The E Ects Of ECB’s Asset Purchase Announcements On Euro .
The Effects of ECB’s Asset Purchase Announcementson Euro Area Government Bond YieldsFrederik Neugebauer April 19, 2018AbstractThis paper employs event study methods to evaluate the effects of ECB’s unconventionalmonetary policy program announcements on 10-year government bond yields of euro areamember states. It covers data from 11 euro area countries from January 1, 2007 to August31, 2017 and distinguishes between the more solvent countries (Austria, Belgium, Finland,France, Germany, the Netherlands) and the less solvent ones (Greece, Ireland, Italy, Portugal, Spain). The paper makes two contributions to the literature. First, it is the first paperto reveal that measurable effects of announcements arise with a one-day delay meaningthat government bond markets take some time to react to ECB announcements. Second, itquantifies the country-specific extent of yield reduction which seems inversely related to thesolvency rating of the corresponding countries. The reduction of the spread between bothgroups in response to an event is due to a stronger decrease in the less solvent group. Byemploying different data as control variables, it turns out that the results are robust for agiven event set.JEL classification: E44, E52, E58, G14Keywords: ECB, unconventional monetary policy, government bond yields, event study WHU – Otto Beisheim School of Management, Department of Economics, Burgplatz 2, 56179 Vallendar,Germany. frederik.neugebauer@whu.edu. The author would like to thank participants at the 11th RGS DoctoralConference in Economics, and the 1st CESifo EconPol Europe PhD Workshop: Economic and Fiscal Policy inEurope.
1IntroductionNational government bond yields include the risk premium of a specific country. That is whythe sole announcement of an asset purchase by a central bank can already reduce the yieldsthrough amended expectations of investors. Event studies by Joyce et al. (2011) and Gagnonet al. (2011) find evidence for short-term yield reductions to quantitative easing announcementsin the UK and in the US, respectively. Previous research, however, indicates that announcementeffects are somewhat specific to the respective country.1 This study aims at quantifying ECBannouncement effects for the euro area. In particular, the study examines whether there is asimilar or varying impact on 10-year government bond yields of different euro area members.Such evidence is of high relevance for ECB’s policy making and communication strategy. Beforemaking an announcement the ECB might want to assess its consequences on individual euroarea members because it matters whether an announcement is perceived differently within theeuro area.In general, massive asset purchases by any central bank provide more liquidity. The presentstudy, however, focuses on the sole announcements of such liquidity provision while the actualamount of asset purchases is not considered.2 A credible asset purchase announcement directly affects investors’ expectations on the (future) attractiveness of particular assets (or assetclasses). As a potential consequence the demand for these assets rises and increases asset prices.In case of government bonds, this, in turn, directly reduces the government bond yields in question. For this short-term mechanism to work, it is irrelevant whether the future quantitativeeasing measures have the expected effects or whether they merely work as a placebo. Morespecifically, it is expected that ECB’s asset purchase program announcements have a strongereffect on the government bonds of stressed countries since the programs intend to foster primarily the euro area economies under stress. In contrast, the yields of more solvent countriesare expected to be less sensitive to such announcements. Although at the announcement dayit is yet not clear what sort of assets the ECB exactly will buy, that is to which economy thepurchased assets belong, the possibility that also assets from the countries under stress mightbe bought, substantially smooths market expectations.1For the case of Japan with its long history at the zero lower bound and quantitative easing measures, noevidence of yield reactions to central bank announcements exists. In contrast, in an event study Bernanke et al.(2004) state that communications by the Federal Reserve change market expectations and thus long-term yieldsin the US while statements by the Bank of Japan do not affect Japanese yields.2For a study that implements actual purchases to assess the impact on sovereign bond yields, see for instanceEser and Schwaab (2016).1
Most existing literature only investigates the announcement effects on the aggregate euroarea as a whole. For instance, Ambler and Rumler (2017) use weighted average real yieldsof all euro area countries to search for announcement effects. Their research indicates thatannouncements lead to significantly lower real bond yields. The few existing disaggregatedstudies compare only a few countries, however. For instance, Altavilla et al. (2016) analyzethe effects of outright monetary transactions announcements on the government bond yields ofGermany, France, Italy and Spain while Briciu and Lisi (2015) look exclusively at the yields ofonly Germany, Italy, and Spain in response to ECB’s announcements. Both studies find yieldreducing effects in response to ECB’s unconventional monetary policy announcements.Many studies consider the effects on yields’ spreads rather than levels. For instance, Falagiarda and Reitz (2015) state that the inter-European spreads on government bond yields decreasein response to ECB’s asset purchase announcements. They find a reduction of long-term yieldspreads