Bachelor’s Thesis
The Effect of the Forward Guidance Announcement
on the Liquidity Premium of Eurozone Countries
Student name: Bogdan Marinel Student number: 10621091
Date of the final version: July 29, 2016 Specialization: Economics & Finance Supervisor: Gabriele Ciminelli
Statement of Originality
This document is written by Bogdan Marinel who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is 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 of completion of the work, not for the contents.
Abstract
The European Central Bank (ECB) is an inflation targeting central bank. One of the policies within the ECB’ package of policy actions in achieving their inflation goal is forward guidance. Forward guidance is an unconventional monetary policy that provides information on the future path of the interest rates given by the central bank itself. In this paper, the effects of the forward guidance announcement are studied using an event-study methodology on three different variables. We first study the effect of the announcement on current EURIBOR rates and on expected EURIBOR three-month future rates on which we find significant negative effects. We also study the outcome of the announcement on two-year zero coupon government interest rates of thirteen Eurozone countries. We find that the interest rates of nine countries are significantly negatively affected by the announcement and two countries’ interest rates are significantly positively affected by the announcement. Lastly, we study the forward guidance announcement outcome on country’ specific term premium changes. We find three countries for which the announcement has a strong positive effect: Portugal, Slovakia and Malta, results that might be affected by country-specific factors. We also find three countries for which the announcement had a significant negative effect: Greece, Spain and Italy, three GIIPS countries for which the forward guidance announcement has the expected outcome.
Contents
1. Introduction 1
2. Literature Review 4
3. Data and Methodology 7
3.1 Data ………... 7
3.2 Methodology ………..… 10
4. Results and Analysis 12
5. Conclusion 20
1
1. Introduction
When entering a monetary union, countries lose more than just instruments of monetary policy such as exchange rate and interest rate. They lose the ability to issue debt in their own currency over which they have full control according to De Grauwe (2011, p.1). This in turn can result in a loss of confidence of investors in the selected country which can lead the country into default. The same researcher posits that if the investors fear a country’s default, they sell that country’s government bonds, increasing their interest rate and invest the common currency in a country with the same currency but with better credit rating.
Four years after the beginning of the housing bubble burst, the high levels of debt in the Eurozone (EZ) were a threat to the recovery of the European economy (Lin &, Treichel, 2012). Throughout such times the European Central Bank (ECB) had to deal with so called “asymmetric shocks”. These are shocks that affect countries in different manners, and hence, cause a problem of synchronization of income, inflation and unemployment rates (Cuestas et al., 2016, p. 2). As ECB has as its sole objective price stability, being an inflation targeting central bank, finding solutions to such shocks is indeed a priority.
Fama (1975, pp. 269-270) argues that short-term interest rates are predictors of inflation, meaning that when setting the nominal interest rates a central bank can obtain relevant information about future inflation. By influencing the short-term interest rate, a central bank can indirectly influence inflation. A policy that aims to do so is called forward guidance, an unconventional monetary policy that provides information on the future path of the interest rates given by the central bank itself. Thus, starting from the 4th of July 2013, the Governing Council of the ECB has been providing forward guidance on the future track of the ECB’s policy interest rates conditional on the outlook for price stability, stating its intention to keep interest rates at prevailing or lower levels for an extended period of time (Praet, 2013, pp. 1-2).
Monetary policy can control with some precision the very short-term interest rates. Nonetheless, the effects of monetary policy on the economy do not only depend on them, but also on the expectations formed by the public of how those rates will evolve in the future. This is so due to the fact that expectations of future short-term rates set by the Central Bank determine longer term market interest rates through the expectations channel plus liquidity premium. The
2
theory behind this channel states that long-term interest rates are equal to the average of the expected short-term interest rates plus a liquidity premium, which represents the country’s specific risk. According to the ECB Monthly Bulletin (2014, pp. 2-3), expectations of future interest rates matter because they affect important economic decisions such as investment and durable consumption. Therefore, by influencing expectations of future short-term interest rates and medium to long-term interest rates, a central bank can ensure that its policy stance is transmitted to the broader economy.
The aim of the forward guidance measure is to guarantee that market expectations on future monetary policy are indeed consistent with the policy intentions of the ECB. As argued by
Cœuré (2013), forward guidance has an effect on long-term interest rates only if agents expect monetary policy to be effective in lowering long-term interest rates and stimulating economic growth. Thus, announcements of the Central Bank should always deliver reliable signals to the market. If the Central Bank fails to provide such signals, monetary policy may become a source of increased credit risk, complicating the economic decisions of private actors.
The credit risk of each country comes in question as the forward guidance introduction can also affect the credit risk premium, a component of risk premia, measuring a country’s default risk. De Grauwe (2011) argues that the ability of a country to repay its debts partly depend on the interest rate itself. When investors expect the country to go into default, they sell the country’s bonds. This leads to an increase in the country’s interest rate and therefore makes it more difficult for the country to repay its debts. When the ECB introduces the forward guidance policy, which aims to keep interest rates at prevailing low levels for an extended period of time, medium-term interest rates should also decrease, leading to lower interest rates on countries’ bonds. If interest rates do actually decrease, then the repayment of the country’s debt should come easier, thus leading to a lower credit risk and hence a lower liquidity premium. This aspect is evaluated in this paper. Since one expects the effect to be stronger in countries with a higher initial level of debt, it is not enough to limit the analysis to only one country. Thus, the analysis is done on a larger sample of countries, thirteen Eurozone countries, in order to obtain a more explicit set of results.
