• Download figure as PowerPoint slide

Source: European Commission and IMF staff calculation.* Adjusted to prices and exchange rate effects.

• Download Figure
• Download figure as PowerPoint slide

2. Reducing the SMEs’ heavy reliance on the domestic market can foster higher growth while increasing their resilience to shocks. The benefits of firms’ internationalization are well recognized. By broadening the customer base through entering new markets, firms are able to expand their production, especially if such expansion is constrained in the domestic market due to saturation or well-entrenched competitors. Furthermore, the diversification of revenue sources can also reduce the firm’s vulnerability to shocks, and as such support more stable growth over time, while some studies, including Robson and Bennet (2000), and Beccetti and Trovato (2002), suggest that firms are likely to become more productive and efficient by entering foreign markets (“learning-by-exporting” hypothesis).³ Relatedly, studies, including the OECD (2008), argue that firms that increase their internationalization through increased integration into global supply chains are likely to benefit from access to new technologies, innovations, and advanced management techniques, and thus improve the production process and increase productivity.

3. Against this background, the purpose of this paper is twofold. First, it aims at assessing the link between the firm’s internationalization (as proxied by the export-to-operating revenue ratio) and firm’s growth in Hungary by applying a micro-level analysis. The second objective is to examine whether firm-level factors, including factor productivity, profitability, leverage, and capital structure play a role in the firm’s degree of internationalization. Identifying these factors could help in shaping up policy priorities to facilitate higher SMEs’ internationalization and growth.

4. The analysis finds a positive link between internationalization and growth among Hungarian SMEs. While the results should be treated with caution given the limited sample, the noisy period, and the narrow measure of internationalization, they show a positive and significant link between the change in the export-to-operating revenue ratio and SMEs’ growth. Moreover, the analysis suggests that larger SMEs are more likely to access foreign markets, while factors such as higher fixed assets ratio, higher labor productivity, and more favorable liquidity position contributed to firms’ internationalization. At moderate levels, higher leverage was also found to contribute to higher export-to-operating revenues ratio, though this effect diminishes at excessive leverage ratio.

5. The rest of the paper is structured as follows: Section B presents some stylized facts on the degree of internationalization and innovation among Hungarian SMEs compared to EU peers. Section C employs a dynamic panel model to assess the link between the SMEs’ internationalization and growth, and examines whether the link varies by the firm’s size. Section D identifies the firm-level determinants for the SMEs’ internationalization; and section E concludes.

B. Some Stylized Facts

6. Cross country comparison suggests that the degree of internationalization among Hungarian SMEs falls behind that of its European peers (Figure 2A in the Appendix). The European Commission’s 2014 Small Business Act Fact Sheet, indicates that, apart from the cost required to export/import, Hungary scores well below the EU’s average in all of the internationalization criteria (chart). Specifically, the share of SMEs trading outside the EU is significantly below the EU average, and the bureaucracy—the time and number of documents required to export/import—is poorer than the EU average. These indicators are consistent with the European Commission’s survey on internationalization of European SMEs,⁴ which shows that the level of internationalization of Hungarian SMEs is relatively low, and in some indicators–particularly SMEs with direct exports and SMEs with cooperation with enterprises abroad–Hungarian SMEs fall behind their regional peers (Table 1A in Appendix).⁵ The survey also indicates that Hungarian SMEs face some critical impediments to expanding their activity beyond the local market. In this regard, the percentage of Hungarian SMEs that reported the “lack of capital or finance” and the “cost or difficult paper work associated with transport” is high and exceeded that in regional peers (chart).

