اثرات نامتقارن در قانون سیاست های پولی لهستانی

In this paper we investigate whether the reaction function of the National Bank of Poland (NBP) is asymmetric according to the level of inflation gap and the level of output gap. Moreover, we test whether these asymmetries might possibly stem from nonlinearities in the Phillips curve. Threshold models are applied and two cases of unknown and known threshold values are investigated. Our results show that the Polish central bank responds more strongly to the level inflation when the level of inflation is relatively high. We find very weak evidence that the level of inflation reacts more strongly to the output gap when the output gap is relatively high. Thus, the asymmetries in the monetary policy rule seem to indicate asymmetric preferences of the central bank.

The aim of this paper is to search for asymmetric effects in the reaction function of the National Bank of Poland (NBP). We check whether the Polish monetary policy rule is asymmetric concerning the levels of the fundamental macroeconomic variables: inflation and output gaps. Encompassing the asymmetric elements in the reaction function might give a better explanation of the central bank's behavior. This, in turn, could help to form better expectations and forecasts and could be used to build more accurate econometric models of the economy. Positive and negative deviations of inflation and output from the reference levels may be treated by monetary authorities differently. On the one hand, central banks may have asymmetric preferences. Some central banks attempt to stabilize output fluctuations accepting inflation being more volatile, it is because they might face some political heat or social pressure. These banks would have greater aversion to recessions than to expansions. Other central banks might be focused on inflation stabilization (e.g. strict inflation targeters) and have greater aversion to high than low inflation because, for instance, they need to build credibility after implementing an inflation targeting strategy. Cukierman and Muscatelli (2008) distinguish between recession avoidance preferences (RAP) and inflation avoidance preferences (IAP). In the former a central bank takes more precautions against negative output gaps, while in the latter more precautions are taken against positive inflation gaps. Such asymmetric preferences lead to nonlinear reaction functions, as the authors show RAP leads to concave Taylor rule while IAP to convex rule in both the inflation and output gaps. On the other hand, central banks might take into account asymmetries in different channels of the monetary transmission process. Most importantly, the aggregate supply curve might be nonlinear. In empirical studies it is often argued that when the output gap is positive it has a positive impact on inflation, while when the output gap is negative it has very small deflationary impact (Baghli et al., 2007, Buchmann, 2009, Laxton et al., 1999 and Pyyhtia, 1999). There are various explanations of this phenomenon, discussed later on, such as for instance nominal wage rigidities, capacity constraints, costly price adjustments, volatility of aggregate demand and supply shocks. Lastly, the uncertainty regarding the NAIRU or the growth rate of productivity may lead to nonlinear interest rate policy, in such a case monetary authorities might need more time and data to make the decision. Therefore central banks might be more aggressive when the output gap reaches a certain threshold and more cautious when the output gap is small. In this paper we follow a single equation approach. The structure of the economy is usually described by two basic equations, one represents the Phillips curve, and one the aggregate demand curve. And the monetary policy is assumed to minimize a loss function. If we assume that a central bank has a quadratic loss function in the inflation and output gaps and minimizes it subject to the linear structure of the economy, then we will obtain a linear reaction function. But the loss function and the structure of the economy may be nonlinear. Thus, when creating the model, there arise problems: how to choose the central bank's loss function, and should the AS and AD curves be forward-looking and/or backward-looking, and, perhaps nonlinear. Concerning the relevant theoretical models, Cukierman and Muscatelli (2008) and Fiodendji (2013) employ nonlinear loss functions making the assumption that the economic structure is linear. While Dolado et al. (2004) and Surico (2007) allow for both a nonlinear loss function and a nonlinear Phillips curve. Each of these studies uses either a fully backward-looking economy structure (Dolado et al., 2004 and Fiodendji, 2013) or a fully forward-looking economy structure (Cukierman and Muscatelli, 2008 and Surico, 2007). Choosing the specific model is difficult and problematic. It requires imposing certain functional forms and making certain simplifications, and may lead to quite complicated formulas for a central bank's reaction function, particularly, for Poland which is a small open economy and is affected by a number of external factors. Moreover, the Polish Phillips curve seems to have both backward and forward-looking components. Therefore, in this study we do not assume any particular structure of the model and estimate single equations. The structure of this paper is as follows. The next section contains a brief review of the literature concerning symmetric Taylor rule and asymmetric effects in both Taylor rule and Phillips curve. 2 and 3 present our empirical strategy and data set. Section 4 reports the empirical results. The last section concludes.

The paper concerns asymmetries in a monetary policy rule, which could appear because of asymmetric preferences of the central bank and because of nonlinearities in the economic system. It might be suspected that monetary authorities are more aggressive to the inflation rate when it is above its target level than when it is below. It also seems probable that monetary authorities have different preferences and react more strongly when the level of economic activity is low than when it is high. The main innovation of the paper lies in analyzing empirically the asymmetric Phillips curve and Taylor rule in Poland. Such an approach allows us to test whether the asymmetries in the Taylor rule might be caused by the asymmetries in the Phillips curve. We estimate a number of models with unknown and known threshold values. When the threshold value is assumed to be unknown we estimate it by minimizing the sum of squared errors from the relevant equation. We consider two different measures of an inflation rate in the Phillips curve as well as two different measures of an inflation target in the Taylor rule. Our preliminary analysis of the Polish Phillips curve suggests that the curve is not asymmetric according to the level of the output gap. We find only very weak evidence, in all cases except one, statistically insignificant, that the rate of inflation is more strongly influenced by the output gap when the output gap is relatively high. The estimations of the asymmetric Taylor rule seem to indicate that the central bank reacts more strongly to the level of inflation when it is relatively high, that might be the result of implementing the inflation targeting strategy and the need to build credibility. The results show that the asymmetric effects in the Taylor rule do not stem from nonlinearities in the Phillips curve but from asymmetric preferences of the central bank.