This brings us to the aim of this paper, analysing how both future short-rate and two-year country specific interest rates react to the announcement of Forward Guidance and what effect
3
this announcement has on the credit risk of every Eurozone country, based on the impact of forward guidance on the liquidity premium within all the selected countries.
To achieve this, we approach an event study methodology, on a time-sample of four years: July/2011-July/2015. The study is performed on only thirteen Eurozone countries as data is not available on the four year period for the rest of them. The macroeconomic announcement that is analysed is the forward guidance announcement of the ECB publicized on the 4th of July 2013. We are estimating the effect of the announcement on both expected future short-term interest rates as well as two-year interest rates of every selected country. Subsequently, we approximate the liquidity premium with the OLS coefficient estimations. Lastly, we are analysing the effects of the forward guidance announcement on the liquidity premium approximations.
Our main findings are as follows. First of all, the forward guidance announcement had a significant negative effect on EURIBOR future rates with a maturity up to 8 quarters, meaning that forward guidance indeed decreased the short-term interest rate expectations. Secondly, the same announcement significantly reduced the two-year government bond rates in nine out of thirteen countries that we studied and significantly increased the rate in two. The three highest swings in the country’ interest rates are on Spain, Italy and Greece, countries with high level of debt. Last of all, we estimated the liquidity premium of the selected countries and conclude that for Spain, Greece and Italy forward guidance has a stronger effect in decreasing their liquidity premium. Explanations for such results may be the high level of debt and the high credit risk present in those countries. Accordingly, forward guidance helps in repaying their debts and reduces their credit risk. On the other hand, we also find two strong positive effects on Portugal and Malta. Such positive results to a forward guidance announcement may be the outcome of country-specific factors.
The development of the rest of the paper is done following standard academic practice. First, a theoretical framework will be presented, providing background researches and theoretical support. Further on, the methodology and data will be discussed, including the explanations for choosing our variables, variables computations and the approach of the research, followed by the analysis on the results of the study presented. Lastly, arguments are weighed and conclusions are drawn, outlining several limitations and offering recommendations for further research.
4
2. Literature Review
As stated by Andrew Filardo and Boris Hofmann (2014, pp. 37-38), forward guidance is not something new. Central banks started using forward guidance since 1990, mainly as qualitative descriptions of their main thrust of the interest rate policies to inform the public. Further development of forward guidance has been adopted in its quantitative form since the late 1990’s by a number of small inflation targeting economies by regularly publishing forecasts of their future interest rate policy path. Nevertheless, what the ECB is pursuing now is forward guidance at zero lower bound. The aim of this type of forward guidance is to clarify the central banks’ future policy path rate. It can also reduce the interest rate volatility and through this channel also the risk premia.
Following its meeting on 4 July 2013 the Governing Council of the ECB communicated that it “expects the key ECB interest rates to remain at present or lower levels for an extended period of time”. Such conditional statements about the future path of the policy interest rates are frequently referred to as forward guidance (ECB, 2014). This monetary policy influences the economy and inflation by affecting medium to long term interest rates. Longer-term interest rates are, indeed, the rates that determine the conditions of borrowing which are the most relevant factors of aggregate spending, consumption and investment. Thus, by analysing the expectation theory of the term structure of interest rates, one can remark that long-term interest rates are the outcomes of the expected path of the short-term interest rates which the central bank controls. Arbitrage opportunities are removed across securities with similar risk characteristics and different maturities, thus returns are equally weighted across the term structure except the liquidity premia which remunerates investors holding relatively longer term securities (Filardo, & Hofmann, 2014, pp. 41-42). Hence, by influencing the expectations of the evolution of those short-term interest rates, the central banks can impact the decisions that matter the most for the economy.
There are several studies which estimate the impact of central banks’ announcements of policy regulations on the interest rates.
Firstly, Brand, Buncic and Turunen (2010, pp. 1267-1268) have led a study regarding the effects of monetary policy announcements and have analysed their impacts on the euro area
5
money market yield curve. The first important finding in their research is that news stemming from the communication of ECB policy have a more substantial and longer-lasting impact on the euro area yield curve than news stemming from actual decisions on policy interest rates. Secondly, their results suggest that market expectations of the path of monetary policy may change considerably during the ECB’s press conferences and have an impact on the longer-term interest rates. These findings validate the relevance of their research for this paper, conducting a similar analysis but with a different methodology and different time-frame. Their methodology differs from the one this paper approaches, as a multidimensional-indicator model is used and one-month EONIA swap rates as expected short-term rates rather than EURIBOR three-month futures. Another difference is that their study refers to all monetary policy announcements whereas we study a single announcement.
Secondly, a similar research conducted by Bernoth and von Hagen (2004, pp. 3-10) argue that the EURIBOR futures rates with a forecast horizon of up to four months are unbiased estimators of future spot rates. Another important aspect is that the volatility of the future rates is significantly higher on days when the ECB Governing Council met than on non-Council days. Nevertheless, the ECB policy announcements have no systematic effect on the absolute size of the prediction error contained in EURIBOR future rates as well as the use of the volatility of the future rates as a measure of surprise caused by the central bank’s policy decision.