Internationalization: Variation from the EU average (measured in standard deviations, EU average=0)

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

Note: Data bars pointing right show better performance than the EU average and data bars pointing left show weaker performance.Source: European Commission, DG Enterprise and Industry

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Internationalization: Variation from the EU average (measured in standard deviations, EU average=0)

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

Note: Data bars pointing right show better performance than the EU average and data bars pointing left show weaker performance.Source: European Commission, DG Enterprise and Industry

• Download Figure
• Download figure as PowerPoint slide

Internationalization: Variation from the EU average (measured in standard deviations, EU average=0)

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

Note: Data bars pointing right show better performance than the EU average and data bars pointing left show weaker performance.Source: European Commission, DG Enterprise and Industry

• Download Figure
• Download figure as PowerPoint slide

Barriers Related to Business Environment ^/1

(Percentage of SMEs)

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

^1/ The culumns indicate the percentage of firms that reported these barriers as important.Source: Internationalisation of European SMEs, European Commission, 2010.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Barriers Related to Business Environment ^/1

(Percentage of SMEs)

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

^1/ The culumns indicate the percentage of firms that reported these barriers as important.Source: Internationalisation of European SMEs, European Commission, 2010.

• Download Figure
• Download figure as PowerPoint slide

Barriers Related to Business Environment ^/1

(Percentage of SMEs)

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

^1/ The culumns indicate the percentage of firms that reported these barriers as important.Source: Internationalisation of European SMEs, European Commission, 2010.

• Download Figure
• Download figure as PowerPoint slide

7. There appears to be a strong link between internationalization and innovation. Although the causality between innovation and internationalization is likely to be bi-directional, it was found that internationally active firms tend to introduce product/service and process innovations more often than non-internationally active firms (European Commission, 2010). Some empirical studies, including Golovko and Valentini (2011), found that exports and innovation positively reinforce each other, as the positive effect of innovation on firms’ growth rate is higher for firms that also engage in exports, and vice versa. This may suggest that the low level of Hungarian SMEs’ internationalization is partly driven by their low level of innovation activities. The proportion of innovative enterprises (i.e., those that introduced a new product/service, process, new marketing or new management methods) during 2008–10 stood at about 31 percent, ranking fifth from the bottom in the EU (chart). Furthermore, the composition of enterprises shows that innovation activities are inversely related to the firm’s size: while more than two thirds of the large firms were engaged in innovation activities in 2008–10, less than half of medium-sized firms and about a quarter of the small firms innovated during this period.

Proportion of innovative enterprises, 2008–10

(% of all enterprises)

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

1/ Excluding Greece.Source: Eurostat.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Proportion of innovative enterprises, 2008–10

(% of all enterprises)

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

1/ Excluding Greece.Source: Eurostat.

• Download Figure
• Download figure as PowerPoint slide

Proportion of innovative enterprises, 2008–10

(% of all enterprises)

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

1/ Excluding Greece.Source: Eurostat.

• Download Figure
• Download figure as PowerPoint slide

Hungary. Proportion of Innovative Enterprises by Size Class, 2008–2010

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

Source: Hungarian Central Statistical Office.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Hungary. Proportion of Innovative Enterprises by Size Class, 2008–2010

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

Source: Hungarian Central Statistical Office.

• Download Figure
• Download figure as PowerPoint slide

Hungary. Proportion of Innovative Enterprises by Size Class, 2008–2010

Citation: IMF Staff Country Reports 2015, 093; 10.5089/9781484314555.002.A002

Source: Hungarian Central Statistical Office.

• Download Figure
• Download figure as PowerPoint slide

C. The Link between Internationalization and Growth

Data

8. The analysis uses the ORBIS database of Bureau Van Dijck, which contains world-wide information on private and public firms. We include all Hungarian SMEs and large enterprises that have a complete record of the variables of interest, and that were “active” during the sample period (thus excluding firms that went bankrupt or were established during the period). Overall, the sample includes annual data for 2006–13 for 478 firms, 300 of them are SMEs that are privately-owned and unlisted in the stock market.⁶

9. Firms’ characteristics are presented in Table 1. Although the comparison may not be representative of all Hungarian firms, it clearly shows that the exports-to-operating revenue ratio increases with the firm’s size: the average export-to revenue ratio for the median large firm is ten-fold the ratio of the median small firm and about threefold the ratio of the median medium firm. The firms’ cash flow-to-operating ratio also increases with the firm size, thus indicating a higher level of profitability. Interestingly, while the comparison indicates that large and medium firms recorded higher growth than small firms (cumulative, 2006–13), there is no clear link between the firms’ size and labor productivity growth.

