European Central Bank Policy Transmission in New European Union Member
States
A Local Projections Approach
Ana Popovici
Student number: 11594349
BSc Economics and Business Economics, Specialization Economics
Thesis supervisor Mr. Andras Lengyel
Faculty of Economics and Business
29.06.2020
Abstract
This paper looks into the transmission of eurozone monetary policy shocks in new European Union member states. The impact of the shock is measured on short-term interest rates and stock prices. The inference was conducted on financial fundamentals from the Czech Republic, Hungary, Poland and Romania. The analysis makes use of a Local Projections approach, as developed by Jord`a (2005). The high frequency identification shock series is retrieved from Jaroci´nski & Karadi (2018). Analyzing transmission effects is important for future policy recommendations and better understanding spillover mechanisms. It follows from the results that the effect on short-term interest rates is heterogeneous and insignificant. However, stock prices are significantly affected by monetary shocks, with the effect lasting throughout the proposed horizon. This implies that, despite the high degree of financial integration, non-euro member states might still retain some monetary policy autonomy, with respect to European Central Bank policies.
Statement of Originality
This document is written by Student Ana Popovici who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision and completion of the work, not for the contents.
1
Introduction
Discussions about the adoption of the euro by new European Union member states (NMS) are important in furthering the process of policy and monetary integration. It is clear that entering the common currency zone implies a loss in monetary policy autonomy, thus reducing the country’s ability to counteract adverse economic shocks (Ferreira et al., 2010). However, it is unclear to what extent the NMS, who are not euro member states, can truly follow independent monetary policy, considering the large degree of financial integration at the European Union (EU) level. Therefore, this paper aims to assess the effect of monetary policy spillover from the eurozone to NMS. This identification is important for policy makers, as it allows for a better understanding of the effectiveness of monetary policy, dependent on EU interrelationships. Furthermore, this research adds to the debate on transmission mechanisms of monetary policy shocks.
Recent identifications of exogenous monetary shocks at the eurozone level make this analysis possible. The identification used in this paper follows from the data set constructed by Jaroci´nski & Karadi (2018). They use high-frequency data from European Central Bank (ECB) announcements to construct a distinct time series of monetary policy shocks. The time series will be used to estimate the spillover effect on short-term interest rates and stock prices. Short-short-term interest rates are representative of monetary policy stances in NMS. As such, this allows for interpretations of the independence of money institutions. The stock price effect is representative of financial convergence at the European level and indicative of the how investors alter their expectations. The effect of the shock on both variables was measured using the Local Projections (LP) approach (Jord`a, 2005). This approach was used as the aim of this paper is uncovering an effect, rather than finding evidence for any underlying mechanism. As such, the LP approach is more robust to any misspecification of the model.
The analysis was conducted on four NMS: Czech Republic, Hungary, Poland and Romania. These countries have relatively large economies that allow for independent monetary policy. There was not enough statistical evidence to conclude that short-term interest rates are significantly impacted by eurozone monetary shocks. Overall, NMS seem to follow independent monetary policy. However, there was enough evidence to conclude that domestic stock prices in all countries are significantly affected by euro area policy shocks. These results prove that there is high financial connectivity at the EU level.
The organization of the paper takes on the following structure. A detailed description of the underlying theoretical framework and current literature can be found in Section 2. Section 3 presents an overview of the model and describes both the independent and dependent variables in detail. Section 4 contains the obtained results and a discussion on their validity and implications. Finally, concluding remarks can be found in Section 5.
2
Theoretical Framework
2.1
Monetary Policy Mechanism
The middle of the 20th century marks the movement away from early Keynesian analysis of structural monetary policy towards reduced form models. Whilst a structural monetary policy model assumes spe-cific causation channels through which aggregate output is affected, the reduced form model looks at the correlation between output and movements in money supply. Early Keynesians dismissed the importance of monetary policy through the structural model as there was no empirical evidence linking movements in nominal interest rate and investment spending (Mishkin, 2013). However, monetarists, focusing on reduced form models, found clear links between monetary shifts and aggregate economy (C. Romer & Romer, 1989). They analysed historical evidence, by looking at exogenous monetary shocks, controlling for reverse causality. This paper considers a reduced form identification.
D. Romer (2012) builds on the reduced form model evidence of monetary policy relevance, further underlining the effect of money growth on inflation. He posits that in the short-run, assuming sticky prices, monetary expansions lower nominal short-term interest rates and encourage aggregate output expansion 1. Inversely, monetary contractions increase short-term nominal interest rates. Empirical evidence of this mechanism is clear, highlighting the importance of monetary policy in practice (Friedman, 1972). Therefore, clear monetary policy rules, aimed at maintaining a stable price level in the long-run, have been at the forefront of ECB goals. The ECB has been perceived as a highly transparent institution, which follows independent policy, that is not dictated by political motives (Haan, 2005).
Changes in interest rates, determined by monetary policy, affect movements in the financial markets through investor expectations (D. Romer, 2012). Building on a standard Gordon dividend growth model2, we can see that a monetary contraction is expected to lower stock prices. This happens as the contraction is assumed to increase the returns on bonds relative to stocks, because of the implied higher nominal interest rate. This will prompt investors to require a higher return on investment in equity, which is going to decrease the present value of future dividend payments. Furthermore, as a monetary contraction indicates a possible slowdown in economic growth, it would also imply a smaller rate of growth in dividends.
Due to the reverse causality implied by the reduced form model, and the limitations imposed by the previously mentioned use of historical data, central bank (CB) announcements have become an important instrument in assessing the effect of monetary policy on financial markets (Kearney, 1996). This is because
1Based on the Fisher’s identity, i ≡ r + πe, expected inflation( πe) will increase following the monetary expansion. This
will cause the nominal interest rate to also increase, which will reduce quantity demanded of real balances. This implies that the price level is rising at a faster rate than the money stock. As prices are assumed to be sticky and individual actors do not adjust their real money balances immediately, the effect could be long-lasting
2Equation for the Gordon dividend growth model: P
0=P∞i=1D0·(1+g) i
(1+ke)i , where P0is the stock price, D0is the most recent
CBs are considered to hold important information about the future course of the economy and financial markets. There is vast literature consensus on the long-term impact of monetary policy on the real economy through the channel of CB overnight interest rates (Bernanke & Blinder, 1992).
CBs enforce policy that is directly correlated with the state of the economy. This implies a very large degree of endogeneity, which does not allow for causal interpretation. Exogenous variation needs to be found in order to adequately assess the impact of monetary policy on economic fundamentals. Assuming that, in a short time period after the CB announcement, the only information hitting the market is the one from the announcement, exogenous variation can be implied. This variation, extracted using a high frequency identification approach, can be used to estimate causal impacts. (Beechey & Wright, 2009).