An important information for our study is what effects other central banks’ forward guidance had on their underlying interest rates. This aids with the comparison of results, benefits and drawbacks between our study and other findings. An important reference point is the article written by Andersson and Hofmann (2009), in which they analyse the effect of forward guidance adopted by three inflation targeting central banks, the Reserve Bank of New Zealand, the Norges Bank and the Riksbank. This is convenient as the ECB itself is an inflation targeting central bank with its main goal being ”a year-on-year increase in the Harmonized Index of Consumer Prices (HICP) for the euro area of below 2%, which is to be maintained over the medium term”. By computing a comparative analysis between the three chosen countries two main important findings arise. First, there is enough evidence that all three central banks have been highly predictable in their monetary policy decisions and that long-term inflation expectations have been well anchored in the three economies, regardless of whether forward guidance involved
6
publication of an own interest rate path or not. Second, for New Zealand, there is weak evidence that a publication of a path may potentially enhance a central bank’s leverage on the medium term structure of interest rates. On the other hand, their findings suggest that if a central bank has already a high degree of transparency and a clearly defined price stability objective, a publication of an interest rate path does not appear to enhance the short-term predictability of monetary policy.
It is true that when the ECB comes into question, we are dealing with a bank that has its targets clearly defined, but we are looking at an economy with different long-term interest rates and risks due to the large number of different countries. Therefore, even if Andersson and Hofmann (2009) claim that publication of an interest rate path does not appear to enhance the short-term predictability of monetary policy for a bank with high transparency and clear goals, the effects of forward guidance can vary across the Eurozone countries due to their country-specific risk and liquidity premium.
A similar approach to ours is Moessner’s (2013) study on the effects of the Federal Open Market Committee policy rate guidance on interest rate expectations as an unconventional monetary policy tool at zero lower bound. The study is performed using an event study methodology, with EURODOLLAR forward rate as the proxy for the expected short-term interest rates and macroeconomic surprises as control variables. The study is addressed to the United States where it resulted that explicit policy rate announcements significantly reduced implied interest rates on horizon of one to five year ahead.
Until now, we focused only on the effects and results of the monetary policy announcements on the interest rates. An interesting approach tackled by Levin et al. (2009) in their article, is related to the limitations of the effectiveness of forward guidance at the Zero Lower Bound. In their analysis they use a New Keynesian model under conditions where a macroeconomic shock makes the zero lower bound interest rates a binding constraint on monetary policy. By applying their method, they found that forward guidance has alone a better outcome on relative large and persistent shocks than a discretionary policy. One cannot conclude that forward guidance outcome is sufficient to keep the interest rates and inflation rates close to the long-run goal, but that it has indeed a role as an unconventional policy when combined with other monetary policies.
7
We consider the event-study approach one of the best approaches in achieving the purpose of our paper, as we are interested in the impact of a single event, the ECBs’ forward guidance announcement, on the changes of our dependent variables. It is indeed true that this is not a perfect method and thus, further research can be done.
The contribution of this paper is to measure the impact of the ECB’s forward guidance on the Eurozone country-specific liquidity premium, by studying its effects on short-term interest rate futures estimated by EURIBOR rates, as well as the effects on long-term rates implied by two-year zero-coupon government bond rates.
3. Data & Methodology
3.1. Data
Firstly, the time-frame that is considered is July/2011-July/2015. We choose this period as it begins two years before the date of the ECBs’ Governing Council announcement regarding the introduction of forward guidance and two years after the announcement. By choosing a four year interval, the number of observations is high, leading to a greater statistical power.
For our analysis we use four kinds of data: expected short-term interest rates, long-term interest rates, dummy variables, and control variables. Values of the interest rates are used as dependent variables whereas the dummy variables and the control variables are used as independent variables.
Our dependent variables are daily changes in expected short-term interest rates and long-term interest rates during the selected period. The measures of the short-term interest rates expectations are three-month EURIBOR deposit interest rates proxy by EURIBOR future rates. According to Bernoth and von Hagen (2004), EURIBOR three-month future rates are unbiased predictors of future spot rates and as well informationally efficient. These are not directly observable, thus the proxy for these futures are derived from short-term interest rate contracts (STIRs). STIRs are derivatives that allow agents to hedge interest rate risk. The expected EURIBOR rate is calculated as follows:
8
Where are the estimated expected future rates and are the trading prices of the three-month EURIBOR future with settlement date q-quarter. We measure the time in quarters due to the fact that EURIBOR futures have four delivery months: March, June, September and December.
Further, for two-year interest rates we use country-specific two-year maturing zero-coupon government bonds, thus leading to thirteen different long-term rates.We use zero coupon bonds as they are the bonds which are most likely to be affected by changes in expectation of future interest rates, since they do not pay a coupon. Data regarding the two-year interest rates has been collected from DataStream. Our study is limited to just thirteen out of seventeen countries that had Euro as a currency since 2011 as data for the other four countries is not available. The thirteen chosen countries are: Austria, Belgium, France, Finland, Germany, Greece, Italy, Ireland, Malta, Netherlands, Portugal, Spain and Slovakia. As the four countries that are not analysed are not considered outliers of the current interest rates value pattern according to the ECBs’ website, we consider that the validity of this paper is not affected by not including them.