^{Table 1}.

Selected Indicators by Firm Size, 2006–2013¹

¹Figures refer to the median firm in each category.

^{Table 1}.

Selected Indicators by Firm Size, 2006–2013¹

	Small firms (10–49)	Medium Firms (50–249)	Large Firms (>249)
Export-to-revenue ratio, percent	7.0	23.3	70.9
Labor productivity growth, cumulative 2006–2013	−7.62	−14.40	32.67
Assets (million EUR)	4	7	12
Cash flow-to- operating revenues, percent	5.0	6.0	9.0
Real Growth of operating revenues, Cumulative 2006–2013	−17.26	−0.05	0.07
Number of employees	32	96	363
Number of firms	88	212	178

¹Figures refer to the median firm in each category.

View Table

^{Table 1}.

Selected Indicators by Firm Size, 2006–2013¹

	Small firms (10–49)	Medium Firms (50–249)	Large Firms (>249)
Export-to-revenue ratio, percent	7.0	23.3	70.9
Labor productivity growth, cumulative 2006–2013	−7.62	−14.40	32.67
Assets (million EUR)	4	7	12
Cash flow-to- operating revenues, percent	5.0	6.0	9.0
Real Growth of operating revenues, Cumulative 2006–2013	−17.26	−0.05	0.07
Number of employees	32	96	363
Number of firms	88	212	178

¹Figures refer to the median firm in each category.

10. Table 2, which covers SMEs, highlights the differences between SMEs with high export-to-operating revenue ratio and the rest. The former are mostly from the manufacturing sector, and they are larger on average, both in terms of total assets and employees. Moreover, SMEs with a high export-to-revenue ratio were more profitable and recorded higher cumulative growth in their operating revenues compared to firms with low export share. The latter is also reflected in their higher labor productivity growth. The liquidity ratio shows that high-export firms maintained lower level of liquidity, suggesting that machinery, buildings and other tangible assets account for a larger part of their assets. Indeed their share of fixed assets to total assets is well above that of low-export firms. The gearing ratio (the ratio between non-current liabilities and shareholder funds) is significantly higher in firms with higher export ratio, thus suggesting that these firms have on average higher access to financing.

^{Table 2}.

SMEs: Selected Indicators by Export-to-Revenue Ratio^1,2

¹Firms with high (low) export ratio are those that their export ratio is above (below) the sample’s median.

²Figures refer to the median firm in each category.

^{Table 2}.

SMEs: Selected Indicators by Export-to-Revenue Ratio^1,2

		High Export ratio	Low Export ratio
Real Growth of operating revenues, Cumulative 2006–13		2.62	−12.03
Labor productivity growth, cumulative 2006–13		−1.5	−26.0
Cash flow (percent of operating revenue)		6.6	4.89
Gearing ratio		41.2	23.8
Assets (in mil EUR)		5.0	6.0
Fixed tangible assets-to-total assets		40.67	28.43
Liquidity ratio		0.84	1.02
# of employees		93	59
# of firms		156	144
	of which manufacturing	110	6
	Small firms	31	57

¹Firms with high (low) export ratio are those that their export ratio is above (below) the sample’s median.

²Figures refer to the median firm in each category.

View Table

^{Table 2}.

SMEs: Selected Indicators by Export-to-Revenue Ratio^1,2

		High Export ratio	Low Export ratio
Real Growth of operating revenues, Cumulative 2006–13		2.62	−12.03
Labor productivity growth, cumulative 2006–13		−1.5	−26.0
Cash flow (percent of operating revenue)		6.6	4.89
Gearing ratio		41.2	23.8
Assets (in mil EUR)		5.0	6.0
Fixed tangible assets-to-total assets		40.67	28.43
Liquidity ratio		0.84	1.02
# of employees		93	59
# of firms		156	144
	of which manufacturing	110	6
	Small firms	31	57

¹Firms with high (low) export ratio are those that their export ratio is above (below) the sample’s median.