Nakamura & Steinsson (2018) retrieve unexpected variation in the real-interest rates, following Federal Open Market Committee (FOMC) announcements, to construct exogenous monetary shocks. They find that the monetary shocks have important long-term effect on the real interest rate, undermining assumptions of monetary-neutrality. Another important finding of the paper is that FOMC announcements also appear to affect the private sector’s view on non-monetary economic fundamentals. Nakamura & Steinsson (2018) argue that this is the case because FOMC is believed to hold important information about the coming state of the economy. The private sector adapts its expectations based on the beliefs that the FOMC holds about future economic performance. Two types of channels following CB announcements can be identified. First, a traditional channel, concerning the standard approach to policy, which is consistent with the mechanism presented in this section until now. In this channel, the economy is affected through the changes in interest rates, and expectations are impacted through monetary fundamentals. Second, an information channel, that adapts the private sector’s expectations and information about the the economic state. This channel has to do with the belief that the CB holds information about the future state of the economy, which was previously unknown. As such, the private sector adapts its expectations based on the perceived information discrepancy. Both of the channels are found to be significant in determining the path of economic aggregates. A parallel analysis can be drawn to ECB announcements.
Jaroci´nski & Karadi (2018) build upon this, identifying systematic deviations from monetary policy to differentiate between monetary policy shocks and information shocks, in order to analyse the long-run response of the economy. A detailed description of their differentiation is presented in Section 3.1. They find that immediate financial market responses, following ECB announcements, are relevant in describing the course of financial and macroeconomic variables in the eurozone. Building on their findings on euro area responses, and the previously mentioned consensus, this paper looks into the extendability of their analysis to non-euro NMS spillovers.
This paper will follow on Jaroci´nski & Karadi (2018) identification of monetary shocks, rather than information shocks. Whilst looking into the effects of the information shock differentiation is an important aspect, the current consensus on monetary policy in the European Union is build on traditional monetary
policy mechanisms. Thus, there is clear framework of spillover effects in NMS through monetary shocks, that can be applied to this analysis.
2.2
Spillover Effects
There is current literature consensus on the transmission of monetary policy shocks from larger to smaller open economies. Studies focusing on United States policy transmission to other developed economies (Kim, 2001) found clear positive spillovers. The degree of financial integration between countries has been shown to play a significant part on the extent of the spillover. The more integrated the countries are, the larger the impact of monetary policy shocks from one to another will be. Extending this analysis to the effects of US shocks on the emerging economies of Latin America, we can see a similar inference. Interest rate fluctuations in the emerging economies seem to be significantly impacted by US’s monetary shocks. The effect appears to be immediate, as there is a peak in impact in just 2 quarters after the initial shock (Canova, 2005). As the EU countries are more integrated than the American economies, and there are large financial and trade flows, it is expected that the transmission will be even more significant. Dewachter et al. (2012) found that European economies are highly sensitive to monetary policy changes in neighbouring countries. As such, it is expected that eurozone monetary shocks will be transmitted across NMS.
In 2004 the European Union underwent its biggest enlargement to date, with 10 NMS joining (EU, 2019). This was followed by the ascension of two other states, Romania and Bulgaria, in 2007. Lastly, Croatia joined in 2013, being the last integrated member. Out of the NMS, the following are still not in the eurozone: Bulgaria, Croatia, Czech Republic, Hungary, Poland and Romania. Building on neo-functionalist3 theories of European integration, monetary policy spillover effects are common for newly integrated member states (Cini & Perez-Solorzano Borragan, 2016). As such, once integrated in the common market, member states are more likely to follow the monetary policy practices of the ECB. According to this line of thought, monetary shocks in the eurozone should have an effect on the development of the economy in NMS. Orlowski (2005) found that one of the most common monetary convergence practices is direct interest rate targeting (DIT)4. This practice is primarily conducted by countries with a large economy and independent financial institutions. The largest four economies, according to nominal GDP measurements, of non-euro NMS are: Poland, Czech Republic, Hungary and Romania. Thus, those states are expected to have a high degree of monetary policy spillover effects following eurozone monetary shock.
Previous studies, assessing the extent of spillover effects, found that the variation in interest rates in NMS is significantly determined by eurozone monetary policy shocks (Fladung, 2007). However, at the
short-3Neo-functionalism is a theory of European integration that focuses on the role of supranational institutions to drive the
integration process. Spillovers are a focal point in such theories, as cooperation in one aspect of public life at the European level, leads to more cooperation in other related aspects.
4Domestic CBs in NMS alter inflation expectations such that interest rates follow a similar target to the one proposed by
run end of the term structure of the interest rates, this variation was less significant that in the long-run. This mechanism is also highlighted by Kearns et al. (2018). In their identification, they evaluate possible mechanisms of monetary shock spillovers. They find that the degree of financial openness between bilateral economies is the most significant determinant of the extent of the spillover. Moreover, they also find that exchange rate regimes could also play a significant role. If a country is to maintain a credible peg to the euro, interest rate convergence is more likely. As the EU economies are highly financially integrated, there is sufficient evidence to assume considerable spillover effect on NMS from ECB policies. However, it is likely that the transmission will be more visible in financial markets.
2.3
Anticipated Results and Hypotheses
Considering that monetary policy in the eurozone has a clear impact on the economy of the NMS, it is expected that a monetary shock in the eurozone will affect key financial variables in NMS. In line with the monetary policy mechanism discussed in Section 2.1, a monetary policy contraction in the eurozone is expected to increase short-term interest rates and decrease stock prices. This result is indeed confirmed by Jaroci´nski & Karadi (2018). If the NMS are indeed significantly affected by monetary policy shocks in the eurozone, we would expect interest rates and stock prices to follow a similar trend. As such, the following alternative hypotheses are proposed:
On short-term interest rate:
H0: The short-term interest rates in NMS are not significantly impacted by eurozone monetary policy shocks
HA: The short-term interest rates in NMS are significantly impacted by eurozone monetary policy
shocks
On stock prices:
H0: Stock prices in NMS are not significantly impacted by eurozone monetary policy shocks
HA: Stock prices in NMS are significantly impacted by eurozone monetary policy shocks
This study focuses on: Poland, Czech Republic, Hungary and Romania, countries which are expected to follow a DIT (Orlowski, 2005). As DIT implies a high degree of monetary policy convergence, it is likely that the chosen countries will react as expected to the monetary policy shocks. Thus, the chosen economies will show similar impulse responses on the chosen financial variables. As financial indicators are highly sensitive to monetary policy shocks, this paper analyses the impact of eurozone monetary shocks on short-term interest rates and stock price indices in NMS. The short-short-term interest rates in NMS are expected to
increase following a eurozone monetary policy contraction. On the other hand, the stock price indices of NMS are expected to decrease following the contraction. This effects are in line with the ones shown by Jaroci´nski & Karadi (2018). A detailed operationalization of these variables is presented in Section 3.1.
Considering the results obtained by Fladung (2007), it is expected that the financial variables will react in a homogeneous way to the monetary policy shocks. This reaction should follow the same trend as for the euro area. However, because of the evidence that short-term interest rates are less affected than long-term rates, it is possible that the interest rate analysis will not yield significant results. However, this outcome will depend on the different financial system fundamentals for each NMS. For the stock market analysis, it is expected that the null-hypothesis will be rejected. This expectation builds on the financial integration spillover mechanisms by Kearns et al. (2018), and the high financial integration policies in the EU.