Our first independent variable is the dummy variable which marks the date of the ECB forward guidance announcements, meaning that the dummy variable takes the value of 1 during the day of the announcements and 0 otherwise. The date of the announcement is 4th of July 2013 and is officially released on the ECBs’ website.
The second dummy variable events all the ECBs’ Governing Council meetings since July 2011. The dummy takes the value 1 on the dates of the meetings and 0 otherwise. The role of this variable is to isolate the effect of the forward guidance meeting from the other meetings effects on the interest rates, leading to a specific estimate of the forward guidance effect. The dates of the meetings have been collected from the ECBs’ official website where every date of the meeting is posted.
We also use control variables for the effects of the macroeconomic news on the interest rates expectations in order to isolate the effect of forward guidance on interest rates from other macroeconomic indicators releases. We use four control variables which we constructed using data from Forexfactory, a website containing data on the actual release and the market
9
expectation of relevant macroeconomic announcements. We use four control variables in our analysis.
The first control variable is the Minimum Bid Rate which is the interest rate on the main refinancing operations that provide the bulk of liquidity to the banking system. The rate is set by the ECB Executive Board and governors of the Euro area central banks. This is an important factor to account for as traders are interested in the changes of this rate as it influences the currency valuation.
As a second control variable we use the German ZEW (German Center of European Economic Research) Economic Sentiment. It is constructed based on surveys of around 275 German institutional investors and analysts which state their opinion about the economic outlook of Germany. They are considered to be highly informed by virtue of their jobs. It is one of the leading indicators of economic health of Germany. As German centrality has created high expectations for it to provide leadership (Bulmer, 2014, p. 1244), such an important indicator for Germany may influence the overall Eurozone.
The third control variable is the Final CPI (Consumer Price Index), which is the overall Europe inflation index, used by the ECB as its inflation target. This variable accounts for inflation, thus the changes in the price of goods and services purchased by the consumers. It is important to account for inflation as control variable in our study as changes in prices can lead the central bank to change interest rates out of respect for their inflation containment mandate.
Lastly, the unemployment rate is used. It measures the percentage of the total workforce that is unemployed and is actively seeking employment in the Eurozone. The number of unemployment people is an important sign of overall economic health because consumer spending is highly correlated with labour-market conditions. As spending is also influenced by the medium-term interest rates that we are using in the analysis, changes in unemployment rate is an important variable to control for.
For all our four control variables we compute the surprise content by taking the differences between Actual Releases and Market Expectations (forecasted values) on the dates the announcements are released on. For example, if the CIP Actual Value on the 4th of September is
10
0.4% and the Forecast Value is 0.3%, the value in which we are interested to control for on the 4th of September is the difference between them, 0.1%.
3.2. Methodology
We adopt an event study approach, regressing changes in future short-term three-month EURIBOR rates, , and changes in medium-term two-year zero coupon bonds, , with m= 2 years ahead and t as the time period measure, in our case being days, on two dummy variables and four control variables. The first dummy variable represents the announcement of the ECBs’ forward guidance, while the second denotes the meetings of the ECBs’ Governing Council, The four control variables are macroeconomic surprises components of Eurozone macroeconomic data releases, , where j takes the value of the surprise number (j=1, 2, 3, 4). Their role is to control for the effects of macroeconomic releases on the interest rate expectations.
The regressions are constructed using the OLS estimation technique as we are dealing with linear regressions. Several assumptions are taken into consideration. First of all the error term is considered to have a zero mean and a zero correlation with the dependent variables. Secondly, we consider no linear dependence, meaning that the observations are not influenced between them, thus being linearly independent. We consider all the assumptions to hold throughout our analysis unless specify otherwise.
The regression equation takes the following form:
∑
(1)
where
is equal to
or , takes the value of 1 on the day when the ECB announces the forward guidance and 0 otherwise and takes the value of 1 on ECBGovernment Council meetings and 0 otherwise. Additionally is the variable denoting unobserved random errors which account for the discrepancy between the actual responses of the dependent variable and the predicted outcomes of the independent variables. By performing OLS we consider White heteroskedasticity-consistent standard errors as they are supported by past
11
literature, thus robust estimation techniques are used when computing the regression (e.g. the letter of Moessner (2013)).
After all the regressions are performed, thirteen regressions on long-term interest rates and nine regressions on short-term expect interest rates, we proceed to the second step of the analysis, the estimation of the change in the liquidity premium term of each country selected for the analysis. The estimation is done using the Liquidity Premium theory which states that the long-term interest rate is equal to the average expected short-term interest rates plus a liquidity premium, thus the liquidity premium theory equation has the form
(2)
where is the liquidity premium of the selected country.
Further, we compute the liquidity premium changes approximations using the forward guidance coefficients estimated by the regressions performed in the first step,
,
integrating them in the liquidity premium theory. It is indeed the case as we are interested in the effect of the forward guidance on the liquidity premium of each selected country, thus it leads us to the formula:(3)
where is the long-term country-specific coefficient of the forward guidance announcement, the short-term interest rate coefficient of the forward guidance announcement and are expected short-term interest rate forward guidance coefficient, which in our case are eight in total, three-month futures within two years’ time-frame. All these coefficients are estimated in equation 1.
