آیا سیاست پولی در کشورهای جدید عضو اتحادیه اروپا نامتقارن است ؟

Estimated Taylor rules have become popular as a description of monetary policy conduct. There are numerous reasons why real monetary policy can be asymmetric and estimated Taylor rules nonlinear. This paper tests whether monetary policy can be described as asymmetric in three new European Union (EU) members (the Czech Republic, Hungary, and Poland) which apply an inflation targeting regime. Two different empirical frameworks are used: (i) Generalized Method of Moments (GMM) estimation of models that allow discrimination between sources of potential policy asymmetry but are conditioned by specific underlying relations, and (ii) a flexible framework of sample splitting where nonlinearity enters via a threshold variable and monetary policy is allowed to switch between regimes. We find generally little evidence for asymmetric policy driven by nonlinearities in economic systems, some evidence for asymmetric preferences, and some interesting evidence on policy switches driven by the intensity of financial distress in the economy.

The monetary policy setting in the Central and Eastern European countries (CEECs) evolved substantially during the economic transition. These countries experimented with diverse monetary policy and exchange rate frameworks until the late 1990s, when their policy regimes fell into line with the then influential bipolar view, i.e., that intermediate regimes between hard exchange rate pegs and free floating are not sustainable. Some countries (the Baltic States and Bulgaria) adopted hard pegs, which put a significant constraint on their monetary policy, while other economies decided to maintain an overall flexible exchange rate, allowing their central banks to pursue internal macroeconomic targets (the Central European countries and Romania). Ongoing nominal and real convergence coupled with EU membership and the obligation to meet the Maastricht criteria put another constraint on policy making in general and monetary policy in particular in the New Member States (NMS). Some countries have merely formalized their previous exchange rate pegs by means of participation in the Exchange Rate Mechanism II (ERM II) and consecutive euro adoption, while others have retained their monetary policy autonomy under the framework of inflation targeting (IT) to the present day. Given the relative success of the latter countries in achieving price stability with decent levels of economic growth, it is of interest to understand their monetary policy conduct in greater detail. In particular, it seems interesting to empirically explore interest rate setting behavior under the IT mandate as well as the subtle differences between these countries. There is a vast amount of empirical research on the way central banks handle interest rate setting. Since Taylor (1993), researchers have been estimating Taylor rules, as they seem to characterize well the interest rate setting of central banks. Clarida et al., 1998 and Clarida et al., 2000 propose that central bankers are proactive rather than reactive and set interest rates with respect to expected values of macroeconomic variables. Estimated monetary policy rules typically take a linear form, assuming that monetary policy responds symmetrically to economic developments. The theoretical underpinning of the linear policy rule is the linear-quadratic (LQ) representation of macroeconomic models, with the economic structure assumed to be linear and the policy objectives to be symmetric, as represented by a quadratic loss function (e.g., Clarida et al., 1999). However, when the assumptions of the LQ framework are relaxed, the optimal monetary policy can be asymmetric. Asymmetric monetary policy implies that the monetary policy rule, which is a schematization of the policy reaction function, is nonlinear. In reality, however, asymmetric monetary policy can arise even when the underlying relations are essentially linear but the policy responses (slope elasticities) are different for positive and negative shocks. Unfortunately, owing to difficulties with shock identification, most empirical research relates asymmetric policy only with departures from the LQ framework and, therefore, nonlinear underlying relations. Departures from the LQ framework involve two different sources of policy asymmetry. The first source lies in nonlinearities in the economic system. A common example of such nonlinearity is a steeper inflation–output trade-off when the output gap is positive. Such convexity of the Phillips curve (PC) implies that the inflationary effects of excess demand are larger than the disinflationary effects of excess supply (e.g., Laxton et al., 1999). This can lead optimizing central bankers to behave asymmetrically (Dolado et al., 2005). However, asymmetric monetary policy can also be related to genuinely asymmetric preferences of central bankers. While central banks in the past were prone to inflation bias due to a preference for high employment or uncertainty about its natural level (Cukierman, 2000), reputation reasons can drive central banks, especially those pursuing IT, to have an anti-inflation bias, which means that they respond more actively when inflation is high or exceeds its target value (Ruge-Murcia, 2004). Looking at monetary policy decisions from the risk management perspective, it seems plausible that central banks would like to avoid tail risk, which implies a disproportional response to certain vulnerabilities bringing about asymmetric policy responses. For example, deflationary risks in the US around 2003 could be seen as a factor behind its policy rate hovering around 1% for a rather extended period. The CEECs may also be more vulnerable to certain