3
Model and Data
3.1
Data and Variables
To ensure the reliability of the data, all financial indicators were obtained from recognized sources. Table 2 in Appendix A presents an overview of the dependent variables used in the analysis: short-term interest rate and stock price indices, for each individual country.
It is important to note that out of the 4 NMS, Romania has a different adherence date, joining the European Union in 2007. As such the timeline proposed for Romania differs slightly from the timeline considered for the other countries. Whilst for Czech Republic, Poland and Hungary the timelines extend from January 1999 until December 2016, the timeline for Romania extends from January 2006 until December 2016. This shorter timeline is justifiable due to the later entry date of Romania. Considering the proposed horizon in the model5this should not pose any significant problems.
The short-term interest rates are computed as the 3-month money market interest rates. The monthly averages are used for the inference. The stock prices indices are representative of the overall performance of publicly listed domestic companies in each NMS. For the stock price index, the end of month closing value is used. Initially, the values were retrieved in the original currency of each member state. In the analysis, the logarithmic values of the closing price are considered, in order to compute the percentage point changes generated by the monetary shocks. As the analysis of the impulse response functions was done over an extended period, using monthly data coincides with the inference at hand. Figures 9 and 10 in Appendix B present an overview of the available data for interest rates and stock prices, respectively. From those figures it can be observed that stock prices, in all NMS, follow a similar trend over the proposed timeline. This is intuitive in the context of financial openness between EU member states. Variation in the short-term
interest rates is considerably larger between the states. However, this variation seems to decrease following the 2008 Financial Crisis, all countries converging to the Zero Lower Bound. This observation could be an indication of spillover of expansionary monetary policy from the ECB, but it could also be a reflection of slowed down domestic economic growth. This is to be determined in the following sections.
The monetary shock variable, that is used as the dependent variable, was retrieved from the Jaroci´nski & Karadi (2018) data set. They have found that the immediate response of the stock market to ECB policy announcements is representative of the long-run economic outlook. The data set they provide offers a differ-entiation between monetary policy shocks and information shocks following monthly ECB announcements. This differentiation was obtained using standard high frequency differentiation, by looking at co-movements between interest rates and stock indices in the eurozone. A monetary policy shock component is identified as a negative co-movement between interest rates and stock prices. This negative co-movement is consis-tent with the theory described in Section 2.1. Whilst different identification methods were used in the Jaroci´nski & Karadi (2018) paper, this inference uses their poor man’s sign restriction identification6. This was computed by looking at the surprises in the 3-month Eonia swap and the Euro Stoxx 50, in a very short period window following the monthly ECB announcements. The poor man’s sign restriction identification was chosen for this analysis as it showed that the monetary shocks have a significant impact on the 1-year government bond yields and on the long-run economic outlook of the eurozone. This significant impact allows this paper to further look into eurozone monetary shock effects on the NMS. If the impact was not significant in the eurozone, the proposed causation chain in Section 2.2 would not be applicable. A more in depth look into the impulse responses that the monetary shocks yield in the eurozone, as found by Jaroci´nski & Karadi (2018), is presented in Section 3.2. Figure 11 in Appendix B presents an overview of the monthly monetary shocks, using the proposed data. As retrieved form the data set, the money shock variable is not standardized. Standardization of the shock variable is required for a facilitated interpretation of the effect. As such, data has been standardized to have 0 mean and 1 standard deviation. Table 1 presents an overview of the summary statistics for the monetary shock, prior to the standardization. Table 3 in Appendix A presents an overview of the standardized shock. The small magnitude of the identified monetary shocks further justifies the choice of looking at spillover effects on interest rates and stock prices. Because of the small magnitude, effects on large macroeconomic aggregates, such as output, are hard to identify (Nakamura & Steinsson, 2018).
6This identification was implemented by Jaroci´nski & Karadi (2018) as a robustness check. It assumes that each month the
Table 1: Monetary Shock Summary Statistics Not Standardized
Mean Standard Deviation Minimum Maximum Monetary Policy Shock 0.00 0.03 -0.17 0.16 Observations 214
Note: The data is given in basis points. It represents the contribution of the shock to the surprise in 3-month Eonia swap. The minimum shock occurred in 2001 and the maximum in 2009.
3.2
Impulse Response Functions
In order to analyze the responses of the proposed financial variables, following the monetary policy shock, an impulse response approach will be considered. This method has been widely used since the 1980s in order to assess the impact monetary policy has on macroeconomic aggregates (Inoue & Kilian, 2013). As such, an impulse response function, in this context, follows the response of the economic system of interest to an external monetary policy shock. In the case of this paper, the systems of interest are interest rate and stock price indices fluctuations in NMS reacting to monetary policy shocks in the eurozone.
Jaroci´nski & Karadi (2018) used impulse responses to analyze euro area economic outlooks following the shocks, as constructed in their series. The impulse responses were constructed using different variations of vector autoregressive (VAR) models. They found that, following a contractionary monetary shock in the eurozone, short-term interest rates increase and stock prices decrease. By using the poor man’s sign restriction identification, eurozone interest rates and stock prices seem to be clearly impacted after for period of 1.5 to 2 years. The same inference will be applied in this paper to observe if the same effect can be induced in NMS. The model proposed for this analysis is presented below.
3.3
Local Projections and Proposed Model
As previously mentioned, Jaroci´nski & Karadi (2018) used a VAR model to determine the impulse responses following isolated monetary shocks. However, there are also other methods that can be used for estimating impulse responses following a discretely observed shock series (Choi & Chudik, 2019). An important consid-eration when building models is the inherent trade-off between bias and estimation. Whilst bias has more to do with the specification of the model, estimation is related to the specificity of the given results (Stock & Watson, 2015). VAR models have been at the forefront of recent analyses that estimate impulse response functions (IRF). However, other alternatives have also been developed that could, in certain inferences, be more adequate. One such example is the Local Projection (LP) model, as proposed by Jord`a (2005).
lagged values of all k series (Stock & Watson, 2015). This implies, that by using a VAR model, an extension of the univariate autoregression over several time-series is possible. As such, it becomes an important instrument in determining the course of economic variables, and has been widely employed. This standard approach is expected to give robust and reliable results if the VAR model is known to be correctly specified. A problem of using such a model, in this monetary transmission scenario, is that not all the channels of spillover are yet identified. This statement builds on the mentioned reduced form approach from Section 2.1. As such, it is improbable to consider an adequately specified model if one builds on this assumption. The LP approach, considered by Jord`a (2005), partially solves this, and allows for an adequate assessment of the resulting IRF.
The idea underlying the LP is that there are sequential regressions of the model, in which the depen-dent variable is shifted several time periods ahead. This sequential regressions can give robust estimates of the regression coefficient even in the absence of a clear specification of the true multivariate system (Jord`a, 2005). As such, this model becomes adequate for estimating effects of reduced form models. Considering that this paper looks at spillover effects of eurozone monetary policy on NMS, and that the transmission system is yet not fully specified, using a LP model for the IRF can be considered adequate. However, in this context, the use of this LP model is likely going to not have very specific estimates. As such, it is expected that the coefficients will have high standard errors. As the purpose of the paper is the analysis of the trend rather than getting an exact magnitude, this problem can be considered secondary. As such, a trade-off in favour of reducing bias was adopted.