After the liquidity premium changes are estimated for every country, we approach our final step, which is also the hypothesis of this paper, whether the announcement of the forward guidance has a significant impact on the liquidity premium of the countries or not, and . Although we do not formally test this hypothesis, an intuitive interpretation is given regarding the outcome. Thus, for every country for which the null hypothesis is accepted
12
we can conclude that forward guidance did not change the risk premium, whereas for the countries where the null hypothesis is rejected the risk premium is indeed different from 0, thus the unconventional monetary policy had an effect on the country’ specific liquidity premium.
4.
Results & Analysis
We first present the results for estimating equation (1). From Table 1 it can be seen that the dummy variable for the forward guidance announcement is significant for both the EURIBOR rate changes and the three-month EURIBOR futures changes on a two year ahead timeline at a significance level of 1%. The largest effect is present at the horizon of seven quarters, where the individual announcement reduces the three-month EURIBOR futures rates by around 8.5 basis points on average. In general, it can be seen that the forward guidance announcement has a negative impact on the short-term interest rates and on future short-term interest rates expectations. In addition to interest rate expectations, EURIBOR futures’ rates also reflect the effects on risk premia. Moessner (2013, p. 172) argues that in addition to lower interest rate expectations, lower futures rates due to explicit policy rate guidance could also reflect decreases in risk premia, which could be associated with reduced uncertainty about the future path of interest rates due to the guidance. This aspect will be discussed later in this paper.
Table 1. Reactions of EURIBOR & EURIBOR futures to the forward guidance announcement Variables Euribor q1 q2 q3 q4 q5 q6 q7 q8 c -0.0018 *** -0.0019 ** -0.002 ** -0.002 ** -0.0021 ** -0.0022 ** -0.0022 ** -0.0022 * -0.0022 * d_FWG (t) -0.021 *** -0.041 *** -0.061 *** -0.065 *** -0.07 *** -0.075 *** -0.08 *** -0.085 *** -0.084 *** d_GV (t) 0.0029 0.0029 0.0026 0.0021 0.0022 0.0024 0.0022 0.0021 0.0017 R-Squared 0.0216 0.0152 0.0197 0.0175 0.0171 0.0172 0.0167 0.0167 0.0158 No. Observations 1040 1040 1040 1040 1040 1040 1040 1040 1040
Note: Above one can observe a total of nine OLS-regressions realized on a sample of 1040 interest changes observations. q- Represents the number of quarters of the future. Coefficients on surprises in 4 EUR macroeconomic variables are not shown in this and other tables but are available upon request. Sample period: 7/04/2011-6/26/2015.
Statistical significance is as follows: * Significance at 10%; ** Significance at 5%; *** Significance at 1%.
13
STATA outputs and other outputs regarding the variables which are not included in the results and tables but which are used in computing the coefficients that are used in our analysis are available upon request.
With equation (1) we also estimate the effects of the forward guidance announcement on the changes of medium-term interest rates within our sample of thirteen Eurozone countries. The countries are: Austria, Belgium, France, Finland, Germany, Greece, Italy, Ireland, Malta, Netherlands, Portugal, Slovakia and Spain. Table 2 presents the outcome of the forward guidance announcement on these countries two-year zero coupon government bond rates. It can be seen that three countries stand out, Greece, Italy and Spain, 3 of the GIIPS (Greece, Italy, Ireland, Portugal, Spain). As argued by Algieri (2013, pp. 233-234), GIIPS are the countries with weak economic and financial performance and high levels of debt. The highest influence is on Italy. The announcement reduced its two-year interest rates by approximately 18 basis points and it is significant at a significance level of 1%.
On the other hand, two countries’ interest rates changes estimates are not statistically significant at any of the three chosen significance levels, Greece and Slovakia. One can argue that for Greece even if the forward guidance had a strong influence on the interest rates, changing it by nearly 15 basis points, the standard error (0.1261) value is large enough to counter the effect, thus leading to a non-significant value. The large volatility of Greece interest rate may be one of the reasons for such results. For Slovakia, the forward guidance effect is not strong enough and it does not influence the interest rate. The forward guidance effect on the Slovakian interest rate is by 0.01 basis points. From the same table it is observed that Slovakia’s interest rates are affected by other central bank announcements and that the ECB government meetings coefficient is indeed significant at a level of 5%. One can conclude that Slovakia’s interest rates are significantly affected by ECB’s announcements but not particularly by the forward guidance on. The announcements coefficient is significant at a significance level of 5% while the forward guidance one is not significant at any significance level.