risks, such as those stemming from other emerging countries, e.g., the 1998 Russian crisis. In general terms, real monetary policy conduct seems to be too complex to be described by a simple linear equation, and nonlinear representation of monetary policy may be more appropriate irrespective of its underlying sources. Several empirical studies have provided evidence that the monetary policy setting of many central banks may really be characterized as asymmetric. An asymmetric loss function was found to affect the decisions of the Bank of England (Taylor and Davradakis, 2006) and the US Fed (Dolado et al., 2004). Bec et al. (2002) confirm that the US Fed, the Bundesbank, and the Bank of France responded more actively to inflation during economic booms. Leu and Sheen (2006) and Karagedikli and Lees (2007) detect an asymmetric response to the output gap by the Reserve Bank of Australia. Surico (2007a) claims that in its early years the European Central Bank (ECB) responded more strongly to output contractions than expansions and that the level of the interest rate itself was a source of policy asymmetry. Surico (2007b) establishes similar evidence of the Fed's asymmetric response to the output gap in the pre-Vocker era and quantifies the inflation bias induced by such policy. Asymmetries due to convexity of the PC found in some European countries (Dolado et al., 2005) and the ECB (Surico, 2007a) were linked to wage rigidity in European countries. A few studies (Maria-Dolores, 2005, Frömmel and Schobert, 2006, Mohanty and Klau, 2007, Orlowski, 2010, Paez-Farrell, 2007 and Vašíček, 2010) provide some evidence of linear monetary policy rules of the CEECs. However, some narratives suggest that monetary policy may also be asymmetric in these countries. In particular, inflation targeters may show anti-inflationary bias and therefore asymmetric policy due to reasons of reputation. The case for anti-inflationary bias could arguably be even stronger in these countries as the announced inflation targets, unlike in most developed countries, had a downward-sloping trend and the countries were de facto targeting disinflation. Under such circumstances, the central banks often declare multi-year targets (Mishkin and Schmidt-Hebbel, 2001) and can treat the bottom of the (short-term) inflation target band or the (short-term) point target undershooting more leniently in order to approach the (lower) long-term inflation target faster. However, Jonáš and Mishkin (2004) argue that such an opportunistic approach to disinflation (Orphanides and Wilcox, 2002) can make monetary policy less predictable, which is in fact problematic for IT credibility. The empirical evidence on asymmetric monetary policy setting in the NMS is very limited. Horváth (2008) employs simple subsample analysis along the sign of the deviation of inflation from the target for the Czech National Bank and finds that monetary policy was asymmetric in the first years after the adoption of IT and symmetric afterwards. The reason was arguably the need to gain credibility and to anchor inflation expectations. On the other hand, IT is flexible enough to allow policy makers not to contract demand when inflation is slightly above the target and the shocks are likely to be short-lived (Blinder, 1997). Similarly, it seems plausible that other concerns, such as economic growth and financial stability, can lead to the temporary dismissal of inflation targets. In this paper, we test the hypothesis of asymmetric monetary policy in three Central European NMS: the Czech Republic, Hungary, and Poland, who have adopted the IT framework and maintain a flexible exchange rate.1 We employ two empirical frameworks to test for policy asymmetry: (i) a framework based on an underlying structural model that modifies the LQ framework, which allows discrimination between sources of policy asymmetry but is conditioned by the specific model setting; and (ii) a flexible econometric framework where monetary policy is allowed to switch between two regimes according to a threshold variable. Besides the common choices for the threshold variable, such as the deviation of inflation from the target and the business cycle stance, we use a variable that tackles potential policy asymmetry along positive and negative shocks. In particular, we focus on the degree of financial stress in the economy to see whether central banks behave differently when the economy is distressed. We find mixed evidence in terms of asymmetric behavior of central banks. First, we find no evidence of asymmetric policy driven by nonlinearities in the economic system. Second, we obtain some indications of asymmetric policy driven by preferences, in particular in terms of inflation and the actual deviation of the interest rate from its long-term equilibrium value. Third, we detect possible policy switches driven by the degree of financial distress in the economy. In particular, central banks seem to alter their monetary policy stance when the economy is faced by severe financial stress. Our empirical findings imply that monetary policy in the three NMS is possibly handled in a slightly different fashion than the legally grounded symmetric IT mandate suggests. It seems especially interesting to further analyze whether financial stability concerns affect policy-making as a (implicit) target or are considered because of their potential effect on inflation. The rest of the paper is organized as follows. The next section briefly reviews the main rationales for asymmetric monetary policy. In Section 3, we present the empirical strategies that will be used to test for policy asymmetry, and in Section 4 we present our dataset. In Section 5, we review the empirical results. The final section concludes.