²Figures refer to the median firm in each category.

Econometric Model

11. In the first stage, we examine whether higher internationalization matters for firms’ growth. We apply a dynamic panel analysis with unobserved panel effects (Arellano and Bond, 1991), in which the firm-specific variables are treated as endogenous and instrumented with their lags.⁷ The sample includes both large firms and SMEs for which data is available over the period 2006–13. The model specification can be represented as follows:

$\begin{array}{cccc}\left(1\right)& {Growth}_{it}& =& {\begin{array}{c}\alpha \end{array}}_0+{\alpha }_1\left({Growth}_{it-1}\right)+{\beta }_1\left({InAssests}_{it}\right)+{\beta }_2\left({dExports}_{it}\right)\\ & & & +{\beta }_3({dExport}_{it}*SME)+{\beta }_4\left({LnAge}_{it}\right)+{\beta }_5\left({Time}_t\right)+{ϵ}_{it}\end{array}$

12. Two alternative indicators are used to measure the firms’ growth (Growth): the difference in the logarithms of the firm’s operational revenues, dlnOprev (in HUF, deflated by the GDP prices) and the difference in the logarithms of the firm’s employment, dlnEmpl. The change in the firm’s internationalization is measured by the change in export-to-revenue ratio, dExport, and it is also interacted with SME— a dummy variable that obtains a value of one for SMEs and zero otherwise—to assess whether the impact of internationalization differ by the firm’s size. In addition, we control for the “traditional” variables that are used in the literature to explain the firm’s growth: (1) the firm’s size, measured by the logarithms of the firm’s total assets, lnAssets (in HUF, deflated by the GDP prices); and (2) the logarithms of the firm’s age lnAge. The specification also includes a time variable, Time, to control for macroeconomic effects.

13. Table 3 presents the correlation matrix for the firm-specific variables. The table shows, as expected, a positive and significant correlation between the two measures for the firm’s growth. Moreover, it indicates a positive correlation between the contemporaneous change in export-to-operating revenues and the two measures for growth and between the firms’ size (lnAssets) and the firms’ growth. The latter is negatively correlated with the firm’s age, as also suggested by some previous studies.

^{Table 3}.

Correlation Matrix, SMEs and Large Firms

* indicates significance level of 5 percent.

^{Table 3}.

Correlation Matrix, SMEs and Large Firms

	dlnEmpl	dlnOprev	lnAge	lnAssets	dExport
dlnEmpl	1
dlnOprev	0.277*	1
lnAge	−0.084*	−0.017	1
lnAssets	0.034	0.087*	0.021*	1
dExport	0.028	0.038*	0.001*	0.015*	1

* indicates significance level of 5 percent.

View Table

^{Table 3}.

Correlation Matrix, SMEs and Large Firms

	dlnEmpl	dlnOprev	lnAge	lnAssets	dExport
dlnEmpl	1
dlnOprev	0.277*	1
lnAge	−0.084*	−0.017	1
lnAssets	0.034	0.087*	0.021*	1
dExport	0.028	0.038*	0.001*	0.015*	1

* indicates significance level of 5 percent.

Results

14. The estimation results are presented in Table 4. The Hansen test indicates that the instruments used are uncorrelated with the residual, and the Arellano-Bond tests for autocorrelation reject the hypothesis that the errors are not auto-correlated in the first order (AR(1)), but cannot reject this hypothesis for the second order (AR(2)).⁸ In addition, the results show that an increase in export-to-revenue ratio contributes to higher SMEs growth, suggesting that a one percentage point increase in export-to operating revenue ratio contributes to about 1½ percent to the firm’s operating revenue growth (in real terms), and about ¼ percent growth in the firm’s employment.⁹ Interestingly, the effect of higher export-to-operating revenue ratio is not significant for large firms, perhaps suggesting that the benefits of higher export ratios are significantly stronger at the initial stages and low levels of internationalization.