14
Table 2. Reaction of two-year zero coupon government bonds to the forward guidance announcement
Country C d_FWG (t) d_GV (t)
R-squared No. Observations AUSTRIA -0.002 * -0.058355 *** 0.002 0.0087 1040 BELGIUM -0.0014 -0.06589 *** -0.010 ** 0.0082 1040 FRANCE -0.0014 -0.078237 *** -0.004 0.0135 1040 FINLAND -0.002 ** -0.057604 *** 0.002 0.0075 1040 GERMANY -0.0019 -0.058002 *** 0.002 0.0073 1040 GREECE -0.0244 ** -0.147177 (0.1261) 0.189 0.0002 1040 IRELAND -0.0094 -0.055365 * -0.036 0.0027 1040 ITALY -0.0012 -0.181073 *** -0.014 0.0046 1040 MALTA -0.0022 0.0307369 *** -0.005 0.0001 1040 NETHERLANDS -0.002 ** -0.062054 *** 0.003 0.0139 1040 PORTUGAL -0.0074 0.0826951 *** -0.045 0.0019 1040 SPAIN -0.0014 -0.172503 *** -0.016 0.0039 1040 SLOVAKIA -0.0012 0.0001881 -0.011 ** 0.0139 1040
Note: Above there are the results of the 13 regressions performed on two-year interest rates of the Eurozone countries. Statistical significance is as follows: * Significance at 10%; ** Significance at 5%; *** Significance at 1%.
When it comes to the other countries than the ones specified previously, the forward guidance announcement coefficient has a significant influence on their interest rates changes. For Austria, Belgium, France, Finland, Germany, Malta, Netherlands and Portugal, the announcement has a significant effect at 1% level whereas for Ireland the effect of the announcement is significant at a 10% level. A general conclusion of the effect of the announcement on the countries medium-term interest rate is that forward guidance significantly reduced the interest rates in nine out of the thirteen countries. The only country in which the
15
interest rate significantly increased is Malta. One explanation for this result is that the interest rates in Malta were already below zero when the announcement was released, with a value of -0.03.
The second step in our analysis is the computation of the liquidity premium changes of every country with the Liquidity Premium Theory, equation (3). One would expect the countries with higher level of debt, the GIIPS, to have higher changes in liquidity premium as forward guidance can decrease the credit risk. The results of the estimates are presented in Table 3. Table 3. Estimators for the liquidity premium changes due to the forward guidance announcement
Note: The results in this table are the raw estimates of the changes in the liquidity premium From Table 3 it can be seen that the countries that have high changes in their liquidity premium are indeed four out of the five GIIPS and Malta, with Portugal having the highest value, nearly 15 basis points increase on average. The expectation of the forward guidance announcement effect is to decrease the liquidity premium, reducing the credit risk of high debt countries by lowering their interest rates. When referring to the five highlighted countries from Table 3, Italy, Spain and Greece, are facing such a decrease. On the other hand, in Portugal and Malta an opposite effect is seen. Such positive results may be the outcome of country-specific
Country ΔLiquidity premium GERMANY 0.006729533 GREECE -0.082445367 AUSTRIA 0.006376333 SPAIN -0.107771767 SLOVAKIA 0.064919433 PORTUGAL 0.147426433 MALTA 0.095468233 ITALY -0.116341167 NETHERLANDS 0.002677433 IRELAND 0.009366333 FINLAND 0.007127233 FRANCE -0.013505167 BELGIUM -0.001158267
16
factors. As an example, according to the BBC timeline, several senior ministers of Portugal resign over the handling of the economic crisis in July 2013.
The countries that stand-out are the countries that did stand-out on the whole analysis, Greece, Spain, Portugal, Malta, and Italy together with Slovakia. By interpreting the results, one can see that the forward guidance announcement strongly decreased the credit risk of Greece, Spain and Italy. For those three countries forward guidance was beneficial in reducing their credit risk, aiding them in repaying their debt. On the other hand, the same announcement strongly increased the credit risk of Portugal, Slovakia and Malta. These three countries interest rates were already at low levels before the introduction of the forward guidance, thus, the announcement of the forward guidance increased this countries interest rate, exposing them to a higher credit risk. These positive effects can be a subject to further studies as country-specific factors should be taken in consideration for a more exact analysis and explanation.
In order to assess whether the estimated effects are persistent rather than fading out after one day, the same analysis is done on a three-day return window. The difference between the methodology used on three-day window returns and the analysis of the one-day window returns is the computation of the dependent variables in equation 1. The changes and are replaced with and , thus looking at three-day returns.
Table 4. Reactions of EURIBOR & EURIBOR futures to the forward guidance announcement on three-day return window
Variables Euribor q1 q2 q3 q4 q5 q6 q7 q8 c -0.0046 *** -0.005 *** -0.0058 *** -0.0061 *** -0.0065 *** -0.0067 *** -0.007 *** -0.0072 *** -0.0074 *** d_FWG (t) -0.0255 *** -0.0424 *** -0.0687 *** -0.0791 *** -0.0954 *** -0.1065 *** -0.1175 *** -0.1283 *** -0.1341 *** d_GV (t) 0.000 0.0025 0.0045 0.0053 0.0069 0.0082 0.0096 0.0105 0.0115 R-Squared 0.0325 0.0288 0.0129 0.0097 0.05372 0.008 0.0076 0.0077 0.0072 No. Observations 1038 1038 1038 1038 1038 1038 1038 1038 1038
Note: Above one can observe a total of nine OLS-regressions realized on a sample of 1038 interest changes observations. q- Represents the number of quarters of the future. Sample period: 7/06/2011-6/26/2015. Statistical significance is as follows: * Significance at 10%; **
17
From Table 4 it can be observed that the results on three-day return window are significantly negative on all EURIBOR and EURIBOR three-month futures rates. This result is similar to the one for one-day returns. It can be seen that on the three-day window results the effect of the forward guidance announcement is stronger. As an example, when analysing the EURIBOR seven-quarter futures, it can be observed that on the three-day window the forward guidance coefficient is -0.1283 and on the one-day window it takes the value of -0.085. Thus, it can be concluded that forward guidance has a consistent negative outcome on the EURIBOR and EURIBOR futures.
