Czechia’s Choice: Would Euro Make a Difference?

• Thesis title in Czech: Volba pro Česko: Jaký rozdíl by znamenalo euro?
• Thesis title in English: Czechia’s Choice: Would Euro Make a Difference?
• Key words: Přijetí eura, Globální VAR, Pravděpodobnostní předpovědi
• English key words: Euro Adoption, Global VAR, Probabilistic Forecasting
• Academic year of topic announcement: 2020/2021
• Thesis type: diploma thesis
• Thesis language: angličtina
• Department: Institute of Economic Studies (23-IES)
• Supervisor: PhDr. Jaromír Baxa, Ph.D.
• Author: Mgr. Ivan Trubelík - assigned by the advisor
• Date of registration: 27.06.2021
• Date of assignment: 27.06.2021
• Date and time of defence: 21.09.2023 09:00
• Venue of defence: Opletalova, O105, místnost č. 105
• Date of electronic submission: 01.08.2023
• Date of proceeded defence: 21.09.2023
• Opponents: prof. Ing. Oldřich Dědek, CSc.

Guidelines

Hypotheses:
1. Hypothesis #1: The degree of connectedness between the Czech economy and the Euro Area is at least as high as between the Euro Area countries.
2. Hypothesis #2: Adopting the Euro would have positive effects on the GDP level of Czechia in the short-term.
3. Hypothesis #3: Adopting the Euro would lead to decrease of prices in Czechia in the short-term.

Methodology:
The hypotheses on the effects on adopting the Euro are generally tested via two distinct approaches: calibrated structural models or forecasts from estimated parametric/nonparametric models. The main methodology of the proposed work would exploit the Global VAR framework developed in Dees et al. (2007). In comparison to the structural models, it does not restrict the dependencies between variables ex ante by an economic theory, offering a possibility to avoid dragging of possibly incorrect researcher’s assumptions into the model. Moreover, it addresses the linkages between economies – the country specific VAR models with foreign countries weighted e.g. by trade shares are integrated into the overall complete econometric model, where all the variables are endogeneous.

Such a model can then be used to produce counterfactual analysis by imposing restrictions on some of the endogeneous variables to simulate the event of certain policy change. Pesaran et al. (2007) propose forecasting not only point estimates, which are hard to distinguish from the forecast errors themselves, but conditional probability distributions of events. The models exploit high-level aggregated data, such as GDP, inflation or short- or long-tem interest rate to name a few. These data are available in the International Financial Statistics database of the International Monetary Fund for up to recent years. Other data needed are available in the database of the World bank, namely the price of oil serving as exogeneous variable in most of the country-specific VAR models.

The presence of linkages between the modelled variables can postulated by the commonly perceived nature of the international trade setting. Nevertheless, it is possible to asses the degree of interconnectedness between the economies, e.g. by employing the recent approach developed by Diebold & Yilmaz (2015). Their analysis can be divided into two parts: first, a forecast error variance decomposition is done using an estimated VAR kind of model. Second, the classical tools of the network analysis are applied to asses the strength of linkages between the given subjects. This promiment new method has a very appealing property that it measures not only the direction (as in a Granger causality analysis), but also the strength of the connections (as done with a multivariate GARCH model).

It would be interesting to look at the strength of connection for the European countries, as well as to compare it with a well-established lasting monetary union serving as a benchmark. For this case, the U.S. could be chosen and the data about GDP of individual states provided by U.S. Bureau of Economic Analysis would provide the set of information needed for estimation of a Large BVAR model, which was proposed by Bańbura et al. (2010). A classical and a new generalized variance decomposition method developed in Lanne & Nyberg (2016) would be compared in their impact on the resulting network properties.

If the connection between the studied economies is found significant, it then would be incorrect to use a methodology not taking this into account – still these are popular for their relative simplicity in estimation and interpretation. These approaches include the difference-in-difference estimator (which is hard to implement in macroeconomics, because of not having dichotomous experimental and control groups), synthetic control estimator (which suffers from far from sufficient similarity between the modelled and control subjects) or event based study, which is particularly prone to forecast error and policy effect confusion (it only forecasts from the time point before the point and compares to realized values).


Expected Contribution:

The results of the proposed work could serve as a supporting pillar in the discussion of Euro adoption in the Czech Republic. It would help to clarify whether the decision not to join the monetary union was indeed right from a macroeconomic perspective. It would therefore not provide an explicit estimate of how forfeiting of the Czech koruna would affect the economy today, but rather look at interesting alternative scenarios in the past.

Comparing counterfactuals with the past is first technically more feasible, because trying to forecast the future and make comparisons with uncertain realizations generated also by a given model could be quite unreliable. Second, even while only finding answers to past what-if scenarios, examing the different possible settings of accession (timing, exchange rate) could give a deeper insight into the optimal conditions for adopting the Euro in the times to come.