of Ireland, Italy, Portugal and Spain. Similarly, Szczerbowicz (2015) evaluates theimpact of ECB’s unconventional monetary policies on 10-year sovereign bond yield spreadsof France, Greece, Ireland, Italy, Portugal and Spain with respect to the German sovereignyield. She also confirms spread reducing effects. Recently, Bulligan and Monache (2018) quantify the spread reduction of Italy, France and Spain (vis-à-vis Germany) for asset purchaseannouncements between September 2014 and July 2017. Nevertheless, the question remainswhich (relative) level effects of the respective spread-defining yields exactly underlie these spreadreductions.Therefore, this study covers a large number of euro area members and focusses on the leveleffects. For policy making, it is essential to see the absolute (level) impact of an announcementto evaluate its costs or benefits. The relative (spread) position to another economy is lessimportant. Furthermore, this study covers a long time span of over ten years. So far, studies inthis field of research are typically constrained to a shorter period. For instance, Christensen andKrogstrup (2018) only consider events during one month, Altavilla et al. (2016) during threemonths and Gagnon et al. (2011) during two years.Hence, this study extends the existing literature in three directions. First, the separateconsideration of individual euro area members allows a comparison of national effects. A euroarea average impact seems not entirely helpful for policy analysis. A study of differences between countries gives important insights on economic conditions of the respective countriesinstead. Second, the focus on the level is more useful than spread analysis. A reduction in2
spread does not explain the inherent direction of yield changes, that is whether both yieldsare increasing/decreasing to a different extent or whether they are moving into opposite directions. Third, the long time span guarantees that announcements are considered at differentstates of the financial crisis. Unlike Bulligan and Monache (2018) who divide their three yearobservation period into subsamples this study aims at a (time-invariant) generalization of thefindings. Given that some programs and their announcements last for a long time and arecontinuously prolonged, it would be inappropriate to include only a part of its announcementhistory. Moreover, a long sample period increases the validity and reliability of findings, byimproving statistical properties with additional observations.By covering data from 11 euro area countries from January 1, 2007 to August 31, 2017and searching for country-specific level effects on 10-year government bonds yields of ECB announcements the paper adds two significant findings to the existing literature. First, to the bestof my knowledge this event study is the first one to reveal that the effects of announcementsarise with a one-day delay meaning that government bond markets take some time to react toECB announcements. Second, it shows that the country-specific quantitative extent of yieldreduction is inversely related to the solvency rating of the corresponding euro area country: Theworse the rating is, the bigger the yield reduction is. This also implies that the observed reduction of the yield spread between core/solvent and periphery/less solvent countries in responseto an announcement event is due to a stronger decrease in the yield of the latter. A group-wisepanel analysis confirms these findings. By employing different data as control variables, it isdemonstrated that these findings are robust for a given event set.The remainder of the paper proceeds as follows. Section 2 describes the methodology andthe data. Section 3 presents the empirical results. Section 4 provides robustness checks, whilethe final section 5 concludes.2Methodology and DataIn order to investigate the short-term impact of ECB’s asset purchase announcements on the10-year benchmark bond yield to redemption of individual euro area members,3 an event studymethodology as in Moessner (2015) is applied. The dataset consists of daily yields (per bankworking day) of Austria, Belgium, Finland, France, Germany, Greece, Ireland, Italy, the Nether3The results hold taking 10-year zero coupon government yields as dependent variable instead of yields toredemption.3
lands, Portugal and Spain from January 1, 2007 to August 31, 2017. These are the foundingmembers of the euro area except Luxembourg while Greece (which joined in 2001) is additionally included. While a small country like Luxembourg might bias the results it is essential toinclude Greece as an economy heavily hit during the sovereign debt crisis. The study is limitedto long-term yields only to overcome the zero lower bound problematic or even negative yieldsthat are partly present for short-term yields.The identical regression is carried out for each government bond yield yt to test whetherthere are different reactions among the euro area countries. The baseline specification usesfirst-differences and is yt α β1 yt 1 β2 stockt β3 CESIt β4 excht γAP At εt ,(1)with t 1, ., T 2784 observations per country denoting the daily observations for each variable and the error term εt 0, σ 2 , while α is a constant.It is assumed that the present day’s yield change is dependent on that of the previousday as common in financial time series. Therefore, a one lag estimator yt 1 is included inthe regression as in Urbschat and Watzka (2017).4 The choice of additional control variablesis motivated as follows. The country-specific stock market indices stock t intend to representthe investors’ perception of an economy. A rising index ceteris paribus reduces the defaultrisk of sovereign