^{Table 4}.

Estimation Results, Internationalization and Growth, Full Sample¹

¹Significance level: * significant at 1 percent, ** significant at 5 percent, *** significant at 10 percent.

^{Table 4}.

Estimation Results, Internationalization and Growth, Full Sample¹

Dependent variable	dlnOprev			dlnEmp
dlnOprev (lagged)	−0.056**	−0.069**	−0.069**
dlnEmpl (lagged)				0.024	0.022	0.019
lnAssets	0.380	0.376*	0.376*	0.153*	0.143*	0.139*
dExport	0.001	−0.003	−0.000	0.001	−0.000	−0.000
dExport*SME		0.016**	0.016**		0.003***	0.003***
lnAge			0.008			−0.069
Time	Yes	Yes	Yes	Yes	Yes	Yes
# of observations	2,240	2,240	2,240	2,267	2,267	2,267
Arellano-Bond test (1) p-value	0.000	0.000	0.000	0.000	0.000	0.000
Arellano-Bond test (2) p-value	0.875	0.225	0.206	0.539	0.687	0.659
Sargan test	0.028	0.115	0.115	0.302	0.280	0.267

¹Significance level: * significant at 1 percent, ** significant at 5 percent, *** significant at 10 percent.

View Table

^{Table 4}.

Estimation Results, Internationalization and Growth, Full Sample¹

Dependent variable	dlnOprev			dlnEmp
dlnOprev (lagged)	−0.056**	−0.069**	−0.069**
dlnEmpl (lagged)				0.024	0.022	0.019
lnAssets	0.380	0.376*	0.376*	0.153*	0.143*	0.139*
dExport	0.001	−0.003	−0.000	0.001	−0.000	−0.000
dExport*SME		0.016**	0.016**		0.003***	0.003***
lnAge			0.008			−0.069
Time	Yes	Yes	Yes	Yes	Yes	Yes
# of observations	2,240	2,240	2,240	2,267	2,267	2,267
Arellano-Bond test (1) p-value	0.000	0.000	0.000	0.000	0.000	0.000
Arellano-Bond test (2) p-value	0.875	0.225	0.206	0.539	0.687	0.659
Sargan test	0.028	0.115	0.115	0.302	0.280	0.267

¹Significance level: * significant at 1 percent, ** significant at 5 percent, *** significant at 10 percent.

15. In line with previous studies (e.g., Wiklund and Shepherd, 2005; Mateev and Anastasov, 2010) the results indicate that larger firms registered on average higher growth. This may reflect the notion that large firms are more likely to exploit economies of scale, have higher negotiating power over clients and suppliers (Serrasqueiro and Macas Nunes, 2008; Singh and Whittington, 1975), and that they generally face lower financing constraints. The impact of the firm’s age was not found to have a significant effect, in contrast to the findings of some other studies such as Navertti et al. (2014).

D. The Determinants of Firms’ Internationalization

17. This section examines the extent to which the change in the firm’s export-to-operating ratio can be explained by firm- specific factors such as profitability, liquidity, leverage, size, labor productivity, and the share of fixed asset. As in the previous section, we apply a dynamic panel approach, which can be represented as follows:

18. The change in the firm’s export-to-operating revenues ratio, dExport, serves as a proxy for the change in the firm’s degree of internationalization. This ratio is explained by its one-period lag and by other firm specific factors, including the firm’s size, measured by the logarithms of the number of employees, lnEmpl; the logarithms of the firm’s age, lnAge; the change in labor productivity, dLprod, which measures annual efficiency gains at the firm level, and fixed tangible assets to total assets, Fassets. The latter may capture the firm’s capital intensity, and also the availability of collateral for bank financing, which affect the risk classification of the firm.¹¹ Furthermore, the specification includes several financial variables, which include the change in the firm’s financial leverage, dGearing (measured by the ratio of non-current liabilities and loans-to-shareholders’ funds); the change in the firm’s liquidity, dLiquidty (measured by the change in the ratio of current assets minus inventories to current liabilities); and the change in the firm’s profitability, dCflow (profit plus depreciation–to–operating revenue ratio). The financial variables are interacted with High—a dummy variable, which obtains a value of one if the firm’s ratios are above the average in the sample, and zero otherwise. Lastly, the estimation allows for a time variable, Time, to capture macro-economic effects. All of the firm’s specific variables, apart from the firm’s age, are treated as endogenous variables, thus instrumented with their lags. This methodology controls for unobserved individual effects, thus variables that are time invariant, such as the sector in which the firms operate, were dropped. Summary statistics are presented in Tables 5.

Table 5.

Summary Statistics, Firm Level Variables N = 2400, 2006–13

Table 5.

Summary Statistics, Firm Level Variables N = 2400, 2006–13

	Mean	Standard Dev.	Min	Max
Export ratio	29.69	30.44	0.01	99.95
# of Employees	86.68	53.06	10.00	249.00
Firm Age	15.09	4.54	0.00	31.00
Labor productivity	41.69	42.67	2.50	486.67
Gearing ratio (%)	64.50	100.58	0.00	968.81
Liquidity ratio	1.21	1.14	0.06	15.42
Cash flow (%)	6.85	7.18	−79.91	43.17
Fixed tangible assets to total assets (%)	34.78	23.47	0.00	100

View Table

Table 5.

Summary Statistics, Firm Level Variables N = 2400, 2006–13

	Mean	Standard Dev.	Min	Max
Export ratio	29.69	30.44	0.01	99.95
# of Employees	86.68	53.06	10.00	249.00
Firm Age	15.09	4.54	0.00	31.00
Labor productivity	41.69	42.67	2.50	486.67
Gearing ratio (%)	64.50	100.58	0.00	968.81
Liquidity ratio	1.21	1.14	0.06	15.42
Cash flow (%)	6.85	7.18	−79.91	43.17
Fixed tangible assets to total assets (%)	34.78	23.47	0.00	100

19. Table 6 presents the correlation matrix. It shows that SMEs with higher export ratio are, on average, more profitable, have more employees, and maintain a higher share of fixed assets. Moreover, SMEs with higher leverage (gearing ratio) maintain lower levels liquidity and are less profitable, perhaps reflecting the high debt service payments. Interestingly, the correlation of export ratio with labor productivity is negative.

^{Table 6}.

Correlation Matrix

* indicates significance level of 5 percent.

^{Table 6}.

Correlation Matrix

	Export ratio	# of Employees	Firm Age	Labor productivity	Gearing ratio (%)	Liquidity ratio	Cash flow (%)	Fixed assets ratio (%)
Export ratio	1
# of Employees	0.301*	1
Firm Age	0.006	0.044*	1
Labor productivity	−0.149*	−0.412*	−0.081*	1
Gearing ratio (%)	0.040*	−0.033	−0.069*	0.070*	1
Liquidity ratio	−0.025	−0.067*	0.053*	−0.049*	−0.252*	1
Cash flow (%)	0.076*	0.042*	−0.076*	−0.158*	−0.194*	0.222*	1
Fixed assets (%)	0.214*	0.231*	−0.005	−0.307*	0.164*	−0.180*	0.205*	1

* indicates significance level of 5 percent.

View Table

^{Table 6}.

Correlation Matrix

	Export ratio	# of Employees	Firm Age	Labor productivity	Gearing ratio (%)	Liquidity ratio	Cash flow (%)	Fixed assets ratio (%)
Export ratio	1
# of Employees	0.301*	1
Firm Age	0.006	0.044*	1
Labor productivity	−0.149*	−0.412*	−0.081*	1
Gearing ratio (%)	0.040*	−0.033	−0.069*	0.070*	1
Liquidity ratio	−0.025	−0.067*	0.053*	−0.049*	−0.252*	1
Cash flow (%)	0.076*	0.042*	−0.076*	−0.158*	−0.194*	0.222*	1
Fixed assets (%)	0.214*	0.231*	−0.005	−0.307*	0.164*	−0.180*	0.205*	1

* indicates significance level of 5 percent.