Thus, the proposed LP model for interest rates and stock price indices takes the following form:
ij,t+h− ij,t−1= αij,h+ γhmt+ υj,hi for h = 1, 2, ..., 20 (1)
IRi(h) = γh
sj,t+h− sj,t−1= αsj,h+ βhmt+ υj,hs for h = 1, 2, ..., 20 (2)
IRs(h) = βh
Equation (1) represents the proposed regression and IRF for estimating the spillover effect on the short-term interest rates Thus, the difference between ij,t+hand ij,t−1represents the short-term interest rate
change at each h horizon. The short-term interest rate is measured as discussed in Section 3.1. Equation (2) does the same, but for estimating the stock price effect. In this case, the difference between sj,t+h and
sj,t−1 is representative of the stock price change. The stock price measurement and operationalization is
also discussed in Section 3.1. In both equations, the variable mt represents the eurozone monetary shock
at the time t, as retrieved from the Jaroci´nski & Karadi (2018) data set. As the analysis is considered for 4 different NMS, the j index takes in turn the nominal values for Czech Republic, Hungary, Poland and Romania. The estimations were run individually, for each specified country.
The parameters of interest that the model aims to estimate are: γh and βh. As such, the IRF of
short-term interest rates, h periods ahead, after a 1 standard deviation increase in mt, can be written as
IRi(h) = γh, where γh is the estimation obtained from the regression in Equation (1). Respectively, the
IRF of stock price effect, h periods ahead, after 1 standard deviation increase in mt, can be written as
IRi(h) = βh. βh is obtained from the regression in Equation (2). αij,h and αsj,h are intercept terms for the
interest rate and stock prices regressions. Respectively, υi
j,h and υj,hs are indicative of the error terms.
The forecast horizon, h for the LP is taken as 20 periods, with lags of 1. As the data consists of monthly averages, this implies that the IRFs will be calculated over a period of 20 months. This time horizon is adequate as the dependent variables have high short-term variability. An horizon of 20 months can already be considered an extended analysis. Furthermore, analysis by Kearns et al. (2018) on spillover effects considered a similar time horizon.
An important advantage of the LP model is that is allows for simple least square estimation. However, because of the rather short time horizon and the consideration of one months lags, it is likely that there is going to be very high auto-correlation (Stock & Watson, 2015). Previous papers that used LP over similar forecast horizons solved his problem by using a least squares dummy variable (LSDV) estimator with the Newey–West covariance matrix. This estimation uses robust standard errors that are corrected for heteroskedasticity and auto-correlation. The following section looks at the results obtained from the specified model.
4
Results and Discussion
This section presents an overview of the results obtained with the LP model, proposed in Section 3.3. In line with the first hypothesis, regarding transmission on short-term domestic interest rates, there is not enough evidence to reject the null hypothesis. Only interest rates in Romania appear to be significantly impacted by eurozone monetary policy shocks. Regarding the impact of the shock on NMS stock prices, the results appear to be homogeneous and significant. As such, there was enough evidence in favour of rejecting the second null hypothesis. A detailed discussion on each independent variable and country will follow. All regression coefficients are presented in Tables 5 to 12 in Appendix A.
4.1
Interest Rate Effect
The results for the interest effect are generally insignificant and heterogeneous. These results are consistent with the implied complexity of the spillover mechanism within the EU. Some of the variation can be explained by the exchange rate spillover mechanism (Kearns et al., 2018). The insignificant results are consistent with the line of thinking that short-term interest rates are generally less affected by spillovers from ECB monetary shocks because of the degree of independence held by domestic financial institutions (Fladung, 2007). This holds for: Czech Republic, Hungary and Poland. On the other hand, the results for Romania present significant long-term effects even after 20 months. These results will be explained through the exchange rate spillover channel. The IRF for short-term interest rates can be observed in Figures 1 till 4. A country by country analysis will be considered for the discussion.
Figure 1 presents the interest rate IRF for Romania. The observed results imply that 1 standard deviation increase in a eurozone monetary shock will increase short-term interest rates by approximately 0.6pp, 1 year after. This result becomes significant at 10%, 6 months after the initial shock. This implies a high degree of monetary policy convergence. However certain considerations need to be made. According to Kearns et al. (2018), countries that are considering maintaining a peg to the originator economy of the shock are likely to showcase almost one to one co-movement in their interest rates. The Romanian leu is maintained at a managed float regime against the euro. This type of regime is likely to imply that short-term interest rates will be more affected by monetary policy shocks, because the Romanian Central Bank will adjust them to maintain the range of the float (Shambaugh, 2004). This is in line with the argument that managed exchange rate regimes reduce the monetary policy autonomy of the country. However, the results seem to point out that Romania is tampering with the exchange rate more than it would be expected. This result would be consistent with the ’fear of floating’ theory. Emerging countries, that claim to have managed float exchange rate, have a tendency to interfere more than needed on the Forex market (Calvo & Reinhart, 2002). On the other hand, it is important to recall that the data set for Romania was considerably shorter than for the other economies, as it joined the EU at a later date. Romania joined the EU approximately one year before the 2008 Financial Crisis, and such, it is possible that the shorter time frame will have an effect on the results. As it can be seen in Figure 9 in Appendix B, short-term interest rates in all NMS appear to converge after the crisis. Thus, the timeline could have a large effect on the observed results. Overall, Romania seems to be the country with the highest degree of short-term interest rate spillover from the sample at hand.
For Czech Republic, there appears to be no significant spillover of the shock on the short-term interest rates. This can be seen in Figure 2. A slight increase can be observed following month 12 after the shock, however it is insignificant. This could imply that the spillover is done through a free floating exchange rate, rather than the short-term interest rates. During the period 1998-2006, the Czech Republic was following DIT and exchange rate stability was not a policy priority. Kemme & Lyakir (2011) show that during this
period there were almost no significant policy interventions on the Forex market. This period represents a significant portion of the used timeline. It is also noteworthy that between 2015 and 2017 the Czech koruna was pegged to the euro. However, this time period seems to not be reflected in the obtained results. Furthermore, during the transition period, the Czech Republic experienced considerably lower inflation than the other NMS (Belhocine et al., 2016). This could be an indication of a more established and independent monetary policy system. This might be the case, as the Czech Republic is one of the best performing NMS economies.