debt. Thus, it decreases government bond yields. The Citigroup EconomicSurprise Index for the Eurozone CESI t is defined as weighted historical standard deviations ofmacroeconomic data surprises and controls for general events taking place all over Europe. Apositive development of this index increases perceived risks of investors, which, in turn, increasesbond yields. The influence of the US- /e spot exchange rate (in price notation) is captured byexcht .5 It intends to control for the link between exchange rate movements and interest ratesdue to international arbitrage considerations. All variables are obtained from Datastream andare end-of-business-day values (‘close prices’).4The application of the model with an additional two-day lag estimator yt 2 shows an insignificant estimatorfor all bonds under consideration. One might argue that a lagged dependent variable might cause endogeneityproblems. Although it is common in related literature to apply such lags (Szczerbowicz (2015), Jäger andGrigoriadis (2017), Urbschat and Watzka (2017)), the model is also applied without a lagged dependent variableto overcome endogeneity concerns as a robustness check. The results (available upon request) persist highlightingthat endogeneity is negligible in this kind of models.5Of course, one could also implement an effective exchange rate such as the rate vis-à-vis the EER-19 tradingpartners. However, due to gaps in the data availability (overall 52 missing observations) the spot exchange rateis convenient. Note that independent of the chosen exchange rate variable the results remain almost identical.4
The timing is important to consider. An announcement event usually takes place in themiddle of the day. For instance, one would expect different reactions in case of start-of-businessday values (‘open prices’) or daily averages. It would be interesting to include market sentimentmeasures such as the index of economic policy uncertainty by Baker et al. (2016) or the consumerconfidence indicator by the European Commission. However, these indices are not available ata daily frequency and a transformation of monthly survey data to a daily basis would bias theresults.AP At is the variable of our main interest. It is a dummy variable taking the value of 1 on theday of a specific ECB asset purchase program announcement, and 0 otherwise. In contrast toFalagiarda and Reitz (2015) who add a dummy variable for each single event, all announcementevents are represented by one common dummy variable in order to detect a generalized effect ofan ECB announcement. An overall effect is more suitable for policy making because the ECB isinterested in gauging the average effect of similar future announcements. If each announcementis considered individually, the result is only valid for an identical announcement in the future.The coefficient γ measures the general announcement effect and it is expected to have a negativesign (γ 0).An integral element in the analysis is the identification of asset purchase announcements. Pressreleases and statements by ECB’s officials are therefore carefully scrutinized according to theircontent. This approach of deliberately determining events is common in related literatureand ‘entails a certain degree of subjectivity’ (Ambler and Rumler (2017), p. 10). Table A1in the Appendix shows an extended list of potentially relevant events that might affect theEuropean government bond market. Out of this list, 23 events are chosen and denoted in bold.Furthermore, the keywords why a certain event is included are denoted in italics. That means,AP At is equal to 1 on these days, and 0 otherwise.All events refer to specific asset purchase programs. For this reason the famous ‘whateverit takes’ statement by Mario Draghi on July 26, 2012 is not chosen as it is not referring toa specific program. Other monetary policy statements, for example press releases regardingconventional monetary policy tools or forward guidance statements6 are also omitted becausethe study focuses merely on unconventional quantitative easing measures. Announcements onpurely technical details of asset purchase programs are not considered as they do not provide6Since forward guidance was recently implemented in the euro area, studies that look explicitly at forwardguidance are limited to the Federal Reserve that implemented it earlier (Moessner (2015), Neugebauer et al.(2017)).5
new information to the market. One might argue that confirming announcements such as thepress releases by the ECB on January 19, 2017 or July 20, 2017 are not effective. However, theyare included for the following reason. Since investors believe that there might be an end of theextreme expansive monetary policy a repeated announcement that contradicts this expectationcan lead to surprise effects. Some studies rely on news from other sources than ECB officials,for example Altavilla et al. (2015) use a news database to screen articles for keywords in orderto detect relevant events. However, this approach is less helpful to work out policy implicationsbecause media is out of control of the ECB and can only be indirectly influenced by its policystatements.The choice of the correct event window width is another debatable element in any eventstudy. While a too long window width induces the risk of contamination of news not relatedto monetary policy, a too short window width might neglect delayed effects of monetary policyannouncements. Recent literature typically uses either one-day windows (e.g. Glick and Leduc(2012), Haitsma et al. (2016), Georgiadis and Gräb (2016)) or two-day windows (e.g. Altavillaet al. (2015), Szczerbowicz (2015), Christensen and Krogstrup (2018)). Figure 1 exemplifies theevent window for the announcement made on March 10, 2016: the dummy variable either is setto 1 on March 10 only (one-day window) or on both March 10 and March 11 (two-day window).Figure 1: Event window13:45: Press release of monetary policy decisions (TLTRO II, CSPP)March 9, 2016March 10, 2016March 11, 2016t-1tt 1one-day windowtwo-day windowThis study sets the window width to just one day as the risk of including effects from otherevents is evaluated higher than the possibility of excluding delayed effects. Furthermore, fre-6