20. The estimation results are presented in Table 7. They show that, while the firm’s profitability and age do not affect the firm’s export ratio, other factors have a significant effect on the firm’s degree of internationalization. In particular, the results suggest that:

• Size. Large firms, with a large number of employees, have, on average, higher export share. This finding is broadly in line with previous studies that suggest that the penetration into foreign markets entails large resources (personnel, financial, research and development, and marketing) that are often beyond the means of small firms (Bonaccorsi, 1992). OECD (2009) indeed highlighted that, beyond the financial barriers (which are also a function of size), lack of managerial skills and knowledge about new markets, which are likely to be more prevalent among SMEs, remain among the top obstacles for internationalization. The positive link between firm’s size and export share can be supported by a possible economies-of-scale effect, and by the notion that smaller firms tend to expand their domestic markets first due to lower uncertainty and risk (Cavusgil and Noar, 1987).

• Labor productivity. SMEs that became more productive (i.e., increased labor productivity) registered higher export-to-revenue share. This finding indicates that more productive firms are more likely to enter the export market, suggesting that competitiveness plays an important role in internationalization (e.g., Golovko and Valentini, 2011).

• Access to finance. Leverage was found to have a positive effect as long as it remains at moderate levels. The positive and significant coefficient of dGearing atio implies that financing constraints not only affect the firms’ growth (Bechetti and Trovato, 2002) but they also limit the firm’s expansion into foreign markets. That said, the estimation shows that “excessive” leveraging (i.e., beyond the sample’s average (65 percent) eliminates this positive effect. This finding is consistent with that of Demijan (2014), who examined the link between financial soundness and performance among firms in Slovenia and found that excessive indebtedness undermine the firm’s exports.

• Liquidity. Liquidity-constrained firms are less likely to increase their export-to-operating revenue ratio, perhaps due to the negative effect on firm’s investment, and ability to direct resources to productive uses. Liquidity-constrained firms are also less likely overcome the cost associated with accessing and operating in foreign markets.¹²

• Fixed assets. The impact of the fixed assets-to-total assets is positive and significant. As discussed above, this variable may reflect several channels, including the availability of collateral for bank financing, which affect the firm’s riskiness and cost of capital, as well as the firm’s degree of capital intensity, technology, and sophistication.

^{Table 7}.

Estimation Results, Determinants of SMEs’ Internationalization¹

¹Significance level: * significant at 1 percent, ** significant at 5 percent, *** significant at 10 percent.

^{Table 7}.

Estimation Results, Determinants of SMEs’ Internationalization¹

	dExport	dExport	dExport	dExport	dExport	dExport
dExport (lagged)	−0.081	−0.092***	−0.086**	−0.088***	−0.073	−0.084***
lnEmpl	3.802*	3.370**	2.660**	1.825***	1.375	0.860
dLprod	0.026	0.0329	0.027***	0.032**	0.033**	0.032***
lnAge		−0.000	0.133	0.185	−0.082	0.003
dGearing			0.027***	0.032**	0.029**	0.026**
dGearing*High			−0.029**	−0.034**	−0.030**	−0.028**
dLiquidity				1.341***	1.212***	1.329***
dLiquidity*High				−1.119	−0.9300	−1.095
dCflow					−0.014	−0.005
dCflow*High					−0.024	−0.064
Fassets						0.035***
Time	Yes	Yes	Yes	Yes	Yes	Yes
Constant	−15.505**	−14.016**	−11.385	−7.547	−5.362	−4.586
# of observations	1,799	1,799	1,799	1,799	1,799	1,797
# of firms	300	300	300	300	300	300
A-B test (1) p-value	0.000	0.000	0.000	0.000	0.000	0.000
A-B test (2) p-value	0.882	0.794	0.795	0.750	0.888	0.758
Hansen test	0.136	0.229	0.599	0.280	0.295	0.335

¹Significance level: * significant at 1 percent, ** significant at 5 percent, *** significant at 10 percent.

