ورود به مجارستان به منطقه یورو: درس هایی برای چشم انداز عضوها از منظر سیاست های پولی

This paper evaluates the conduct of monetary policy in Hungary using standard Taylor rules as well as extended rules that incorporate real exchange rate effects. Moreover, we explicitly consider the impact of future euro area entry by estimating instrument rules that permit an influence from Maastricht Treaty inflation requirements via the estimation of Markov switching models as well as by estimating a differential rule vis-à-vis the existing euro area. Lastly, the paper also considers the impact on policy rules from the large data revision that affects real exchange rate and output estimates. I find that interest rate setting behavior in Hungary does not resemble that of the euro area. Also, counterfactual experiments reveal that the potential macroeconomic costs of entry into the euro area sooner rather than later may be lower than if membership in the single currency area is delayed beyond 2008.

In May 2004 Hungary, along with nine other states, joined the European Union (EU). Since there were no derogations, Hungary will someday join the euro area (EA). The only question is when. Participation in European Monetary Union (EMU) consists of several steps, namely adoption of the acquis communautaire and meeting the relevant Maastricht Treaty provisions. The former conditions involve, in essence, the requirement of central bank autonomy prior to accession, and meeting the strictures of the Stability Growth Pact (SGP) following accession (see http://europa.eu.int/comm/enlargement/negotiations/chapters/chap11). The Treaty provisions are well-known (e.g., see ECB, 2004). While it is debatable whether the National Bank of Hungary (MNB) is sufficiently autonomous, particularly in light of recent events concerning governance questions (e.g., see Newsmakers, 2005), the EU at least has closed the book on this stage of the process (though not the ECB; see ECB, 2004, p. 58). It should be emphasized, at the outset, that while the paper focuses attention on the Hungarian experience, the broad findings reported below are likely to reflect considerations that are common to all states that joined the EU in 2004. It is useful to note that, as of this writing, there are recent EU members that are keen to join the EA as quickly as possible (e.g., Estonia, Lithuania, Malta) while others, such as Poland, the Czech Republic and Hungary, have indicated a desire to possibly delay adoption of the euro. Decisions hinge not only on Maastricht and SGP considerations but also on a perceived trade-off, as it were, between the benefits of EA membership (e.g., relatively lower nominal interest rates, improved access to credit markets) versus the short-run benefits of greater fiscal and exchange rate flexibility on generating improved economic growth. While the road to the euro involves fulfilling conditions that cover both fiscal and monetary policies, this paper focuses on the monetary policy challenges facing Hungary as it seeks to join EMU. Optimism that the date of entry might be as early as 2008 has been replaced with skepticism that even 2010, or 2012, are feasible dates for euro adoption (e.g., Associated Press, 2005, MTI, 2005 and Miller, 2006).1 Three provisions of the Maastricht Treaty are particularly germane to our analysis. They are: the price stability objective, that is, attaining an inflation rate that does not exceed the reference value,2 the absence of severe tensions in the nominal exchange rate over a 2-year period following membership in the new exchange rate mechanism (called ERM II), and a nominal long-term interest rate that does not exceed the reference value.3 Within these limits we examine, using a mixture of estimates of instrument rules and counterfactual experiments on these same rules, how the conduct of monetary policy has evolved in Hungary since 1996, and the interest rate prospects if Hungary were to join the euro area in the not too distant future. Why adopt such an approach? For good or ill, evaluation of central bank behavior is thought to be usefully summarized via the device of reaction function estimation. While no central bank rigidly follows such rules, there is growing acceptance that interest rate setting behavior in a reasonably transparent and accountable central bank can be reasonably summarized via estimates of Taylor type rules (e.g., Poole, 2005). One could interpret the similarity of reaction functions with the existing euro area and the new member states4 (NMS) as an expression of the latter group's desire “… to show they were normal countries and now want to join the euro to show they are good Europeans” (The Economist, 2006).5 Hence, the paper posits that membership in the euro area ought to result in an instrument rule followed by Hungary's MNB that “looks” like existing ones estimated for the euro area (e.g., Gerdesmeier and Roffia, 2004).6 It could be argued that the period studied is not a homogeneous one, as Hungary struggled with economic shocks that would, in all likelihood, perhaps not be relevant to a member of the EU, still less a member of the euro area. The adoption of a form of inflation targeting in mid-sample further underscores the dangers of drawing strong implications from empirical estimates based on the 1995–2005 sample. This partly explains the resort to counterfactual experiments. Nevertheless, this negative view ought to be counterbalanced by the fact that other countries that have undergone changes in monetary policy strategies (e.g., Australia, Canada, and New Zealand),7 including the adoption of inflation targeting, also experienced significant changes in the economic environment before, and after, the ‘regime’ shift without negating the usefulness of the tools employed in the analysis below. The bottom line is that successive Hungarian governments were more or less pre-occupied with fiscal control while monetary policy was explicitly given the job of delivering lower and more stable inflation. Further, monetary policy has been guided by principles that were widely adopted and accepted in the industrial world. Therefore, there is no reason to believe, insofar as overall macroeconomic objectives are concerned, that the period under investigation cannot be used to obtain useful inferences about the prospects for joining the EA.

This paper has considered the problem of evaluating the conduct of monetary policy in Hungary based on estimates of instrument rules. The monetary policy of the MNB does not resemble that of the ECB or central banks in the euro area. Estimates of Taylor rules are quite fragile and only forecast-based rules come closest to mimicking rules estimated for the euro area. A counterfactual experiment suggest that the MNB's interest rate policy has been lower than would be needed to meet the Maastricht standards, and that delaying the introduction of the euro to 2010 instead of 2008 might be relatively costly from a macroeconomic perspective, or reflects domestic considerations (viz., management of debt costs) that can be fulfilled via a relatively autonomous monetary policy. Since estimates of the output gap for Hungary may be unreliable, it may be worth considering estimation of reaction functions that rely on a wider variety of output gap estimates. Next, the paper suggests that the fragility of estimated rules is tied to the operation of a managed floating regime. A clearer indication of the role of the exchange rate regime might be obtained by generating estimates of the pass-through effects of monetary policy. It is well-known that pass-through effects have declined in the industrial world but the same may not be true for Hungary. Lastly, our instruments are generally weak. We should consider augmenting the list of instruments to include money growth or, possibly, fiscal variables. Lastly, even if one accepts resort to the estimation of instrument rules, one must keep in mind the reservations and limitations associated with Taylor rules (e.g., see Svensson, 2005 and McCallum and Nelson, 2005). These extensions are left for future research.