Numerous empirical studies try to describe monetary policy decisions by means of estimated Taylor rules. There are different reasons why monetary policy can in fact be asymmetric, in the sense that the intensity of the central bank response varies according to economic developments. Our empirical analysis tries to reveal whether monetary policy could be described as asymmetric in three NMS that apply IT (the Czech Republic, Hungary and Poland). However, this study aims at providing some intuition about the sources of policy asymmetry rather than a specific test of whether policy is asymmetric or not. We find that the overall evidence is mixed. When we use GMM estimation of nonlinear policy rules derived from specific underlying models (Dolado et al., 2004, Dolado et al., 2005, Surico, 2007a and Surico, 2007b) we do not find any rationale for asymmetric policy in terms of nonlinear economic relations. On the other hand, there is some indication of asymmetric preferences in inflation; in particular, the Czech National Bank seems to have weighted situations where inflation exceeded the target more heavily than those where it was below target during the initial period of the IT, while the opposite pattern was identified for Hungary. While the former finding is consistent with asymmetric policy handling during the disinflation period, the latter does not have any clear interpretation but can possibly be linked to inconsistencies between inflation and exchange rate targets pursued in Hungary. Interestingly, for all three countries we reveal a preference to limit the volatility of the current interest rate from its equilibrium value. For the Czech Republic and Hungary, we detect distaste for actual interest rates exceeding the equilibrium value, and for Poland we find that too low interest rates were of concern. In addition, a preference for lower rather than higher interest rates can be an indication of a preference to avoid contractions, while the opposite points to a preference for price stability. The previous results rely on the specific nonlinear form because they are derived from specific parametric models. Although such an approach allows for discriminating between different sources of policy asymmetry, it can turn problematic when the underlying relations are not observable. Consequently, as an alternative we use a method of sample splitting where nonlinearities enter via a threshold variable and monetary policy is allowed to switch between two regimes (Hansen, 2000 and Caner and Hansen, 2004). Besides the inflation and output gaps, we used a financial stress index as a competing threshold variable. The threshold effects are most evident with the financial stress index. While the Czech and Polish central banks seem to face financial stress by decreasing their policy rates, the opposite pattern is found for Hungary. The policy implications of our empirical findings can be summarized as follows. First, the mixed evidence on asymmetric responses to expected inflation should be confronted with the fact that the IT regime, as implemented by most central banks, de facto legally implies symmetric policy handling. The target is usually expressed by a point value or band with no recognition that positive deviations are less desirable than negative ones. However, the narratives suggest that policy handling can often be asymmetric when IT is used as a disinflation strategy, which can either be related to the aim to approach the long-term inflation target faster or to a significant uncertainty which is associated with the achievement of the mid-term targets. In this regard, it seems interesting to ask to what extent the asymmetric preferences in inflation revealed for the Czech National Bank contributed (along with shocks such as unexpected koruna appreciation) to the inflation target undershooting that often occurred during the first years of Czech inflation targeting.22 Second, the indication that the degree of financial distress can alter the monetary policy stance has two possible interpretations; it is an indication either that financial stability is a goal which is pursued (even implicitly) by central banks (irrespective of its effect on future inflation and output), or that central bankers adjust policy rates in the awareness that financial instability can affect the future path of the macroeconomic variables it targets. The latter view can be linked to the financial accelerator literature (Bernanke et al., 1996), suggesting that financial vulnerabilities can amplify adverse shocks to the economy. In the aftermath of the recent crisis, the former approach seems to be favored over the latter. Still, there is a discussion whether monetary policy that pays attention to financial stability should be reactive or pre-emptive (see, for example, Borio and White, 2004, or Cúrdia and Woodford, 2010). Our study de facto tests reactive policy, as the EM-FSI is a coincident indicator of the stability of the financial system. Finally, the forecasting models employed in central banks commonly describe monetary policy in a symmetric form (for example, in the case of IT central banks only in terms of inflation). Therefore, the potential asymmetry of actual policy decisions could open a gap between the modeling apparatus of central banks and practical policy decisions. There are different avenues of future research. First, it could be interesting to compare the behavior of central banks in the NMS and in other emerging countries that use IT but have faced very different economic challenges, such as South Africa, Mexico or Chile. Second, the models that were used for the derivation of nonlinear policy rules (Dolado et al., 2004, Dolado et al., 2005, Surico, 2007a and Surico, 2007b) could be extended to include different aspects of small open economies to derive model-based nonlinear policy rules that are more suitable for the NMS. Third, with respect to the threshold model, the assumption of an exogenous threshold variable can be too restrictive given the forward-looking nature of IT. Recently, Kourtellos et al. (2009) extended the model of Caner and Hansen (2004) to include an endogenous threshold variable. Finally, more complex econometric techniques such as Markov switching models (Assenmacher-Wesche, 2006) or state space models (Kim and Nelson, 2006) could be employed to take into account the possibility that monetary policy is asymmetric, but also that it evolves over time.