Skip to main content
SpringerLink
Log in

The impact of Basel II on the debt costs of German SMEs

  • Original Paper
  • Published:
Journal of Business Economics Aims and scope Submit manuscript

Abstract

We investigate the ex-post effects of the Basel II reform on the debt costs of German small and medium-sized enterprises (SMEs). We assume that Basel II formalized the credit assessment procedure of banks and that especially SMEs might face higher costs of debt as they tend to have comparably high proportions of bank credit, and consequently lower ratings than bigger companies. Beyond that, banks might try to refinance those additional rating costs by imposing higher interest rates on debtors. The results presented in our paper indicate a significant overall rise of debt costs of SMEs that have debt relations with IRBA-certified banks since 2007. Furthermore, low-risk companies benefit from comparably lower loan rates, whereas risky firms face comparably higher loan costs than before. The results are less obvious for SMEs that are rated under RSA, but still present for highly risky companies. We test for several influences of the financial crisis. However, we do not find convincing arguments that would reason a rise of the costs of debt in regard of the crisis. Yet, we still observe higher debt costs of SMEs, particularly of highly risky companies, since 2007 that indicate a change of bank lending behavior in Germany.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

Notes

  1. All empirical analyses on the SME definition of the EU recommendation 2003/361: SMEs are categorized into micro firms (<10 employees and either <€ 2 million of sales revenues or € 2 million of total assets), small firms (10–49 employees and either € 2–10 million of sales revenues or € 2–10 million of total assets) and medium-sized firms (50–249 employees and either € 10–50 million of sales revenues or € 10–43 million of total assets). Definitions of SMEs by the Institut für Mittelstandsforschung differ slightly from those of the European Commission and include companies with fewer than 500 employees or € 50 million annual turnover.

  2. Mean calculations in our sample display an average equity ratio of 27 %. The ratio increases from about 25 % in 2003 to about 30 % in 2010. SMEs seem to have gradually improved their capital structure over the last decade.

  3. Banks had to build up a required minimum quote of regulatory equity capital at the level of 8 % of their risk-weighted assets. The individual risk weights RW were determined by category of the on-balance-sheet assets (Basel Committee on Banking Supervision 1988): Cash, claims on governments and central banks RW = 0 %; Claims on banks incorporated in the OECD RW = 20 %; Loans fully secured by mortgage on residential property RW = 50 %; Claims on private sector (including corporations) and other assets RW = 100 %.

  4. For a more detailed explanation of the components of Basel II see for example Szczesny (2003) or Wilkens et al. (2001).

  5. For details about the different approaches compare Szczesny (2003) and Basel Committee on Banking Supervision (2003a).

  6. In our sample, 64.7 % of all observations have debt relations with at least one IRBA certified bank.

  7. The risk weighting factor for retail loans is 75 % (Basel Committee on Banking Supervision 2004).

  8. Empirical results of a survey of German banks support this assumption. See Haller et al. (2008).

  9. Dietsch and Petey (2004) indicate that by clearly declaring SME as riskier compared to larger companies.

  10. Missing values in our dataset are coded as zero. Therefore, we eliminated all observations equal to zero for our calculations. Thereby, we could have eliminated observations with economically correct zero values, but we did not expect this effect to significantly influence our calculations, as this might only be accurate in rare cases. We limited our sample to observations with an equity ratio, an interest rate and a ratio of amounts owed to credit institutions to total liabilities with values between 0 and 1. A detailed examination of the dataset indicates a bias by outlying observations. Thus, we excluded the 100th percentile of total assets. This restriction leads to an unbalanced panel with 58,497 observations with non-missing data for our major regression model. As consequence of this harsh reduction of observations, we assumed that we sorted out small companies in particular. Creating a balanced panel would lead to a severe reduction of observations. Therefore, we allowed the panel to be unbalanced.

  11. Our dataset only allows for a distinction between consolidated group accounts and ‘unconsolidated’ accounts. Within the unconsolidated accounts, we were not able to reliably differentiate between companies that do and do not belong to a group. Thus, our dataset might be distorted by observations that are affiliated. Affiliated companies might have the possibility to negotiate loan conditions with the consolidated group accounts and consequently receive better conditions in general. However, Szczesny and Valentincic (2013) also use data from unconsolidated financial statements for their analysis.