For Poland, the effect is also insignificant, but seems to follow a negative rather than the predicted positive trend. This does not imply any fault in the theory, as a variety of factors might have impacted the results. Section 4.3 presents an in depth explanation of the limitations of the analysis, which will relate to reasons for which this effect might have been noticed. Additionally, it is important to consider that after the 2008 Financial Crisis, Poland had one of the lowest Calvo-Reinhart “Fear of Floating” indices (Belhocine et al., 2016). Thus, the country might allow the majority of the spillover to enter through the exchange rate channel. By looking at Figure 9 in Appendix B, we can observe that after 2004 the short-term interest rates in Poland and Czech Republic follow the same trend. However, in the brief period in the data set included before the adherence, Poland has considerably higher short-term interest rates, reaching almost 20%. This is indicative of very high pre-adherence inflation. This could be reflected in the IRF results, as the magnitude is rather significant. In the eurozone, the same period had relatively low interest rates, and also presented a series of negative monetary policy shocks. These shocks, at the beginning of 2000, are representative of the burst of the dot-com bubble (Jaroci´nski & Karadi, 2018). Overall, there is no evidence that short-term interest rates in Poland are significantly impacted by monetary policy shocks in the eurozone. The IRF for Poland can be observed in Figure 3.
Lastly, Figure 4 presents the IRF of short-term interest rates in Hungary following an ECB monetary policy shock. Again, the results do not show any significant impact. A slight increase in interest rates can be observed 10-months after the shock, however this is not statistically significant. An explanation for this result is again that Hungary, having a floating exchange rate regime, retains a high degree of independence for monetary policy. The same mechanism as for the other NMS holds.
Overall, the results for short-term interest rates in NMS are country dependent. The insignificant results are consistent with theories stating that short-term interest rates are less affected by monetary policy spillovers. This seems to support previous the findings of Fladung (2007) and Kearns et al. (2018). This happens regardless of the high-degree of integration across NMS. The comparatively more restrictive exchange rate, that Romania has implemented to the euro, could partially provide an explanation for the significant results obtained. These results would be consistent with the mechanism provided my Kearns et al. (2018). However, the later adherence date of Romania should also be considered, as the majority of the data set consisted of post-crisis values and could capture the effect of unconventional monetary policy. This mechanism is not discussed in this analysis.
Figure 1: IR Short-term Interest Rates Romania Figure 2: IR Short-term Interest Rates Czech Republic
Figure 3: IR Short-term Interest Rates Poland Figure 4: IR Short-term Interest Rates Hungary
4.2
Stock Price Effect
The stock price effect is homogeneous across the NMS. Generally, the impact of ECB monetary policy shocks becomes significant in domestic financial markets in the first months following the shock. The impact is in line with the expected monetary shock transmission mechanism assuming high financial integration. A monetary contraction shock is indeed met with a stock price decrease. As Jaroci´nski & Karadi (2018) found a significant impact of the shocks on the S&P500 index, it was expected that a significant effect will be found on NMS stock market indices as well. The inference at hand is supportive of the hypothesis that domestic financial markets in NMS are significantly impacted by eurozone monetary policy shocks. The magnitude of the results is in line with previous studies, which found that a high degree of financial integration allows for significant spillover effects on the stock market (Ehrmann & Fratzscher, 2009). The spillover is in line with the argumentation of Kearns et al. (2018). Figures 5 to 8 present the IRF for each individual NMS. A country by country analysis is considered.
at this time frame, 1 standard deviation increase in the monetary shock will decrease the BET index by 0.075pp. Romania has a high degree of financial integration with the EU. A significant effect was expected (Ehrmann & Fratzscher, 2009). After the first 6 months, the index continues to show a decrease, peaking at a 0.1pp, one year after the shock. The effects then level off, becoming almost insignificant 20 months after. One curious observation is that the stock price trend is less cyclical for Romania than for the other NMS. Again, this observation could be a result of the reduced time data set for Romania, as a result of the later adherence date. For the Czech Republic, the magnitude of the effect is comparable with that of Romania. A 1 standard deviation increase in the monetary policy shock causes the PX-50 index to decrease by 0.05pp approximately 12 months after. In this case, the effect is significant even at the maximum calculated horizon. Overall a significant downwards trend can be observed.
For Poland and Hungary the magnitude of the results appears to be more significant in the first months than for the Czech Republic and Romania. However, this changes in later months, when the effects become comparable. As such, in Poland, at almost 1.5 years after the shock, a 1 standard deviation increase in the monetary shock causes a 0.075pp decrease in the WIG20 index. The same can be said for Hungary, in the case of the BUX index. Similar to Czech Republic, Poland and Hungary show a significant trend almost all throughout the proposed horizon. One possible explanation is that the time series for those countries was longer, due to the earlier adherence date.
By observing Figure 10 in Appendix B we can see serious co-movements of the domestic NMS indices. Thus, the homogeneity of the results was to be expected. The small differences in trends can be attributed to domestic market characteristics and time-series inconsistencies. Generally, the results appear as anticipated. The high financial integration at the EU level, prompts domestic stock markets in NMS to significantly react to any policy changes. As such, results in favour of spillover effects through financial integration are found (Kearns et al., 2018). It is important, however, to mention that the magnitude of the results could appear relatively small. The domestic stock markets considered do not have very high trading volumes. They consist of domestic companies from each respective NMS. Whilst those companies have significant domestic importance, the traded volume is significantly lower than for other stock markets that have a larger international base. Nevertheless, the observed trend is still significant and there is clear evidence of eurozone monetary shock spillover.
Figure 5: Impulse Responses BET index Romania Figure 6: Impulse Responses PX-50 index Czech Republic
Figure 7: Impulse Responses WIG20 index Poland Figure 8: Impulse Responses BUX Hungary
4.3
Limitations
Whilst the results are generally in line with previous studies and are overall consistent, there are still several limitations to the proposed inference. Firstly, for this analysis the LP approach was chosen as it reduces biases in incompletely specified models. Inherently to this line of thinking is then, the incompleteness of the reduced form approach that has been used. Whilst this approach is indeed adequate for observing if there are any trends and effects, it gives little insight into the underlying mechanism. As such, any inference from the evidence could be part of discussed spillover transmissions, such as the ones provided by Kearns et al. (2018), but cannot be said to attest to any. This is a common form of criticism for reduced form models in monetary policy. As such, there is a very high degree of uncertainty and also unaddressed non-linearity. Despite the concern being valid, it is important to recall that the objective of this paper was to uncover any effects, rather than explain the mechanism from which they follow. Thus, this limitation calls for future research that can further examine the exact transmission mechanism. A detailed account into the transmission mechanism will allow for a better understanding of policy coordination at the EU level. However, knowing the extent of transmission is the stepping stone for undertaking such research.
Another limitation, directly impacting the observed results, is the reduced time frame and the high auto-correlation that results from it. Firstly, the data set is dependent on the still young implementation of the euro. There was no common structure before 1999 from which monetary policy shocks could be derived. Furthermore, all the NMS presented had an economically tumultuous transition period following the fall of communism, and then, the adherence to the block. As such, to avoid the impact of the transition periods, a data set starting at later date could have been used. As the monetary shock series stops in 2016, this would have been a too short a series to draw any clear statistical inference. Furthermore, it would have not been a reflection of the full transition period of the NMS. Additionally, during the proposed timeline there have been clear changes in exchange rate regime across the sample. The presented IRFs might, thus, not capture the true effect of the transmission mechanisms. Inference on reduced time frames could be valuable in providing more insight into how the transmission mechanism changes over time.
