A new look into the prevalence of balance sheet or competitiveness effect of exchange rate depreciation in a highly euroised economy

Abstract

This article empirically tests the impact of exchange rate depreciation on sectoral performance proxied by investment or, alternatively, sales. It measures the balance sheet and the competitiveness effect in a country that records very high levels of liability euroisation. Panel data methodology is applied on a dataset of 20 Croatian non-financial sectors combining macroeconomic and sectoral financial information. Results confirm there are strong negative effects of liability euroisation on both investment and sales. Negative balance sheet effects and very small positive competitiveness effects are found as well, adding up to a negative overall exchange rate depreciation effect on sectoral performance. Moreover, we find evidence that the corporate sector does not hedge against exchange rate exposure and that the domestic financial system is a constraining factor for corporate investment growth. We also find evidence of size asymmetries related to bank lending relationships.

Notes

1. Details about the definition and sources of all variables can be found in Table A1 of the Appendix.

2. HRK is the conventional abbreviation for the Croatian kuna, the official currency in Croatia.

3. A large number of other balance sheet and macroeconomic variables were included in the analysis, such as indicators of performance, illiquidity, tradability, GDP growth rates, a ‘recession dummy’ etc. However, they did not appear to be statistically significant in explaining investment or sales dynamics. Therefore, those results are not presented in the article but are available upon request.

4. Equations (4) and (5) were additionally estimated using different sector-specific and macroeconomic variables as explanatory variables. When we exclude the competitiveness effect and the interest rate variable from the specification with investment as the dependent variable, the signs and coefficients remain unchanged. Similarly, excluding the short-term debt ratio from the model specification with sales does not affect our results either. Including different indicators of revenue, turnover, earnings and capital yielded results very similar to those reported. This implies the results are robust.

5. The Sargan test and the autocorrelation tests indicate that the number of lags used in the dynamic panel data estimation were appropriate.

6. The recession dummy is defined as a variable that takes the value 1 for a recessionary year and 0 otherwise. For the case of Croatia, the value 1 is given for the years 2008 and 2009. The recession dummy variable is included in the model, as we believe that in a period of economic downturn companies find it rather difficult to borrow at longer maturities, because creditors opt for less risky loans with shorter maturity. Consequently, companies' short-term to total debt ratio increases.

7. We ran several regressions with different model specifications and confirmed that the results presented here are robust. When we exclude the export ratio variable, all the coefficient signs remain unchanged and the coefficients themselves change by negligible amounts. Alternatively, when we substitute the export ratio variable for a tradability indicator, the signs and coefficients stay unchanged. We tried also including different variables, such as indicators of illiquidity, labour and earnings, and the signs and coefficients remained unchanged from those reported here.