  12. This differentiation might lead to some distortion as SMEs might have debt relations with both RSA and IRBA-certified banks. As we cannot identify the corresponding amounts of credit owed to single credit institutions, we assign those observations to the IRBA subsample. We identified IRBA-certified banks by data from the Deutsche Bundesbank (2013d).

  13. The following insolvency risk groups are included: 1 = low risk, 2 = moderate risk, 3 = high risk, 4 = very high risk.

  14. Please compare Creditreform Rating (2015) for an overview of qualitative and quantitative data the agency likely uses for the rating.

  15. We define the dummies as follows: 1 = LRISK, 2 = MIDRISK, and 3 and 4 = HRISK.

  16. We also included a control variable for the overall credit volume granted from German banks to German companies based on aggregated data of the Deutsche Bundesbank (2013b), but the variable had to be excluded because of multicollinearity.

  17. Graham et al. (2008) use the LIBOR to show the macroeconomic development of refinancing interest rates of banks.

  18. We furthermore included CREDIT (=amounts owed to credit institutions/total assets) and other debt capital proxies in our model 1 to control for other debt financing influences. We observe a significant positive effect of the reform for IRBA banks, but no effect for RSA banks. We did not include CREDIT in our main models as a comparison of a random sample of our companies with actual annual statements showed that the data in our database for this certain balance sheet position is not reliable. About 25 % of our randomly sampled companies disclosed values for bank loans, but their data was not included in our database whereas data for other main balance sheet positions were correct.

  19. Results of Model 1 indicate a spread that equals \(\beta_{3}\) before 2007, and \(\beta_{3} + \beta_{4} + \beta_{5}\) since 2007.

  20. We confirmed this assumption by a comparison of mean values in our dataset.

  21. We added corresponding Wald-tests in our calculations.

  22. A t test of REFIN before and after the Basel II amendment discloses a significant rise of the variable. Nevertheless, a Wald-test confirms that REFORM and REFIN measure different effects.

  23. See Deutsche Bundesbank (2013a) for further details about the development of interest rate levels.

  24. The cash flow variable available in the DAFNE database does not equal the operating cash flow. Therefore, we use the balance sheet approach to calculate the operating cash flow by subtracting total accruals from earnings before extraordinary items. We use the following approach mentioned by Ball and Shivakumar (2005) to calculate accruals: \({\text{Acc}}_{t} = \varDelta {\text{Inventory}}_{\text{t - 1,t}} + \varDelta {\text{Debtors}}_{\text{t - 1,t}} + \varDelta {\text{Other Current Assets}}_{\text{t - 1,t}} - \varDelta {\text{Creditors}}_{\text{t - 1,t}} - \varDelta {\text{Other Current Liabilities}}_{\text{t - 1,t}} - {\text{Depreciation}}_{t}\).

  25. Former studies used Tobin's Q instead of sales growth to control for future profitability. However, as we do not have data to calculate Tobin's Q for private companies we follow Lenger and Ernstberger (2011).

  26. Descriptive data reveals that SMEs’ average investments in our sample rose in the early stage of the financial crisis that is in 2007 or 2008, and almost equaled zero in 2009 (compare Table 8). We also observe this development if we subdivide our sample into our four risk categories. Results contradict those of Duchin et al. (2010). In the US, real effects of the crisis are already noticeable in 2008.

  27. A closer look at the average volume of amounts owed to credit institutions per year does not reveal a significant change of the bank debt ratio during the crisis. In addition, we inspect the development of newly granted bank credits during the respective period. We observe an exceptional rise of new credit granting, a slight rise in 2008, and a decline in 2009. Hence, only the drop in 2009 might be an indicator of credit amount restrictions in our sample. As we assume that a credit crunch mostly likely took place in 2007 and 2008, this does not directly shake our other statements. However, we cannot control for the fact whether SMEs did seek additional debt financing that was not granted eventually.

References

Download references

Author information

Authors and Affiliations

  1. Chair of Management Accounting, University of Wuerzburg, Stephanstraße 1, 97070, Würzburg, Germany

    Alexandra Schindele & Andrea Szczesny

Authors
  1. Alexandra Schindele
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Andrea Szczesny
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Andrea Szczesny.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Schindele, A., Szczesny, A. The impact of Basel II on the debt costs of German SMEs. J Bus Econ 86, 197–227 (2016). https://doi.org/10.1007/s11573-015-0775-3

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11573-015-0775-3

Keywords

JEL Classifications

Search

Navigation