The analysis from the short-term interest rate implies that 3 out of 4 NMS have a rather independent monetary policy. This finding is in line with the macroeconomic trilemma. Thus, the three countries that have a free floating exchange rate and free capital flows with the EU, do not have their interest rates affected by eurozone monetary policy shocks. This is because they can implement independent monetary policy. However, recent research such as the Edwards (2015) paper, questions the empirical applicability of independent monetary policy under free-floating exchange rate regimes. The paper analysis monetary spillovers from the United States of America into Latin American economies that have floating exchange rate regimes. The author finds high convergence between these economies and the USA economy. This could also be the case at the EU level with the NMS. Thus, it is important to further investigate the results. Edwards (2015) excludes the period following the start of quantitative easing, whereas the analysis at hand does not. Future research that takes into consideration capital flows would be useful in determining the transmission mechanism of eurozone monetary policy to NMS, following the Edwards (2015) framework. However, an extended time series should be applied.
Additionally, unconventional monetary policies following the 2008 Financial Crisis could also have an effect on the results. Chen et al. (2014) found that, following the US asset purchases programs, monetary policy had larger spillovers per unit of surprise. This appears be the result of the change in structural factors, pertaining to the new instruments. As such, this could be a justification for the significant results obtained in Romania due to the reduced timeline. However, further analysis is needed in order to identify if this mechanism was also at play within the European Union.
Overall, the aim of the analysis was do determine weather there are significant spillover effects on short-term interest rates and stock prices in NMS following eurozone monetary policy shocks. As such, the transmission mechanisms of the shocks were not tested. The obtained results remain relevant for future research analyzing expected transmission mechanisms. The results do show that there is significant spillover of eurozone monetary shocks on NMS equity markets. On the other hand, the effect on short-term interest rates is generally not significant. Those results are consistent with the predictions and previous literature.
Any trade-offs undertaken in the model are consistent with the aim of observing the effect, rather than determining its causes.
5
Conclusion
This paper aimed to assess the effects of spillovers from eurozone monetary policy shocks in new EU member states. This was done by employing the Jaroci´nski & Karadi (2018) identification of monetary shocks following European Central Bank announcements. The impact on short-term interest rates and domestic stock market, following the shock, was identified using a Local Projections approach.
Overall, short-term interest rates in new member states are heterogeneously affected by euro area monetary shocks. Only Romania presents a significant spillover effect on the short-term interest rates. For the other member states, there was no significant spillover effect. These results are justified by differences in exchange rate adjustments and financial systems. The results are in line with previous research. For domestic stock prices in NMS, there are significant spillover effects which were homogeneous across all 4 countries. The magnitude of the effect was small, but it persisted a long time after the shock. Those results are in line with underlying financial convergence between European Union member states.
The obtained results can be used to better understand spillover mechanisms at the EU level. As such, research testing specific chains of causation is recommended. A more in depth analysis into the exchange rate mechanisms is set to help individual countries with their domestic policy, but also improve economic coordination. Additionally, NMS countries benefit in their future policy recommendations of grasping the full extent of their independence in monetary policy. Considering the result by Fladung (2007), additional research carried on long-term interest rates should also be undertaken. Whilst a higher degree of monetary convergence on long-term interest rates does not imply complete loss of monetary policy autonomy, it is still likely to affect the long-run economic outlook of NMS. From the financial market perspective, the results suggest that there is high degree of financial convergence at the European level. Future research looking into how information shocks propagate differently from monetary shocks could also bring more insight into the monetary policy transmission mechanism on stock markets. Lastly, the inference at hand adds insight into the existence of transmission mechanisms, strengthening results obtained in previous literature, with a different model approach. This lays a broad foundation for future studies that will help policy makers in adopting concrete action plans.
References
Barro, R. J. (1989). Interest-rate targeting. Journal of Monetary Economics, 23 (1), 3-30.
Beechey, M. J., & Wright, J. H. (2009). The high-frequency impact of news on long-term yields and forward rates: Is it real? Journal of Monetary Economics, 56 (4), 535-544.
Belhocine, N., Crivelli, E., Geng, M. N., & Wiegand, M. J. (2016). Taking stock of monetary and exchange
rate regimes in emerging europe. International Monetary Fund.
Bernanke, B. S., & Blinder, A. S. (1992). The federal funds rate and the channels of monetary transmission.
The American Economic Review, 82 (4), 901-921.
Calvo, G. A., & Reinhart, C. M. (2002). Fear of floating.(statistical data included). Quarterly Journal of
Economics, 117 (2), 379-408.
Canova, F. (2005). The transmission of us shocks to latin america. Journal of Applied Econometrics, 20 (2), 229-251.
Chen, J., Mancini-Griffoli, T., & Sahay, R. (2014). Spillovers from united states monetary policy on emerging markets: Different this time? IMF Working Papers.
Choi, C.-Y., & Chudik, A. (2019). Estimating impulse response functions when the shock series is observed.
Economics Letters, 180 , 71-75.
Cini, M., & Perez-Solorzano Borragan, N. (2016). European union politics (Fifth edition. ed.). Oxford: Oxford University Press.
Dewachter, H., Houssa, R., & Toffano, P. (2012). Spatial propagation of macroeconomic shocks in europe.
Review of World Economics, 148 (2), 377-402.
Edwards, S. (2015). Monetary policy independence under flexible exchange rates: An illusion? World Economy, 38 (5), 773-787.
Ehrmann, M., & Fratzscher, M. (2009). Global financial transmission of monetary policy shocks. Oxford
Bulletin of Economics and Statistics, 71 (6), 739-759.
EU. (2019). About the EU - further expansion. Retrieved 2010-09-30, from http://web.archive.org/web/ 20080207010024/http://www.808multimedia.com/winnt/kernel.htm
Ferreira, P., Dion´ısio, A., & Pires, C. (2010). Adopt the euro? the gme approach. Journal of Economic
Interaction and Coordination, 5 (2).
Fladung, M. (2007). Spill-over effects of monetary policy: a progress report on interest rate convergence in
Friedman, M. (1972). Monetary policy. Proceedings of the American Philosophical Society, 116 (3), 183-196. Haan, J. d. (2005). The european central bank : credibility, transparency, and centralization. Cambridge,
Mass: MIT Press.
Inoue, A., & Kilian, L. (2013). Inference on impulse response functions in structural var models. Journal of
Econometrics, 177 (1), 1-13.
Jaroci´nski, M., & Karadi, P. (2018). Deconstructing monetary policy surprises the role of information shocks. Frankfurt am Main: European Central Bank.
Jord`a, (2005). Estimation and inference of impulse responses by local projections. The American economic
review, 95 (1), 161-182.
Kearney, A. A. (1996). The effect of changing monetary policy regimes on stock prices. Journal of
Macroe-conomics, 18 (3), 429-447.
Kearns, J., Schrimpf, A., & Xia, F. D. (2018). Explaining monetary spillovers: the matrix reloaded. BIS Working Paper.
