The Spillover Effect of Euribor on Southeastern European Economies: A Global VAR Approach

Abstract

In this paper we employ a Global Vector Autoregressive Model (G-VAR) to examine macroeconomic and international monetary spillovers within the South Eastern European Countries over the period 2002–2016. In particular, we investigate how shocks in Euribor affect domestic interest rates, real effective exchange rates, foreign exchange reserves and the industrial production in six countries, namely Bulgaria, Croatia, Greece, North Macedonia, Romania and Slovenia, using monthly data. Our analysis shows that a negative Euribor shock has a positive effect, but of different size, on the industrial production across the countries. There is limited evidence regarding the effects of Euribor on foreign reserves, interest rates, and the real effective exchange rate. Our results provide significant implications and insights for authorities and policy makers.

Keywords:

• Euribor
• global vector autoregression
• international spillovers
• South East Europe
• transition countries

Acknowledgments

We thank the editor and the four anonymous referees for their valuable comments. This submission to Journal of East-West Businessis derived from the following PhD thesis: Petros Golitsis, (2018, March).Interactions Of Monetary

Notes

1 Golitsis, P. (2018) Interactions of Monetary Policies in South East Europe in a European Monetary Union Context. A Global-Vector Autoregressive Model (Doctoral dissertation, University of Sheffield). Sheffield. Link: http://etheses.whiterose.ac.uk/21736/1/Petros_Golitsis_PhD_White_Rose.pdf. This PhD thesis is the original source of this publication.

2 For the various monetary transmission channels see Mishkin (1996).

3 Slavov (2017), apart from Marcal et al. (2014), uses the REER as we also do.

4 Note that some limited access to the interbank rate of Slovenia in terms of the start date is present. Specifically, the starting date of this variable is the 1st of January of 2003 instead of 2002, and thus a backward linear interpolation through the econometric software EViews took place to add one year of observations. The same applies for the starting date of Foreign Exchange Reserves in the case of Greece given that the available starting date of this variable, provided by the Central Bank of Greece, is 2003 instead of 2002.

5 Their computation is based on averages of trade flows over the last 5-year period. W_ij is measured as the total trade of country i with all of its trading partners, where w_ii=0 for all i.

6 See Dees et al. (2007) for a more analytical derivation of the G-VAR model.

7 For Structural vector autoregressive (VAR) models which are important tools for empirical work inmacroeconomics, finance, and related fields, see the book of Kilian and Lütkepohl (2017).

8 It has to be added that for estimation purposes of the model, star variables (x_it*) has to be treated as integrated of order one (i.e. as I(1) and also as weakly exogenous with respect to the parameters of the VARX* model) or differently stated as “long-run forcing” variables.

9 For a more detailed description of the above see Chudik and Pesaran (2014) and Smith and Galesi (2014).

10 For the fiscal ones see Belke and Osowski (2019); Dragomirescu-Gaina and Philippas (2015); Koukouritakis et al. (2015); Niehof (2014); Ricci-Risquete and Ramajo-Hernandez (2015); Hebous and Zimmermann (2013); Favero, Giavazzi, and Perego (2011); Caporale and Girardi (2011); Nickel and Vansteenkiste (2013).

11 For a detailed presentation and analysis of the specification and estimation of the country—specific models, see Supplementary Appendix. A restriction was introduced on the cointegrating equations and was equal to minus one lag (within a VAR, at level, based on AIC). Thus, testing the over-identifying restrictions on the coefficients of the co-integrating relations—as they are associated and tested in capturing the long run relationships between the variables—we have chosen one, but for the yield output we have set them equal to zero.

12 A result that will be integrated further with the rest of the empirical findings of the G-VAR and especially the ones of the GIRFs. Still, we can report here that the results could probably indicate a more stable economy-like the Slovenian one- that moves more strongly towards a sustainable Eurozone participation.

13 On this section we use the.Generalized Impulse Response Functions (GIRFs) that within the G-VAR context and methodology were introduced by Koop et al. (1996) and were upgraded by Pesaran and Shin (1998).

14 Structural breaks were not detected in the examined period on the applied G-VAR, possibly due to the fact that shocks are roughly probable in occurring at the same time frame for all investigated SEE countries. Thus, the two financial crises (including the Euro crisis), which negatively affected Euribor's indicators,arereflected on the selected global variable (Euribor), which supports its selection as such.

15 Before we proceed to the presentation and the analysis of our results, we should add that the time frame of the forecast horizon of GIRFs being used is a 40 month one (following Smith and Galesi, 2014). This is a reasonable and relatively extended time frame, one could suggest, that is attributed to the special properties of SEE that by being not that ‘fully’ integrated with the rest of the Eurozone and Europe (that in their own turn are not integrated, in an Optimum Currency Area (OCA) context at least) it would need more time to absorb the shocks. Also, the figures, as reported, display the bootstrap mean estimates with 90% bootstrap error bounds, an error bound which is commonly used in G-VAR literature (see for example Dees et al. (2007) and Smith and Galesi (2014)).

16 This pattern is similar excluding partly Romania, in terms of the trend of the response, which is still a positive one. The drop of the Euribor, economically speaking, is expected to be beneficial for the economies given that at lower interest rate for example the Gross Fixed Capital Formation (GFCF) tends to increase, and thus subsequently the economies should be better off in an international context.