Next, the work would use a model that gains its parameters from estimation of the data. This could be an advantage in comparison to structural models that have been employed in various related works and that might possibly be prone to misspecification while the researcher is designing the mutual effects of variables from a favored economic theory. One study with such a model was done by Svačina (2015), which is a unique contribution in the sparse literature on the estimated effects of the Czech adoption of the Euro. To sum it up, this work would fill the blank space in the emprical research on the common currency question in Czechia, which is a neccessary condition of acquiring a sound and complete picture of this major policy decision.

References

Bańbura, M., Giannone, D., Reichlin, L. (2010): Large Bayesian Vector Auto Regressions. Journal of Applied Econometrics 25(1): 71–92.

Caselli, F.G. (2017): Did the Exchange Rate Floor Prevent Deflation in the Czech Republic? Review of Economics & Institutions 8(2): 1-31.

Conti, M. (2014): The Introduction of the Euro and Economic Growth: Some Panal Data Evidence. Journal of Applied Economics 17(2): 199-211.

Dees, S., Mauro, F.d., Pesaran, M.H. (2007): Exploring the International Linkages of the Euro Area: A Global VAR Analysis. Journal of Applied Econometrics 22(1): 1-38.

Diebold, F.X., Yilmaz, K. (2015): Financial and Macroeconomic Connectedness: A Network Approach to Measurement and Monitoring. Oxford University Press, USA.

Lanne, M., Nyberg, H. (2016): Generalized Forecast Error Variance Decomposition for Linear and Nonlinear Multivariate Models. Oxford Bulletin of Economics and Statistics, 78(4): 595-603.

Pesaran, M.H., Smith, L.V., Smith, R.P. (2007): What if the UK or Sweden Had Joined the Euro in 1999? An Empirical Evaluation Using a Globar VAR. International Journal of Finance and Economics 12(1), 55–87.

Svačina, D. (2015): Impacts of the Euro Adoption in the Czech Republic. Master’s Thesis. Charles University in Prague, Institute of Economic Studies.

Preliminary scope of work

Motivation:
The Euro – a symbol of political unity from the pan-European perspective, but most importantly the leading thread in many parts of economic life, ranging from travel, trade payments and investment on the individual level to nation-wide monetary conditions and bank supervision. It embodies an unprecedented cooperation project in the history of the European continent and as such offers new perspective opportunities to Europeans, as well as inherent potential threats on the other side.

When joining the European Union in 2004, the Czech Republic obliged to adopt the Euro while signing the Maastricht Treaty, yet until now, it has not done so. One of the reasons is undoubtedly the complexity of such a decision when considering the benefits and losses in future public welfare. Indeed, the adoption effects differ across the Eurozone members – with evidence given e.g. by Conti (2014) for the estimated difference in GDP per capita increase. Having information on probability or expected magnitude of relevant economic events (changes in output, inflation, interest rates) after adoption of the common currency would constitute an analytical support for the decision with regards to the economic side of the problem and thus help to clarify the discussion about joining the monetary union.

Formally, Czechia is supposed to join EMU at some point of time and, therefore, the research should also focus on the optimal timing of the potential accession. The underlying idea of different effects of implementing such an economic policy stems from the notion of changing global economic environment together with varying interconnected between countries. To put forward an example, a country could join the monetary union that is in prosperous economic upturn or, in contrast, in times of onset of an economic crisis. The two scenarios would most probably have a different impact on the new member, which would furthermore be dependent on the degree of connectedness between the economies. In the Czech Republic, there is in addition a reinforced fear from the loss of the independent monetary policy, fed by the recent experience of CNB interventions in 2013-2017 in the foreign exchange market, which Caselli (2017) estimated as successful in evading the deflation territory.

Outline:
1. Motivation: the empirical literature about Czech adoption of the Euro does not provide estimates to possible effects with a complex econometric model. However, this is a key question from the economic point of view to be answered, which cannot be absent while the mosaic of making a decision impacting such a wide spectrum of different areas of life is put together.
2. Literature on Euro adoption: the section would discuss the methodologically related foreign literature on the effects of the common currency, but also look studies approaching the topic with different tools. The implications of the optimal currency areas on the level of synchronization of the economies in question will be discussed and such existing estimates will be surveyed.
3. FEVD networks: data will be presented and connection networks will be estimated from variance decompositions for the U.S. states and European countries, separately, in different time points. Relevant network metrics will be computed to evaluate the degree of integration between the countries.
4. Data & Methodology of GVAR: the variables used in the Global VAR model will be described. The principles of the model will be outlined, together with a discussion of its advantages and possible pitfalls.
5. Estimation Results: the results of different counterfactual settings will be presented and the hypotheses about the Euro adoption evaluated for each of them. The robusstness of the results will be also be addressed.
6. Conclusion: the last section will summarize the findings about whether the Czech and Eurozone economies are sufficiently connected before joining the common currency and if doing so in the past would have been a favourable decision.