View Table

^{Table 7}.

Estimation Results, Determinants of SMEs’ Internationalization¹

	dExport	dExport	dExport	dExport	dExport	dExport
dExport (lagged)	−0.081	−0.092***	−0.086**	−0.088***	−0.073	−0.084***
lnEmpl	3.802*	3.370**	2.660**	1.825***	1.375	0.860
dLprod	0.026	0.0329	0.027***	0.032**	0.033**	0.032***
lnAge		−0.000	0.133	0.185	−0.082	0.003
dGearing			0.027***	0.032**	0.029**	0.026**
dGearing*High			−0.029**	−0.034**	−0.030**	−0.028**
dLiquidity				1.341***	1.212***	1.329***
dLiquidity*High				−1.119	−0.9300	−1.095
dCflow					−0.014	−0.005
dCflow*High					−0.024	−0.064
Fassets						0.035***
Time	Yes	Yes	Yes	Yes	Yes	Yes
Constant	−15.505**	−14.016**	−11.385	−7.547	−5.362	−4.586
# of observations	1,799	1,799	1,799	1,799	1,799	1,797
# of firms	300	300	300	300	300	300
A-B test (1) p-value	0.000	0.000	0.000	0.000	0.000	0.000
A-B test (2) p-value	0.882	0.794	0.795	0.750	0.888	0.758
Hansen test	0.136	0.229	0.599	0.280	0.295	0.335

¹Significance level: * significant at 1 percent, ** significant at 5 percent, *** significant at 10 percent.

E. Conclusions and Some Policy Implications

21. The SME sector remains an important driver for Hungary’s economic growth and employment creation. However, its heavy reliance on the domestic economic activity has limited its opportunities to expand, develop, and reap the benefits of globalization. Moreover, the collapse of domestic demand in 2009 and the weak domestic economic activity in subsequent years adversely affected the SMEs’ performance, and their employment and value added have yet to return to their pre-crisis levels.

22. This paper investigated the effects and the firm-level determinants of internationalization among Hungarian SMEs. While the results need to be treated with caution, given the limited sample size, and the turbulent sample period, which largely covers the effects of the global financial crisis, we find strong evidence that an increase in the SMEs’ export-to-operating revenue ratio has a nontrivial effect on SMEs’ growth, both in terms of revenues and number of employees. This finding perhaps corroborates the “learning–by–exporting” hypothesis, which suggests that greater degree of internationalization can generate positive spillovers through technology and knowledge transfer, and ultimately lead to productivity and competitiveness gains.

23. The analysis also identified some of the firm-level factors that contribute to higher internationalization. Specifically, larger firms with a high share of fixed tangible assets tended to increase more their level of internationalization. Moreover, higher labor productivity, more favorable liquidity position, and higher leverage are also associated with higher degree of internationalization, though the positive effect of the leverage ratio diminishes when its level is “excessive”.

24. These results suggest that there is scope to take policy actions to alleviate the barriers for internationalization to support SMEs’ growth. Developing a stronger infrastructure to improve the capacity of SMEs to trade internationally, including by simplifying procedures for exports—an area that Hungary is lagging behind its peers, is critical. Additionally, there is scope to expand public sector’s support to help SMEs penetrate foreign markets, including by reducing skill mismatch; and promote innovation activities so as to enhance factors’ productivity and external competitiveness. Fostering greater cooperation of SMEs with large multi-national companies can also facilitate positive spillovers through knowledge and technology transfer and better infrastructure for internationalization. As the results also show that greater access to financing is an important contributor to SMEs’ internationalization, it is important to ensure that financial intermediation is repaired, including by cleaning up banks’ portfolio and reducing the heavy tax burden. Finally, developing an infrastructure for non-bank financing may also alleviate funding constraints.