Kemme, D. M., & Lyakir, G. (2011). From peg to float: Exchange market pressure and monetary policy in the czech republic. Review of International Economics, 19 (1), 93-108.
Kim, S. (2001). International transmission of us monetary policy shocks: Evidence from var’s. Journal of
Monetary Economics, 48 (2), 339–372.
Mishkin, F. S. (2013). The economics of money, banking and financial markets (European ed.). Harlow, England: Pearson.
Nakamura, E., & Steinsson, J. (2018). High-frequency identification of monetary non-neutrality: The information effect. The Quarterly Journal of Economics, 133 (3), 1283-1330.
Orlowski, L. T. (2005). Monetary convergence of the eu accession countries to the eurozone: A theoretical framework and policy implications. Journal of Banking and Finance, 29 (1), 203-225.
Romer, C., & Romer, D. (1989). Does monetary policy matter? a new test in the spirit of friedman and schwartz. NBER Macroeconomics Annual, 4 , 121-170.
Romer, D. (2012). Advanced macroeconomics (Fourth edition. ed.). New York: McGraw-Hill/Irwin. Shambaugh, J. C. (2004). The effect of fixed exchange rates on monetary policy. The Quarterly Journal of
Economics, 119 (1), 301-352.
Stock, J. H., & Watson, M. W. (2015). Introduction to econometrics (Updated third edition, Global edition. ed.). Boston: Pearson.
Appendix A
This section of the Appendix presents descriptive and regression tables from the proposed inference.
Table 2: Dependent Variables Data Sources
Country Variable Data Source
Czech Republic Short-term interest rate OECD database PX-50 index Yahoo Finance
Hungary
Short-term interest rate OECD database BUX index Yahoo Finance
Poland Short-term interest rate OECD database
WIG20 index Warsaw Stock Exchange- historical data
Romania Short-term interest rate European Central Bank database
BET index Bucharest Stock Exchange- historical data
Note: Data sources are presented from each dependent variable. All data files and work files are saved for each country, and are available upon request.
Table 3: Monetary Shock Summary Statistics Standardized
Mean Standard Deviation Minimum Maximum Monetary Policy Shock 0.00 1.00 -5.91 5.49 Observations 214
Note: The data is presented basis points. The summary statistics are representative of the standardized shock series from Jaroci´nski & Karadi (2018). The standardization was performed in the following manner: mt,s= mtσ−µ. This was done in order to facilitate the interpretation
Table 4: Dependent Variables Summary Statistics
Mean Standard Deviation Minimum Maximum
Hungary
Short term interest rate 7.45 3.76 0.66 15.80 BUX closing price 16714.88 6819.51 5489.64 32003.05
Romania
Short term interest rate 5.82 3.83 0.51 18.21 BET closing price 6024.51 1907.36 1899.14 10262.82
Czech Republic
Short term interest rate 2.45 1.89 0.28 8.49 PX-50 closing price 955.46 384.22 331.90 1908.30
Poland
Short term interest rate 6.51 4.88 1.65 19.82 WIG20 closing price 2146.06 632.86 1022.62 3877.62 Observations 216
Note: The table presents an overview of the summary statistics for the dependent variables. The short term interest rates are given in % terms and are defined as stated in section 3.1. The stock price indices are given in the domestic currency of each NMS. This was accounted for in the inference as stated in Section 3.1.
Table 5: Regression Coefficients Romania Short Term Interest Rate Effect
Month 1 Month 2 Month 3 Month 4 Month 5 Month 6 Month 7 Month 8 Monetary Shock ECB -0.11 0.09 0.15 0.22 0.27 0.34∗ 0.31∗ 0.34
(0.11) (0.10) (0.13) (0.16) (0.17) (0.19) (0.18) (0.23)
Observations 131 130 129 128 127 126 125 124
Month 9 Month 10 Month 11 Month 12 Month 13 Month 14 Month 15 Monetary Shock ECB 0.50∗∗ 0.54∗ 0.63∗∗ 0.62∗∗ 0.62∗ 0.64∗ 0.56
(0.23) (0.28) (0.28) (0.29) (0.33) (0.36) (0.40)
Observations 123 122 121 120 119 118 117
Month 16 Month 17 Month 18 Month 19 Month 20 Monetary Shock ECB 0.64∗ 0.69∗ 0.71∗∗ 0.65∗ 0.66∗
(0.38) (0.36) (0.32) (0.35) (0.34)
Observations 116 115 114 113 112
Standard errors in parentheses
∗p < 0.10,∗∗ p < 0.05,∗∗∗p < 0.01
Note: The table presents the regression coefficients for Romania, as presented by γhin Equation 1 The reaction to the shock
Table 6: Regression Coefficients Romania Stock Price Effect
Month 1 Month 2 Month 3 Month 4 Month 5 Month 6 Month 7 Month 8 Monetary Shock ECB 0.00 -0.01 -0.03 -0.04 -0.04 -0.07∗∗ -0.06∗ -0.06∗
(0.01) (0.01) (0.02) (0.03) (0.03) (0.03) (0.03) (0.03)
Observations 127 125 123 122 121 120 119 118
Month 9 Month 10 Month 11 Month 12 Month 13 Month 14 Month 15 Monetary Shock ECB -0.07∗ -0.07∗∗ -0.09∗∗ -0.09∗∗ -0.08∗ -0.08∗ -0.08∗
(0.04) (0.04) (0.04) (0.04) (0.04) (0.04) (0.05)
Observations 117 116 115 114 113 112 111
Month 16 Month 17 Month 18 Month 19 Month 20 Monetary Shock ECB -0.08∗ -0.08∗ -0.07∗ -0.07 -0.07
(0.05) (0.04) (0.04) (0.04) (0.04)
Observations 111 111 111 110 109
Standard errors in parentheses
∗p < 0.10,∗∗ p < 0.05,∗∗∗p < 0.01
Note: The table presents the regression coefficients for Romania, as presented by βhin Equation 2. The reaction to the shock
Table 7: Regression Coefficients Czech Republic Short Term Interest Rate Effect
Month 1 Month 2 Month 3 Month 4 Month 5 Month 6 Month 7 Month 8 Monetary Shock EU -0.00 0.00 0.01 0.01 0.02 0.01 0.01 0.01
(0.01) (0.02) (0.02) (0.02) (0.02) (0.02) (0.03) (0.03)
Observations 214 214 213 212 211 210 209 208
Month 9 Month 10 Month 11 Month 12 Month 13 Month 14 Month 15 Monetary Shock EU 0.01 0.01 0.01 0.03 0.05 0.06 0.01
(0.04) (0.04) (0.04) (0.05) (0.06) (0.06) (0.07)
Observations 207 206 205 204 203 202 201
Month 16 Month 17 Month 18 Month 19 Month 20 Monetary Shock EU -0.00 -0.01 0.07 0.09 0.06
(0.08) (0.09) (0.09) (0.10) (0.10)
Observations 200 199 198 197 196
Standard errors in parentheses
∗p < 0.10,∗∗ p < 0.05,∗∗∗ p < 0.01
Note: The table presents the regression coefficients for Czech Republic, as presented by γhin Equation 1. The reaction to the
Table 8: Regression Coefficients Czech Republic Stock Price Effect
Month 1 Month 2 Month 3 Month 4 Month 5 Month 6 Month 7 Month 8 Monetary Shock EU -0.00 -0.01∗∗ -0.02 -0.02 -0.02 -0.03∗ -0.02 -0.03∗
(0.00) (0.01) (0.01) (0.01) (0.01) (0.02) (0.02) (0.02)
Observations 214 214 213 212 211 210 209 208
Month 9 Month 10 Month 11 Month 12 Month 13 Month 14 Month 15 Monetary Shock EU -0.03∗ -0.03∗ -0.04∗∗ -0.04∗∗ -0.04∗ -0.04 -0.03
(0.02) (0.02) (0.02) (0.02) (0.02) (0.02) (0.03)
Observations 207 206 205 204 203 202 201
Month 16 Month 17 Month 18 Month 19 Month 20 Monetary Shock EU -0.03 -0.04∗ -0.05∗∗ -0.05∗ -0.05∗
(0.02) (0.02) (0.02) (0.02) (0.03)
Observations 200 199 198 197 196
Standard errors in parentheses
∗p < 0.10,∗∗ p < 0.05,∗∗∗ p < 0.01
Note: The table presents the regression coefficients for Czech Republic, as presented by βhin Equation 2. The reaction to the
Table 9: Regression Coefficients Poland Short-Term Interest Rate Effect
Month 1 Month 2 Month 3 Month 4 Month 5 Month 6 Month 7 Month 8 Monetary Shock EU -0.06∗∗ -0.10∗ -0.06 -0.09 -0.10 -0.16 -0.21 -0.17
(0.03) (0.06) (0.05) (0.06) (0.08) (0.10) (0.14) (0.13)
Observations 214 214 213 212 211 210 209 208
Month 9 Month 10 Month 11 Month 12 Month 13 Month 14 Month 15 Monetary Shock EU -0.13 -0.12 -0.12 -0.14 -0.14 -0.11 -0.04
(0.13) (0.14) (0.15) (0.17) (0.18) (0.19) (0.19)
Observations 207 206 205 204 203 202 201
Month 16 Month 17 Month 18 Month 19 Month 20 Monetary Shock EU -0.08 0.04 -0.17 -0.21 -0.11
(0.21) (0.20) (0.27) (0.30) (0.27)
Observations 200 199 198 197 196
Standard errors in parentheses
∗p < 0.10,∗∗ p < 0.05,∗∗∗ p < 0.01
Note: The table presents the regression coefficients for Poland, as presented by γhin Equation 1. The reaction to the shock is
Table 10: Regression Coefficients Poland Stock Price Effect
Month 1 Month 2 Month 3 Month 4 Month 5 Month 6 Month 7 Month 8 Monetary Shock EU -0.00 -0.02∗∗∗ -0.03∗∗∗ -0.02∗ -0.02∗∗ -0.04∗∗ -0.03∗∗ -0.04∗
(0.00) (0.01) (0.01) (0.01) (0.01) (0.02) (0.02) (0.02)
Observations 214 214 213 212 211 210 209 208
Month 9 Month 10 Month 11 Month 12 Month 13 Month 14 Month 15 Monetary Shock EU -0.04∗ -0.05∗∗∗ -0.06∗∗∗ -0.06∗∗∗ -0.06∗∗ -0.05∗ -0.04
(0.02) (0.02) (0.02) (0.02) (0.02) (0.03) (0.03)
Observations 207 206 205 204 203 202 201
Month 16 Month 17 Month 18 Month 19 Month 20 Monetary Shock EU -0.04 -0.05∗ -0.06∗∗∗ -0.07∗∗∗ -0.06∗∗
(0.03) (0.03) (0.02) (0.02) (0.03)
Observations 200 199 198 197 196
Standard errors in parentheses
∗p < 0.10,∗∗p < 0.05,∗∗∗p < 0.01
Note: The table presents the regression coefficients for Poland, as presented by βhin Equation 2. The reaction to the shock is
Table 11: Regression Coefficients Hungary Short Term Interest Rate Effect
Month 1 Month 2 Month 3 Month 4 Month 5 Month 6 Month 7 Month 8 Monetary Shock EU 0.07 0.09 0.12 0.09 0.09 0.11 0.08 0.11
(0.07) (0.08) (0.09) (0.09) (0.10) (0.11) (0.12) (0.13)
Observations 168 166 173 162 164 164 156 161
Month 9 Month 10 Month 11 Month 12 Month 13 Month 14 Month 15 Monetary Shock EU 0.03 0.14 0.19 0.18 0.16 0.19 0.11
(0.12) (0.16) (0.16) (0.16) (0.18) (0.17) (0.17)
Observations 160 153 156 156 150 151 150
Month 16 Month 17 Month 18 Month 19 Month 20 Monetary Shock EU 0.09 0.16 0.07 0.14 0.17
(0.19) (0.18) (0.15) (0.18) (0.17)
Observations 146 148 146 144 146
Standard errors in parentheses
∗p < 0.10,∗∗ p < 0.05,∗∗∗ p < 0.01
Note: The table presents the regression coefficients for Hungary, as presented by γhin Equation 1. The reaction to the shock
Table 12: Regression Coefficients Hungary Stock Price Effect
Month 1 Month 2 Month 3 Month 4 Month 5 Month 6 Month 7 Month 8 Monetary Shock EU -0.00 -0.02∗∗ -0.02∗∗ -0.02 -0.03∗∗ -0.04∗∗∗ -0.03 -0.03
(0.00) (0.01) (0.01) (0.02) (0.01) (0.02) (0.02) (0.02)
Observations 214 214 213 212 211 210 209 208
Month 9 Month 10 Month 11 Month 12 Month 13 Month 14 Month 15 Monetary Shock EU -0.04∗ -0.04∗∗ -0.05∗∗ -0.05∗∗ -0.05∗∗ -0.04 -0.04
(0.02) (0.02) (0.02) (0.02) (0.02) (0.03) (0.03)
Observations 207 206 205 204 203 202 201
Month 16 Month 17 Month 18 Month 19 Month 20 Monetary Shock EU -0.04 -0.05∗ -0.07∗∗ -0.06∗∗ -0.05∗∗
(0.03) (0.03) (0.03) (0.02) (0.03)
Observations 200 199 198 197 196
Standard errors in parentheses
∗p < 0.10,∗∗ p < 0.05,∗∗∗ p < 0.01
Note: The table presents the regression coefficients for Hungary, as presented by βhin Equation 2. The reaction to the shock
Appendix B
This section of the Appendix contains all descriptive figures mentioned in the analysis.
Figure 10: Overview Stock Market Movements: The original data was obtained in local currency. The graph presents logarithmic values. This was done because the interest of the paper is in the percentage
changes caused by monetary shocks, not nominal values.