For the two-year government bonds, there are several differences in outcomes when considering different time-frame returns. When we analyse the effect of the announcement on one-day returns we find that Italy, Spain and Greece have the highest coefficients. When establishing a longer period of returns, it can be observed from table 5 that the effects on the same three countries coefficients are standing out but with lower values. Among the countries standing out, Finland and Portugal are noticeable as well when we refer to the three-day window. Finland seems to not react that sudden to the announcement when one-day returns are used but it has a clear strong effect when we analyse the announcement on three-day returns.
Another difference is that besides Portugal, all other countries’ interest rates are negatively influenced by the announcement. When the one-day window is used, Malta and Slovakia’s interest rates reacted positively to the announcement. When we use the three-day window, forward guidance affected the countries’ interest rates negatively. Thus, when a three-day period return is used, the forward guidance announcement has a consistent negative effect on twelve out of thirteen countries.
When using the three-day window returns we expected Portugal to follow a similar pattern as the other GIIPS countries. The effect on Portugal on the other hand is even stronger in a positive direction (nearly 17 times stronger). As we stated previously in this paper, Portugal’ results may be influenced by country-specific factors for which we did not control in our analysis.
After we estimate the liquidity premium for the three-day window return data, we observe that five out of the six countries that stand out also stand out when one-day window data
18
is used. The country that changes is Spain. While for the one-day window return data Spain’ liquidity premium seem to be strongly affected by the announcement, when we look at the three-day windows return data we can observe that this effects is not persistent and it is actually fading out. On the other hand Finland does not encounter an immediate effect. When one-day window return data is used their liquidity premium is not significantly affected by the announcement, but when three-day window return data is used the outcomes show a persistent significant effect. Table 5. Reaction of two-year zero coupon government bonds to the forward guidance announcement on three-day return window
Country c d_FWG (t) d_GV (t) R-squared No. Observations AUSTRIA -0.0061 *** -0.0925 *** 0.0058 0.0046 1038 BELGIUM -0.0074 * -0.0848 *** 0.0033 0.0042 1038 FRANCE -0.0059 *** -0.0957 *** 0.0066 0.0066 1038 FINLAND -0.0062 *** -0.1296 *** 0.0079 0.0115 1038 GERMANY -0.0059 *** -0.0881 *** 0.0077 0.008 1038 GREECE -0.0716 -0.1512 0.5533 0.0005 1038 IRELAND -0.0361 ** -0.0699 -0.0154 0.001 1038 ITALY -0.0053 -0.1345 *** -0.0304 0.0045 1038 MALTA -0.0081 -0.0109 * 0.0017 0.0024 1038 NETHERLANDS -0.0062 *** -0.0824 *** 0.0094 0.0088 1038 PORTUGAL -0.0353 * 1.3927 *** -0.0249 0.0147 1038 SPAIN -0.0079 -0.1027 *** -0.0134 0.0028 1038 SLOVAKIA -0.0066 *** -0.0135 * -0.0050 0.0246 1038
Note: Above there are the results of the 13 regressions performed on two-year interest rates of the Eurozone countries. Statistical significance is as follows: * Significance at 10%; ** Significance at 5%; *** Significance at 1%.
19
For the five countries that are significant at the both one-day window and the three-day window returns, it can be concluded that the forward guidance announcement effects on their liquidity premium are persistent and they do not disappear one day after the announcement. Table 6. Estimators for the liquidity premium changes due to the forward guidance announcement using three-day return window
Country ΔLiquidity premium AUSTRIA -0.003888889 BELGIUM 0.003811111 FRANCE -0.007088889 FINLAND -0.040988889 GERMANY 0.000511111 GREECE -0.062588889 IRELAND 0.018631111 ITALY -0.045888889 MALTA 0.077711111 NETHERLAND 0.006211111 PORTUGAL 1.481311111 SPAIN -0.014088889 SLOVAKIA 0.075111111
Note: The results in this table are the raw estimates of the changes in the liquidity premium Comparing the results of the forward guidance announcement effects with other unconventional monetary policies announcements can give insight towards the direction of such publication’ outcomes. According to a study pursued by Altavilla et al. (2014, pp 2-5), the ECB announced on the 8th August 2012 the introduction of Outright Monetary Transactions (OMT), an unconventional monetary tool. The results of their study suggest that the OMT announcement decreased the Spanish and Italian two year government bond yields while the bond yields of the same maturity bonds are left unchanged for Germany and France. Krishnamurthy et al. (2014) argues that the decreases in the bond yields due to the OMT lead to a decrease of the default risk in the countries in question.
The result of the OMT announcement is similar to the result of our study regarding the four countries in question. For Germany and France we have a significant small change in the two-year government bonds and a non-significant change in their liquidity premium. On the
20
other hand Italy is facing a consistent decrease in its interest rate and liquidity premium. However, in the case of Spain, the liquidity premium is significantly affected for one day only while its interest rate falls significantly due to the announcement.
