Financing obstacles and growth: an analysis for euro area non-financial firms

Abstract

This paper investigates the determinants of financing obstacles (FOs) and their impact on firm growth. For this purpose, we rely on both balance sheet data and survey data for a sample of non-financial firms in the euro area. The latter allows us to devise a direct measure of the firms’ probability of facing FOs. First, our results indicate that FOs are linked to characteristics such as the age of the firm, its size, its sales level or the sector in which it operates. Second, we find that, though based on few variables, our measure of FOs appears to be relevant in explaining firm growth in four out of the five countries considered; likewise, growth is found to be positively linked to cash flow.

Keywords:

• financing constraints
• firm growth
• panel data

JEL Classification:

• E22
• G32
• L25

Notes

1. SMEs represent 99.8% of all EU-27 enterprises in the non-financial business economy in 2006, employing about 67% of the workforce and generating more than half (57.7%) of its value added (Eurostat).

2. Sutton (1997) provides a review of the theoretical and empirical literature. Among others, see also Mansfield (1962), Hall (1987), Evans (1987a, 1987b) and Goddard, McKillop, and Wilson (2002) for the USA; Hart and Oulton (1996), Kumar (1985), and Dunne and Hughes (1994) for the UK; Goddard, Wilson, and Blandon (2002) for Japan; Mata (1994) and Oliveira and Fortunato (2006a) for Portugal; Wagner (1992), and Harhoff, Stahl, and Woywode (1998) for Germany; Solinas (1995), and Lotti, Santarelli, and Vivarelli (2003) for Italy. Tschoegl (1983), instead, performs a multi-national study.

3. Survey data are also used, for instance, by Becchetti and Trovato (2002). The authors test Gibrat's Law by including a dummy variable which takes a value of one if the firm reports that it is credit constrained. Their analysis rejects the LPE and shows that variables measuring the availability of external finance (subsidies, leverage, and financing constraints) affect firm growth in a significant way.

4. For additional details on the survey, see Appendix in Coluzzi, Ferrando, and Martinez-Carrascal (2009). Data and documentation are available on the website of the World Bank. The WBES has been used fully or partially in a number of papers for different purposes. Related to our paper, see Beck et al. (2006), Beck, Demirgüç-Kunt, and Maksimovic (2004, 2005); and Beck, Demirgüç-Kunt, and Levine (2006).

5. In addition to simple regression analysis, we specified the model either as a probit or as a logit. We controlled for country effects, for sector effects and for country and sector effects.

6. Comparing results at the country level is subject to caveat (see the user agreement for WBES, which states that ‘Given inherent error margins associated with any single survey results, it is inappropriate to use the results from this survey for precise country rankings in any particular dimension of the investment climate or governance’).

7. Later on, we will measure the growth rate of firms in terms of changes in total assets. Unfortunately, the WBES does not include this information. For the same period and countries, the growth of sales in AMADEUS is about 15%. It is not possible to compare the growth of employees, however, as a large number of firms in the AMADEUS sample does not report this information.

8. The reclassification of the balance sheets appears reliable, since no attempt is made to reconstruct items that are missing from the original balance sheets or are difficult to reconstruct. However, the limitations of the dataset are well known as accuracy and coverage of the data depend on how demanding the accounting standards of a country are. For instance, in France, there is a compulsory format for the preparation of company accounts, while in Germany, Austria and the Netherlands just a recommended or suggested format. Therefore, the sample is biased towards countries with more demanding accounting standards. The use of the dataset is nevertheless becoming widespread as it covers an extensive number of private firms with different sizes across countries. For more information on the dataset see http://bvdinfo.com/Products/Company-Information/International/AMADEUS.aspx).

9. In terms of employment, the size classes are defined as micro with fewer than 10 employees, small with fewer than 50, medium with fewer than 250 and large with more than 250. The compounded size classification used by the European Commission allows us to retain a large number of observations from the dataset, in particular for SMEs that do not often report the number of employees in AMADEUS. When this information is available, we checked the classification based only on total assets and turnover, and in most cases, the matching is correct. This is important for our analysis since the size classification used in the WBES is based only on employees.

10. An anonymous referee pointed out that the growth rate of assets and the growth rate of sales are likely to be highly correlated and that causality could go from the growth rate of assets to that of sales instead of vice versa. We acknowledge that the estimated coefficient for the growth rate of sales, our proxy for growth opportunities, might be biased due to this problem, although the estimation procedure used and the inclusion of the lagged growth rate of assets in the specification should help to mitigate it. In any case, even if after instrumenting some endogeneity problem persists with regard to this variable, we do think it is unlikely to affect the estimation of the financial obstacles indicator coefficient. Therefore, it is unlikely to condition the results of the hypothesis tested in this paper.

11. We also estimated the model-using log of age instead of the dummy young. We obtained a negative and significant relationship confirming that younger firms suffer more from financing obstacles. In addition, when we allow for a different impact of medium or large companies, both dummies are non-significant.

12. A large amount of research provides evidence that financial development has a significantly positive effect on economic growth (see Wachtel 2003; Levine 2005; Papaioannou 2007 for comprehensive surveys). For the euro area Guiso et al. (2004) and Ferrando, Köhler-Ulbrich, and Pál (2007) find some evidence supporting the hypothesis that the lack of financial development limits more severely the growth of SMEs.

13. Marginal impacts are evaluated at the sample mean. Bootstrap analysis with 5000 random resampling confirms the estimation results.

14. In the WBES, Germany recorded a percentage of companies reporting financing obstacles above that for Italy but a lower predicted probability of financing obstacles this is linked to factors such as the higher percentage of large and listed companies, which, according to the results shown above, are negatively related to financial difficulties.

15. The instruments used are not common across countries because of specification problems in some countries. Hence, the results are not strictly comparable across countries. Summary statistics along with correlation coefficients among regressors are available upon request.

16. Since FO is a generated regressor, standard errors should be adjusted to take into account error heteroskedasticity and autocorrelation problems. To avoid both problems, we use standard errors robust to heteroskedasticity and we include sectoral dummies (for each country, sector is the only grouping factor behind the FO variable) to avoid error autocorrelation.

17. These instruments could be divided into two categories. The fist is high growth and innovative SME facility (GIF) which provides risk capital for innovative SMEs in their early stages (GIF1): the European Investment Fund can usually invest 10–25% of the total equity of the intermediary venture capital fund or up to 50% in specific cases. They provide also risk capital for SMEs with high growth potential in their expansion phase (GIF2): the EIF can invest 7.5–15% of the total equity of the intermediary venture capital fund or, exceptionally, up to 50%. The second category is the SME guarantee facility (SMEG), which provides loan guarantees to encourage banks to make more debt finance available to SMEs, including microcredit and mezzanine finance.