quent trading on financial markets supports this choice.7 To capture potential delayed reactionsto the announcement we will lag the one-day to the March 11 rather than expanding the windowwidth.Table A2 in the Appendix presents the descriptive statistics. It shows that the stock indices vary considerably across countries. Their standard deviations range from 83 points (theNetherlands) to 7,367 points (Italy). The yields differ substantially in their level. They rangefrom a minimum of -0.22% in case of Germany to a maximum of 48.6% in case of Greece. Thisunderpins the choice of applying first-differences instead of absolute values.The data set motivates to distinguish two country groups: The first group consists of Austria,Belgium, Finland, France, Germany and the Netherlands. Their bonds show moderate yieldsover time with an average smaller than 3% and a maximum smaller than 6%. In contrast, thesecond group consisting of Greece, Ireland, Italy, Portugal and Spain possesses high bond yieldswith an average higher than 3% and a maximum up to 48%. The former groups is labeled ‘Corecountries’ and the latter group is referred to as ‘Periphery countries’ in the following. Theseexpressions are synonyms for the more solvent and the less solvent countries, respectively.The grouping corresponds to the common distinction of PIIGS countries and other euro areamembers.8Plotting the dependent variable over time gives additional insights. Figure A1 in the Appendix shows that the divergence in yields emerged from 2010 on and it clearly demonstratesthe varying levels across countries. Since the middle of 2014, all yields except for Greece persistat a lower level than in 2007. To detect possible differences within both groups, Figure A2 andFigure A3 in the Appendix plot the yields of Core countries and Periphery countries, respectively. While the yields of Core countries are similar and follow the same (negative) trend, theyields of Periphery countries do not. The high peaks of Ireland, Greece and Portugal explainits high standard deviations of more than 2 percentage points (cf. Table A2). In contrast, Italyand Spain only exceed the 6%-threshold marginally in 2012. The figures indicate that the dataseems to be non-stationary. An augmented Dickey-Fuller test shows that all variables
Germany, France, Italy and Spain while Briciu and Lisi (2015) look exclusively at the yields of only Germany, Italy, and Spain in response to ECB’s announcements. Both studies nd yield reducing e ects in response to ECB’s unconventional monetary policy announcements. Many studies c
6.1.6 ECB Hand/Off/Auto 6-9 6.1.7 ECB Mode of Operation 6-9 6.1.8 Bypass Status Word Bit Examples 6-11 6.1.9 ECB Auto Bypass 6-11 6.1.10 ECB Run Permissive 6-12 6.1.11 ECB Overload 6-12 6.1.12 ECB Safety Interlock 6-13 6.1.13 ECB Common Run/Stop 6-13 Contents Option Panel Instruction Manual MG14I122 - VLT is a registered Danfoss trademark
May 02, 2018 · D. Program Evaluation ͟The organization has provided a description of the framework for how each program will be evaluated. The framework should include all the elements below: ͟The evaluation methods are cost-effective for the organization ͟Quantitative and qualitative data is being collected (at Basics tier, data collection must have begun)
Silat is a combative art of self-defense and survival rooted from Matay archipelago. It was traced at thé early of Langkasuka Kingdom (2nd century CE) till thé reign of Melaka (Malaysia) Sultanate era (13th century). Silat has now evolved to become part of social culture and tradition with thé appearance of a fine physical and spiritual .
On an exceptional basis, Member States may request UNESCO to provide thé candidates with access to thé platform so they can complète thé form by themselves. Thèse requests must be addressed to esd rize unesco. or by 15 A ril 2021 UNESCO will provide thé nomineewith accessto thé platform via their émail address.
̶The leading indicator of employee engagement is based on the quality of the relationship between employee and supervisor Empower your managers! ̶Help them understand the impact on the organization ̶Share important changes, plan options, tasks, and deadlines ̶Provide key messages and talking points ̶Prepare them to answer employee questions
Dr. Sunita Bharatwal** Dr. Pawan Garga*** Abstract Customer satisfaction is derived from thè functionalities and values, a product or Service can provide. The current study aims to segregate thè dimensions of ordine Service quality and gather insights on its impact on web shopping. The trends of purchases have
International Market Research 5 ECTS Modern Strategic Analysis 5 ECTS International Standards of Production-Management 5 ECTS Corporate Finance 5 ECTS Sales Management 5 ECTS Business Marketing 2.5 ECTS Russian Civilization 1.5 ECTS Management of Investments 5 ECTS Die Vorlesungszeiten s
Chính Văn.- Còn đức Thế tôn thì tuệ giác cực kỳ trong sạch 8: hiện hành bất nhị 9, đạt đến vô tướng 10, đứng vào chỗ đứng của các đức Thế tôn 11, thể hiện tính bình đẳng của các Ngài, đến chỗ không còn chướng ngại 12, giáo pháp không thể khuynh đảo, tâm thức không bị cản trở, cái được