5. Conclusion
This paper’s aim is to determine if the forward guidance announcement had an effect on the short-term EURIBOR futures rates and medium-term two-year zero coupon Eurozone countries government bond rates as well as its impact on the countries specific liquidity premium. Our results answer all the aspects of the research question. First of all, the forward guidance announcement had a significant negative effect on EURIBOR future rates with a maturity up to 8 quarters. Thus, the forward guidance announcement had its desired outcome on these rates. Secondly, the same announcement significantly reduced the two-year zero coupon government bond rates in nine out of thirteen countries that we studied and significantly increased the rate in two out of thirteen countries when a one-day window returns are used. When three-day window returns are used, twelve out of thirteen countries interest rates negatively react to the announcement, leaving Portugal the only country positively affected by the announcement. Country-specific factors may be the cause for such an unusual result for Portugal. Last of all, forward guidance had a strong impact on the term premium of six out of the thirteen countries divided in three significantly positive effects and three significantly negative effects when we use one-day return data and a similar outcome when we use three-day return data. We consider the positive effects to be caused by country-specific factors which we do not take in consideration whereas negative effects are the expected outcome of the forward guidance on the countries with higher level of debt, in our case Italy Greece and Spain. The effect of the announcement is persistent in five countries: Portugal, Greece, Italy, Malta and Slovakia.
There are several limitations to our study. The choice of the variables can be an important limitation. This study is limited to five control variables. Adding more control variables can lead to a less bias estimation of the coefficients. Controlling for country specific factors can also improve the overall results.
As we also compare the outcomes between our study on the forward guidance announcement effects and the outcomes of the outright monetary transactions announcement
21
effects, an extension of this study can be the effects of other unconventional monetary policy announcements on the liquidity premium of the studied countries. One such policy is Quantitative Easing, a policy which has been announced by the ECB on 22nd January 2015.
When it comes to external validity, as our study is done on thirteen countries with shared monetary policy, one cannot generalize the results to other economies due to the variety of characteristics and country-specific economic environment encountered in a monetary union. Thus, our study is relevant just for further studies regarding the Eurozone.
As the aim of this paper is to study only the direction of the forward guidance announcement effect on the selected variables, further studies can be done more in depth on the reasons for such effects. There are several remarkable results which come rather unexpected: the inconsistent outcomes of Spain which lead to diminished effect of the announcement after one day and the positive effects on the liquidity premium of forward guidance for Portugal, Slovakia and Malta. Both results can also be subjects for further studies.
22
Bibliography
Algieri, B. (2013). An empirical analysis of the nexus between external balance and government budget balance: The case of the GIIPS countries.Economic Systems, 37(2), 233-253. Altavilla, C., Giannone, D., & Lenza, M. (2014). The financial and macroeconomic effects of
OMT announcements.
Andersson, M., & Hofmann, B. (2009). Gauging the effectiveness of quantitative forward guidance: evidence from three inflation targeters.
Bernoth, K., & Von Hagen, J. (2004). The Euribor futures market: Efficiency and the impact of ECB policy announcements. International finance, 7(1), 1-24.
Brand, C., Buncic, D., & Turunen, J. (2010). The impact of ECB monetary policy decisions and communication on the yield curve. Journal of the European Economic Association, 8(6), 1266-1298.
Bulmer, S. (2014). Germany and the eurozone crisis: between hegemony and domestic politics. West European Politics, 37(6), 1244-1263.
Cœuré, B. (2013). The usefulness of forward guidance. discurso pronunciado en Money
Marketeers Club of New York, Nueva York, 26.
Cuestas, J. C., Gil‐Alana, L. A., & Taylor, K. (2016). Inflation convergence in Central and Eastern Europe vs. the Eurozone: Non‐linearities and long memory. Scottish Journal of
Political Economy.
De Grauwe, P. (2011, July). Managing a fragile Eurozone. In CESifo Forum(Vol. 12, No. 2, p. 40). Institut für Wirtschaftsforschung (Ifo).
Fama, E. F. (1975). Short-term interest rates as predictors of inflation. The American Economic
Review, 65(3), 269-282.
Filardo, A. J., & Hofmann, B. (2014). Forward guidance at the zero lower bound. BIS Quarterly
Review March.
Krishnamurthy, A., Nagel, S., & Vissing-Jorgensen, A. (2014). ECB policies involving government bond purchases: Impact and channels. Unpublished working paper.
Levin, A. T., López-Salido, D., Nelson, E., & Yun, T. (2009). Limitations on the effectiveness of forward guidance at the zero lower bound.
Lin, J. Y., & Treichel, V. (2012). The crisis in the Euro zone: did the euro contribute to the evolution of thecrisis?. World bank policy research working paper, (6127).
Moessner, R. (2013). Effects of explicit FOMC policy rate guidance on interest rate expectations. Economics Letters, 121(2), 170-173.
Praet, P. (2013). Forward Guidance and the ECB. Forward Guidance: Perspectives from Central
Bankers, Scholars and Market Participants, Vox eBook. The ECB’s Forward Guidance (2014), Retrieved from:
