an overview of european economic indicators ...5caf75ee-6277-4f29-92ab-013a6c80867b/mop... · this...

This contribution provides an overview of the most common short-term indicators of economic develop-ment in the euro area. These indicators are useful when official data are released with long time lagsor if they are subject to major revisions. Indicators based on surveys among businesses, households, finan-cial market analysts or forecasters have the advantage of providing detailed and timely information on in-dividual sectors on a monthly basis and largely without later revision. As an additional instrument, compo-site indicators, which are calculated by combining a variety of measures into a single indicator with the helpof regression and factor analysis, offer an attractive tool for drawing conclusions from different, often di-vergent signals. Even the most reliable economic indicators, however, can only be interpreted as constit-uent elements of comprehensive economic analysis. With regard to the new EUMember States, coverageis found to be limited as yet. This study also shows that the forecasting quality of the European Commis-sion�s business and consumer surveys for the newMember States is not as high as for the other EUMem-ber States. As the reliability of economic indicators increases as forecasting institutions and respondentsgainmore experience, coverage of established indicators should be extended early on to this group of coun-tries, in particular as some of the new Member States may soon join the euro area.

JEL classification: 0110, 520Keywords: reading indicators.

1 Short-Term EconomicIndicators — IntegralComponents ofEconomic Analysis

New data on short-term economic in-dicators regularly make headlines.The release of new Ifo index data forGermany, for instance, can even topthe latest GDP growth figures in termsof media presence. This is because pru-dent economic and monetary policiesare geared toward future economic de-velopment and reflect all available datathat help gauge current and future eco-nomic trends. To direct economic pol-icy according to past data alone wouldbe like steering a car while lookingonly in the rear-view mirror.

For the very same reason, eco-nomic indicators are an important toolfor Eurosystem central banks. To fulfillits mandate, the Eurosystem pursues afuture-oriented strategy, which isgeared toward economic developmentin the medium term. Monetary policyis unable to respond to short-term fluc-tuations owing to the lags in the trans-mission process and owing to the de-

gree of uncertainty that surroundsthe effects of monetary policy becauseof the complexity of the transmissionprocess. With a medium-term mone-tary policy strategy, excessive activismand the introduction of unnecessaryvolatility into the economy can beavoided (ECB, 2004b).

The Eurosystem central banks basetheir economic analysis — one of thetwo pillars of monetary strategy — onnot only the latest economic data avail-able but also on short-term economicindicators and forecasts that are, inturn, based on such data and indica-tors. Forecasting models are most reli-able when the economy is on a stablegrowth track. By contrast, they arefar less reliable in signaling turningpoints. Economic indicators help to re-duce this uncertainty and are thereforean integral component of economicanalysis in the Eurosystem�s monetarystrategy.

Furthermore, economic indicatorsenjoy great popularity because officialdata on real GDP growth — a key refer-ence measure for indicators — are not

1 Translation from German.2 JEL codes: C43, C53, E32 (Leading indicators, business cycle fluctuations, forecasting).3 The author would like to thank Markus Arpa, Jesu«s Crespo Cuaresma, Doris Ritzberger-Gru‹nwald and Martin

Schneider for their valuable comments and assistance as well as Maria Dienst, Angelika Knollmayer and AndreasNader for their support in data collection.

Maria AntoinetteSilgoner3

Refereed by:Martin Schneider,Economic AnalysisDivision.

An Overview of European Economic Indicators:Great Variety of Data on the Euro Area, Need forMoreExtensive Coverage of the New EUMember States1, 2

66 Monetary Policy & the Economy Q3/05�

adequate for short-term economicanalysis owing to a number of prob-lems. First, real GDP figures are pub-lished only on a quarterly basis, and re-lated monthly series for the most partrefer to the manufacturing industry,with the service sector being coveredunsatisfactorily. Second, they are re-leased with long time lags and fre-quently subject to major revisions.Lastly, the series are subject to meas-urement errors and problems in bothdata gathering and processing, and theyare not comparable internationally dueto methodological differences.

Of these aforementioned prob-lems, the release lags are most criticalin the analysis of economic perform-ance. As a case in point, the first releaseof quarterly real GDP growth figuresfor the euro area does not becomeavailable until about two months afterthe end of a given quarter, and even aflash estimate based on the data ofsome Member States is released witha time lag of one and half months. Bycontrast, data on consumer and indus-trial confidence calculated by the Euro-pean Commission are available on thelast day of the quarter for all threemonths of the quarter. The EuropeanCommission�s Euro Area GDP Indica-tor, a range estimate of quarterlygrowth, is released even five monthsbefore the confidence indicators andsubsequently updated monthly in thelight of new information.

A large number of economic datacommonly come under the umbrellaof short-term economic indicatorsfor the euro area. These can be brokendown into the following categories:— Measurable economic data can help

to assess the performance of GDPgrowth in a timely manner. First,these can be data on GDP subcom-ponents (individual countries or in-dividual sectors) that are released

earlier on. For instance, growthin industrial production is oftenused as an indicator for GDPgrowth. Second, data reflectingthe early stages of the productioncycle may be very useful. Thesecan be data from sectors or coun-tries specialized in intermediategoods but also data on inventories,building permits and overtimehours.

— Surveys are a common method ofobtaining data from economic ac-tors (consumers, company execu-tives, financial analysts, forecast-ers) on their assessment of the cur-rent or future economic situation.Individual reponses are aggregatedto derive sentiment indicators.

— Composite indicators, finally, are aproduct of statistical methods thatextract a single indicator from alarge number of data that, in addi-tion to aforementioned variables,also include key determining fac-tors of future economic develop-ment, e.g. oil prices and interestrates.This study focuses on sentiment in-

dicators based on surveys (section 2)and composite indicators (section 3)that are particularly closely watchedby the media and by economic expertsin the euro area. Section 4 presents afew rather peculiar indicators that arealso repeatedly mentioned in the me-dia. All in all, the indicators presentedhere do not necessarily reflect all exist-ing types of indicators, but they repre-sent the key methods and problemareas. Although the focus is on indica-tors for the euro area as a whole, na-tional indicators are also presented ifthey are followed in the euro area.

The overview of indicators in eachsection starts with a technical descrip-tion of surveymethods, sampling prop-erties and availability of data and also

An Overviewof European Economic Indicators: Great Varietyof Dataon the Euro Area,

Need for More Extensive Coverage of the New EU Member States

Monetary Policy & the Economy Q3/05 67�

addresses critical aspects of the calcula-tion method of which one should beaware for interpretation purposes.The indicators are assessed accordingto various quality criteria providedthey are directly comparable. To illus-trate the uncertainty that may sur-round indicators, the post-9/11 periodis taken as a specific example of themost recent past when, after an initialmood of crisis, it was impossible to saywhich way the economy was going. Inearly 2002, several sentiment indica-tors issued mistakenly strong signalsof an upturn that never materialized.In the final quarter of 2002, GDPgrowth came to a mere 1.1% year onyear. Even if this period cannot be de-scribed as anything but exceptional, itis nevertheless worth looking at the ex-perience with individual indicators, asit highlights the problem that respond-ents often do not see the situation anyclearer themselves in times of greatuncertainty.

Section 5, finally, examines the ex-tent to which comparable indicatorsfor the ten new EU Member States(NMS) are already available andwhether they qualitatively differ fromindicators for countries that have pub-lished such measures for decades.Given that some of the new EU mem-bers could join the euro area shortly,the availability of such indicators forthe NMS may soon be of relevancefor the euro area, on which this contri-bution — and most economic analysis —focuses primarily. Although the eco-nomic importance of most of thesecountries is limited, they currentlyconstitute the most dynamic regionin Europe to which greater attentionwill be paid in future. Above all, thestill inadequate quality of official eco-nomic data in many caseswill stimulateinterest in reliable short-term eco-nomic indicators.

2 Sentiment Surveys:Indicators ofLong-Standing Tradition

Many of the most common economicindicators are determined in the formof surveys among businesses, house-holds, financial analysts or forecastinginstitutions. Although the surveys forthe most part ask qualitative questions,quantitative indications may be re-quired too. Whereas the results of sur-veys are primarily used to anticipatethe performance of key economic var-iables, they can also throw light on un-derlying factors or help assess the con-sequences of extraordinary eventsearly on.

In a summary article, the EuropeanCentral Bank (ECB, 2004a) cites anumber of advantages sentiment indi-cators have over officially publisheddata. First, they are released far earlieron than the latter. Second, data are re-leased on a monthly basis whereas theirreference series are frequently availa-ble only as quarterly data. Third, sur-veys can provide data that cannot be di-rectly gathered (e.g. capacity utiliza-tion in manufacturing industry).Fourth, survey data tend to be less vol-atile, as they are not (or less) influ-enced by one-off events (storms,strikes) and should therefore identifyturning points sooner. Lastly, surveydata are rarely revised.

All these advantages are also ac-companied by certain drawbacks. Forinstance, surveys provide primarilyqualitative data that are not easy to con-vert into quantitative assessments. Fur-thermore, survey data on different sec-tors may not necessarily be compara-ble. Finally, the quality of the resultsdepends to a great extent on howstrong the motivation of respondentsis to answer questions carefully. Thequality of the survey is itself difficultto monitor, as series cannot be subject




to quality checks on an ongoing basis.Despite these drawbacks, long experi-ence with some of these sentiment in-dicators puts them among the mostpopular short-term economic indica-tors.

A common methodological ap-proach, which was developed by theU.S. National Bureau of Economic Re-search (NBER) for a U.S. indicator,consists in providing respondents withset responses for their assessment ofthe current or future economic situa-tion. These can be broken down intothe following categories: positive (e.g.�the situation will improve considera-bly (PP)� or �the situation will improveslightly� (P)), neutral (�the situationwill remain unchanged�), negative(�the situation will deteriorate slightly�(N), �the situation will deteriorate con-siderably� (NN)) and the nil response(�No idea�). If p, pp, n and nn representthe respective share of respondents inthe corresponding reponse category,then the index value is given by the dif-ference between the positive and nega-tive reponses according to the for-mula

I ¼ ðppþ �pÞ � ð�nþ nnÞ;where � is the weight (generally 1�2or 1) with which slight movementsare downgraded relative to strongones. If respondents are given onlyone negative and one positive responseamong which to chose, the formulais I ¼ p� n.

Sections 2.1 and 2.2 present severalindicators from consumer, businessand financial market surveys whichare calculated either directly for theeuro area or are related to individualeuro area countries that are consideredto be important for the region as awhole. The box �Internet References�provides the web link for the relevantindicators. Table 1 presents a compari-son of key features of indicators.

Whereas all these indicators are pub-lished monthly, they do differ in termsof release dates in relation to the firstrelease of GDP growth figures, histor-ical availability, statistical correlationwith the reference series and their rep-resentativeness for the economy as awhole.

To analyze the statistical relationbetween indicators and economic de-velopment, this paper uses growth inthe euro area�s seasonally adjusted in-dustrial production rather than GDPgrowth as a reference series. Althoughindustrial production accounts for onlysome 25% of GDP in the euro area, it ispublished on a monthly basis and is,moreover, responsible for more thanhalf of the fluctuations in GDP. In addi-tion, many services (transportation,supplies, repairs) are directly relatedto industrial production.

Table 1 shows the maximum corre-lation coefficient between a given indi-cator and growth in industrial produc-tion that can be reached by adjustingthe time lag between the two series.The series are standardized in a waysuch that they have mean 0 and stand-ard deviation 1. The relative lag isshown in parentheses, with a negativefigure indicating an actual lead of theindicator, a positive figure representinga lag and 0 signifying that the correla-tion is highest when both series are co-incident. If, for example, the correla-tion coefficient is highest when the in-dicator series is lagged by two months(—2) relative to industrial production,then the January indicator will offerthe best insights into the growth of in-dustrial production in March. If, how-ever, the indicator data are lagging, say,by one month (+1) with respect to in-dustrial production, then only an ear-lier indicator release date could offeradded value. In other words, a coinci-dent or slightly lagging indicator can




act as a leading indicator in practice if itis published sufficiently early.

Furthermore, a Granger causalitytest is used to test the statistical rela-tion between the indicator and refer-ence series. Under ideal circumstan-ces, the indicator (I) is Granger causalfor growth in industrial production(IP) but not vice versa (this is denotedin table 1 as I!IP). Mutual (I$IP) cau-sality can be established only in two in-stances. A final test lastly checks howmany months earlier (negative value)or later (positive value) an indicatorreaches a turning point than industrialproduction figures. Table 1 shows theaverage lead or lag across the entiresample as well as the maximum andminimum time lead or lag in parenthe-ses.4 This is intended to illustrate thegreat uncertainty that surrounds theactual timing of an economic turningpoint signaled by the latest indicatorvalues.

A further statistical test is to checkhow well reference series can be pre-dicted using short-term indicators.One possibility is to use the historicalrelation between the reference seriesand indicator data (estimated on the ba-sis of the full range of data) to predicteconomic performance at individualreference dates and then check theforecasts against actual outcome (in-sample approach). The other possibil-

ity is to rerun estimates for eachrefernce date using only the data avail-able prior to the given forecastingperiod (out-of-sample approach).5 Ex-amples of such forecasting excerisesare Dreger and Schumacher (2005) orHu‹fner and Schro‹der (2002) for vari-ous German indicators. This study sys-tematically analyzes the forecastingquality only for the individual compo-nents of the European Commission�sESI (section 5).

2.1 Sentiment Indicators in theEuro Area

The Economic Sentiment Indicator (ESI),the origins of which go back to the1960s and which has been publishedby the European Commission (2004a)on a monthly basis since 1985, followsthe methodological approach of con-structing a balance of positive and neg-ative reponses from sentiment surveysas described above.While initially onlyfive countries took part in the project,today data are collected with a standar-dized questionnaire for all EUMemberStates (with the exception of Malta) aswell as for Bulgaria and Romania,which are scheduled to accede in2007 or 2008. EU and euro area aggre-gates are also published. Some of thesurveys are conducted by public insti-tutions and some by private national in-stitutions.6 The indicator, which is pub-

4 Turning points were calculated as extreme values of the three-month moving averages of both indicator series andgrowth in industrial production. Since the early 1990s, growth in industrial production has accordingly posted fivepeaks and troughs. Since the fourth peak and fifth trough represent only a slight economic improvement in thequarters post 9/11, followed by a further dent in growth (and not an upturn and a downturn in the current mean-ing of a business cycle), neither of these turning points was taken into account here. For most indicators, the lead orlag properties also have a historically atypical pattern in this period. If the same test is repeated with all ten turningpoints, the average lead or lag of the indicators differs from the value recorded in table 1, but the pecking order of theindividual indicators will remain essentially unchanged.

5 Inoue and Kilian (2004) show that in-sample tests more frequently indicate good forecasting properties than out-of-sample tests. For instance, a model based on past data may have fairly good predictive powers whereas a structuralbreak in the respective forecast horizon gives rise to forecasting errors.

6 Examples of the wide range of forecasting institutions are the Nationale Bank van Belgie‹/Banque Nationale deBelgique (NBB/BNB), Germany�s Ifo Institute for Economic Research, the Austrian Institute of Economic Research(WIFO), Hungary�s GKI Economic Research Institute, the Czech Statistical Office and the U.K.�s Confederation ofBritish Industry (CBI).




lished on the last working day of eachreferencemonth, is seasonally adjustedand standardized in a way such that thelong-term average has a value of 100.

The questionnaire and the sectorscovered have steadily grown in sizeand number. Some questions relate tothe economy as such (business situa-tion, production expectations, orderbooks, level of inventories); others re-

late to the inflation and the employ-ment expectations of households andto their financial situation, saving rateand big-ticket purchases. The ESI com-posite indicator aggregates five confi-dence surveys for the following sec-tors: industry (weight: 40%), services(30%), consumers (20%), construction(5%) and retail trade (5%), with each ofthese components being drawn from

Table 1

Comparison of Sentiment Indicators:

Indicator Quality for Growth in Industrial Production in the Euro AreaPublishinginstitution

Pub-lishedsince

Currentlyreleased forthe followingEU countries

Lead onGDPpublication1)

Samplesize in1,000

Sectorscovered

Numberofsubindi-ces2)

Maximumcorrela-tion coef-ficient3)

Grangercausa-lity4)

Lead/lagof turningpoints5)

ESIEuropeanCommission

1985 EU, euro area,24 countries

62 141 Consumer,industry,construction,retail, services

15+27 0.85 (+1) I!IP +1.9(—1; +5)

Industry confidenceindicator

EuropeanCommission

1985 as above 62 36 Industry 3+11 0.89 (+1) I!IP +1.5(—1; +4)

Service sectorconfidence indicator

EuropeanCommission

1996 as above 62 28 Services 3+2 0.69 (+1) I!IP +2.7(—1; +5)

Consumerconfidence indicator

EuropeanCommission

1985 as above 62 33 Consumer 4+10 0.71 (+3) I!IP +4.0(0; +11)

Constructionconfidence indicator

EuropeanCommission

1985 as above 62 21 Construction 3+2 0.39 (+5) I!IP x

Retail tradeconfidence index

EuropeanCommission

1985 as above 62 23 Retail trade 2+2 0.47 (+3) x x

Production expec-tations component

EuropeanCommission

1985 as above 62 36 Productionexpectations

1 0.90 (—1) I$IP —0.3(—3; +5)

PMI(Manufacturing)

NTC 1997 EU Euro area,11 countries

60 5 Manufacturing 8 0.87 (—1) I!IP —0.2(—3; +2)

Ifo businessclimate-index

Ifo 1984 Germany 66 7 Manufacturing,construction,trade

8 0.64 (0) I!IP —1.6(—5; 0)

Ifo business situationcomponent


4 0.58 (+3) I$IP +2.4(+1; +7)

Ifo business expec-tations component


4 0.69 (—2) I!IP —2.9(—6; —1)

ZEW indicator ZEW 1991 Germany 73 0.35 Financialmarket

1 0.80 (—5) I!IP —4.6(—9; —3)

Belgian BusinessSurvey

NBB/BNB 1954 Belgium 69 6 Manufacturing,construction,trade

3+1 0.79 (—1) I!IP —1.0(—3; +1)

1) Interval between the publication of the indicator value of the last month of every quarter and the first release of GDP growth of the corresponding quarter, measured in days,average of the first three quarters of 2005.

2) For the indicators published by the European Commission and the Belgian central bank, the first figure indicates the number of subindices included in the calculation of therelevant indicator. The second figure refers to the additional indicators released for each field.

3)Maximum correlation coefficient between indicator and growth in industrial production in the euro area. The degree of the lead/lag (inmonths) between the series, for which themaximum correlation is reached, is indicated in brackets; a negative value implies a time lead of the indicator.

4) Test at the 5% level. For the European Commission�s retail index, the null hypothesis (no Granger causality) cannot be rejected in either direction.5) Average interval (in months) between the turning points of the indicator and those of growth in industrial production. Maximum and minimum interval in brackets. Turningpoints are calculated using the relevant three-month moving average. A negative value implies a time lead of the indicator. The exercise was carried out only for indicatorsshowing all tested peaks and troughs in industrial production.




several individual questions.7 This indi-cator is thus composed of 15 individualcomponents in all. In addition, data de-rived from a further 27 questions,some of which are surveyed only on aquarterly basis, are also presented. Fur-thermore, an investment survey is con-ducted every six months in the indus-trial sector. Overall, 108,000 enter-prises and 33,000 households through-out the EU take part in these monthlysurveys.

ESI�s great advantage is its long his-torical time series and large sample aswell as the EU-wide standardizationof its surveymethod. At the same time,a distorting effect may arise from thefact that the balance of opinion reflectsa rather rough quantification of the ex-pected degree of improvement and/ordeterioration (�somewhat better� or�much better�). Moreover, it shouldbe borne in mind that the ESI is aslightly lagging indicator relative togrowth in industrial production, asthe analysis of correlation and turningpoints demonstrates. In other words,the index value published in a givenmonth is in fact an indicator for a pastmonth of the reference series. In prac-tice, the ESI serves as a leading indica-tor nonetheless as it is publishedaround two months before the indus-trial production series. Indeed, all fivemain components are lagging indica-tors, with industrial and service sectorconfidence having the best indicatorproperties for industrial production(short lag and high correlation). A trulyleading indicator is a subcomponent ofindustrial confidence (also shown in ta-ble 1), which asks explicit questionsbearing on production expectations in

the three months ahead and of whichparticularly good leading propertiesrelative to industrial production cantherefore be expected. Section 5 exam-ines the properties of the ESI in detail.

The Purchasing Managers Index(PMI) is modeled on its U.S. equiva-lent and has been calculated on amonthly basis for the euro area since1997 by NTC Research on behalf ofReuters for the manufacturing andservice sectors. All in all, more than5,000 businesses from eight countries(Germany, France, Greece, Ireland,Italy, the Netherlands, Austria andSpain), accounting for a total of 92%of the euro area, are covered by thesurvey. The PMI is published on thefirst working day following the endof each reference month and is brokendown by sector and country. Thequestionnaire for the the most fre-quently used Manufacturing PMI cov-ers the reassessment of output, em-ployment, new orders, suppliers� de-livery times and inventories (eightsubindices in all) compared with theprevious month. The PMI is standar-dized in a way such that an indexabove 50 shows expansion while an in-dex lower than 50 reflects contractionin the economic situation. However,the signaling function of this thresholdvalue may be somewhat flawed attimes, which is why fluctuations invalue should always be interpreted inrelation to the level as well. ThePMI is popular also on account of itsinternational comparability. After all,every G-8 country has been surveyedaccording to the same methodologysince 2002. The PMI also enjoys greattrust because its questionnaire is based

7 The weights are determined from both the relevant component�s importance for GDP and the level of correlationwith the reference series. The service sector has been surveyed since 1996 and has been a component of ESI only since2004 (European Commission, 2004b). The European Commission expects that the inclusion of the service sectorsurvey will increase the correlation of the index with the reference series and shorten the length of the indicator�s lag.




on actual facts and not on expecta-tions. Accordingly, table 1 shows theManufacturing PMI to be a slightlyleading indicator with a high correla-tion. NTC Research (2002) shows thatthe British PMI has in the past had abetter handle on the definitive GDPgrowth figure than the first releaseof GDP figures.

As chart 1 illustrates, the ESI com-posite indicator has proved to be rela-tively accurate in the post-9/11 period.In early 2002 it trended up only slightly,reflecting the weak and temporary up-turn in the economy relatively well. Bycontrast, the PMI rose steeply, over-stating the ensuing economic trends.

2.2 Germany and Belgium — Repre-sentatives for the Euro Area

In addition to these indicators that ex-plicitly cover the euro area, national in-dicators are also often considered to bean important gauge of the euro area�seconomic health. Those hogging thespotlight are Germany�s Ifo BusinessClimate Index, Germany�s ZEW Indi-cator of Economic Sentiment and Bel-gium�s Business Survey.

The Ifo Business Climate Index is pub-lished on the 25th of each reference

month by Germany�s Ifo Institute forEconomic Research. Senior managersin more than 7,000 businesses in Ger-man trade and industryare asked togivetheir assessment of the current businesssituation and their business expecta-tions for the six months ahead. The bal-ance of responses is determined accord-ing to the aforementioned methodol-ogy. The geometric mean of both theseindices is the most frequently used IfoBusiness Climate Index, which is stand-ardizedatan intervalof+/—100.The in-

�� !��4777��!

��

��

��

��

��>

��

��

��

��>

��

��

��

��

)�&��"� ��"��&��&�� <

��

+��()*%��!��'*��1��/��!%��'

��:9��

4��

F+D/��"+�� .�� %(�� '

/"0��!��%(�� '/"0��%(�� '+�� .�� /��%��'

��

��

��

��

��

��

��

��

��

��4��

��4�� 4�� 4�� 4�� 4�� 4�� 4�� 4�� 4�� 4�� 4�� 4��




dex is also broken down by subsector(manufacturing, construction, whole-sale and retail trade). Separate indicesused to be published for eastern andwestern Germany until 2004. Now thisdistinction is no longer applied, as theeconomic trends of both regions havesufficiently converged.The IfoBusinessClimate Index is coincident or evenslightly lagged relative to GermanGDP. That the index nonetheless hassuch a high profile in the media is alsoits association with a very memorablerule of thumb, according towhich threerises or falls in the index in successionherald a turning point in GDP growth.This rule is frequently applied also tothe euro area on account of Germany�shigh GDP weight.

This rule of thumb never failed inthe first 40 years of the Ifo index�s ex-istence. However, in the aftermath of9/11 the index slumped temporarilyto then rise three times in a row with-out being followed by a turning pointin GDP growth. This false signal forthe very first time in its historyprompted a debate about the indica-tor�s reliability. Although the situationat the time should be seen as an excep-tional event since an excessive down-ward correction was immediately fol-lowed by signs of equally unfoundedeuphoria, this case underlines thatsometimes just when the degree of un-certainty about the future is at a peak,economic indicators are also subjectto increased uncertainty. Ever since,the two subcomponents of the Ifo in-dex have received greater attention,as the false signal in early 2002 ema-nated only from the business expecta-tions component but not from its cur-rent business situation counterpart

(chart 1). Although the correlationanalysis in table 1 shows that the cur-rent business situation indicator has athree-month lag,8 while the expecta-tions index has a two-month lead, leadsshould not overrule reliability in timesof great uncertainty. Kunkel (2003)goes as far as concluding that the threesuccessive signals issued by the Ifo busi-ness climate index only indicate a turn-ing point reliably when they are subse-quently confirmed by three successivesignals from the business situation indi-cator.

The ZEW Indicator of Economic Sen-timent is a perfect foil to the Ifo Indexsince it consults precisely those expertsin Germany who are not included inthe Ifo sample, i.e. financial analysts.The Centre for European EconomicResearch (ZEW) has been surveying350 German financial experts fromthe banking, insurance and major in-dustrial sectors on a monthly basissince 1991. The ZEW indicator asksthese experts for their opinion on thesix-month prospects of the Germaneconomy. It also asks them for their as-sessment of key financial indicatorssuch as interest rates, equity prices,oil prices and inflation, as well as fortheir views on economic trends in theeuro area, Japan, the U.K. and theU.S.A.

As table 1 shows, the ZEW indica-tor has a lead of some five months onindustrial production in the euro area,thereby enjoying a significant lead rela-tive to the Ifo business expectations. Itis also published a week or so beforethe Ifo index. Hu‹fner and Schro‹der(2002) show that the ZEW index ismore suitable for medium-term fore-casts of the German economy than

8 The significant lag of the business situation index conforms to mutual Granger causality (I$R). The type ofquestions bearing on the current business situation make past industrial production trends a key determining factorfor the indicator.




the Ifo index of business expectations.The ZEW indicator is, however, gener-ally more volatile than the Ifo index forthe following reasons: its more limitedsample size, its smaller questionnaireand the fact that its respondents reactmore strongly to general market senti-ment and political or economic newsand are not themselves involved inbusiness life. Hu‹fner and Schro‹der(2002) show that the Ifo index of busi-ness expectations is more reliable forshort-term forecasts (up to threemonths). Like the Ifo index, theZEW indicator also falsely suggestedan upturn in the aftermath of 9/11(chart 1).

Whereas the high profile of boththese indicators can be easily explainedin view of Germany�s size, the signifi-cance attached to the Belgian BusinessSurvey — conducted on a monthly basissince 1954 by the Nationale Bank vanBelgie‹/the Banque Nationale de Belgi-que (NBB/BNB) among 6,000 seniormanagers of Belgian industry (manu-facturing, construction, trade, corpo-rate services) — requires a few wordsof explanation. Belgium is a small,open economy, with the euro area asits main trading partner. It specializesin intermediate goods and has a highshare of small and medium-sized busi-nesses. This is why economic changesin Belgium can be ascertained earlieron than for its trading partners in theeuro area. As a result, turning pointsin the Belgian business cycle have a sig-nificant lead on those in the euro area.Accordingly, the Belgian business cli-mate index also has a lead relative toGDP growth in the euro area (table 1;Vanhaelen et al., 2000). Furthermore,the popularity of the Belgian economicsurvey is based on the long historicaltime series and the internationallycomparable methodology used forthe European Commission�s ESI.

To sum up, the great advantage ofthese sentiment indicators is that theyhave been in use for many years, theircalculation method is simple, they arepublished early on and they are byand large not revised retrospectively.Past experience has shown that some-times a longer lead comes at the ex-pense of reliability in periods of greatuncertainty owing to a future-orientedperspective. In practice, this meansthat economic signals issued by leadingindicators should be substantiated bysignals from indicators that are moreclosely related to the present.

2.3 Surveys of Forecasters

The last two indicators in this sectionfocus on a completely different groupof respondents. Unlike the previous in-dicators, for which consumers, busi-nesses and financial analysts are inter-viewed, these indicators reflect theopinion of professional forecasters.The idea, which is also substantiatedby the relevant literature (Batchelor,2001; Blix et al., 2001; Zarnowitz,1984), is that although individual fore-casters may outperform the averageof a group of forecasters in certaincases, an individual forecaster rarelyoutperforms systematically. A consen-sus forecast should thus minimize therisk associated with forecasts and pro-vide a more reliable indicator.

Since 1989 Consensus Economics, aprivate British survey firm, has beenconducting a monthly survey of 400economists worldwide for their fore-casts of GDP growth, inflation, thecurrent account balance and interestrates in more than 70 countries. Theforecasts are classified by individualcountry and regional group and re-leased in four volumes (industrializedcountries, Asia-Pacific, Latin Americaand Eastern Europe). Twice a year,Consensus Economics also undertakes




special surveys on long-term forecasts.In April 2005, for instance, long-terminflation expectations in the euro area(reference year: 2010) stood at 2.0%,as did potential growth in the euro area.

Since early 1999 the ECB has beenconducting the Survey of ProfessionalForecasters (SPF), which asks a panel ofnearly 90 EU-based participants (finan-cial institutions, research institutes, aswell as employers� associations andtrade unions) in quarterly intervalsfor their predictions for the euro area(Garcia, 2003). SPF forecasters are freeto use a forecasting method of theirchoice (model forecasts, rule-of-thumb forecasts, subjective forecasts).Typically, about two-thirds of thosepolled will respond. The SPF question-naire asks for expected GDP growth,inflation and employment. A distinc-tive feature of the SPF is that, unlikeConsensus Economics, it does not onlyask for point estimates but also forcomplete probability distributions. Ac-cordingly, forecasters are to allocatesubjective probabilities to intervals(i.e. a range of possible outcomes)witha width of 0.5 percentage point. Thisthrows light on the risk spread aroundthe most probable forecast value andhighlights the uncertainty surroundingthe forecast. The main results are pub-lished in the ECB�s monthly bulletin.

Once a year, long-term forecasts (fiveyears ahead) are also collected. In thethird quarter of 2005, for instance,long-term inflation expectations stoodat 1.9%, potential growth at 2.1% andstructural unemployment at 7.6% (ref-erence year: 2010).

A basic problem with surveys offorecasters is that the expense and timeinvolved to make the forecasts cannotbe verified in practice. Although a cer-tain continuity of participants is ex-pected, model forecasts are likely tobe made only at large intervals of timewhereas a purely subjective update canbe issued in between these periods. Byway of surveys of interest rate forecast-ers, Bewley and Fiebig (2002) showthat the latter tend to indicate valuesin the safe consensus range so as notto stick their neck out with forecaststhat dramatically deviate from themean. This would lead to a bias in thedistribution toward the mean, result-ing in an unsatisfactory picture of therisk profile. In this sense, it is good thatSPF participants remain anonymousand that the survey is conducted onlyon a quarterly basis. This ensures thatforecasters do not come under exces-sive pressure to participate in surveysevery time — even if a current forecastupdate is not available.

��

�� '��$�(��)��89:233;

��(!��233<6 ��!��

��

��

��

��

��

��

�

�� <

�� G��H�

��

��

��

��

��

�

<

�� G��H�

" ��

< <




3 Composite IndicatorsWith the emergence of suitable statisti-cal methods of calculation and corre-spondingly powerful computers, com-posite indicators, which sometimes re-flect hundreds of data series, have expe-rienced a boom in recent years. Thebasic approach is to obtain informationfrom data that are considered to beleading indicators signaling economictrends, that respond quickly to eco-nomic fluctuations (e.g. overtimehours) or are themselves the cause ofeconomic fluctuations (e.g. oil prices,interest rates and exchange rates).The aim is to extract the �essence�from all these data series and to filterout disruptive factors such as contra-dictory signals issued by individual in-dicators, measurement errors and cal-endar or base effects. This would makethe movements of composite indica-tors more straightforward and easierto interpret. Likewise, economic fluc-tuations can have various causes andcharacteristics and be reflected in a va-riety of indicators.

A range of statistical methods isavailable to derive composite indica-tors from data series. These methodsdiffer in terms of how the compositeindicator�s input series are selected,standardized (correction for differentmargins of fluctuation), synchronized(time lead or lag in the composite indi-cator�s constituent series comparedwith the reference series), correctedfor outliers/seasonal fluctuations andweighted, and also differ with regard

to how the indicator is extracted. Ofthe statistical methods available, re-gression analysis and factor analysishave prevailed in practice. These twomethods are roughly outlined here. Inaddition, there are many other meth-ods (e.g. Markov switching model,state space model etc.), of which anoverview can be found in Marcellino(2006).

For both regression analysis andfactor analysis, the data series suitablefor further testing initially need to beselected from a large number of poten-tial candidates. The selection criteriainclude statistical (long time series,few revisions, low volatility, timely re-lease) and economic factors (stable em-pirical correlation with the referenceseries and economic plausibility). In re-gression analysis, the definitive selec-tion of series, their time lead or lagand the calculation of their weightsare carried out by using regression equa-tions. These weights are then used tocalculate the economic indicator or di-rectly predict growth based on the lat-est economic data. Theweights and theselection of series are usually kept con-stant over a certain period of time butare subject to regular reviews. This isnecessary, as composite indicators areonly ex post efficient since the correla-tions between input series and refer-ence series are subject to changes overtime (Emerson and Hendry, 1996).

A far more sophisticated approachis factor analysis, which became popu-lar in the 1990s. The Dynamic Factor

Internet References

Economic Sentiment Indicator (ESI): europa.eu.int/comm/economy—finance/indicators—en.htmPurchasing Managers Index (PMI): www.ntc-research.comIfo Business Climate Index: www.ifo.deZEW Indicator of Economic Sentiment: www.zew.deBelgian Economic Survey: www.nbb.beConsensus Forecasts: www.consensuseconomics.comSurvey of Professional Forecasters: www.ecb.int/stats/prices/indic/forecast




Model (Sargent and Sims, 1977;Geweke, 1977) was developed in the1970s on the basis of the Static FactorModel (Burns and Mitchell, 1946).The development of the GeneralizedDynamic Factor Model by a researchgroup at the Centre for Economic Pol-icy Research (Forni et al., 2000; Forniand Lippi, 1999; Forni and Reichlin,1998) made this method also applicableto large data sets at the end of the 1990s.The basic idea is to decompose eachdata series on different variables andcountries in a large panel of time seriesinto two unobservable components, ofwhich one is strongly correlated withthe other (common component) whilethe other one is not correlated at all orcorrelated only weakly (idiosyncraticcomponent). The common compo-nent is driven by a small number ofcommon �factors� or shocks, whichcan be interpreted as a synthetic indica-tor. The weights are therefore derivedfrom the properties of individual timeseries, within the entire gamut of con-stituent data series.

In Europe, factor analysis is usedfor two purposes. First, it offers ameans of estimating on a timely basisdata series that are only released witha large time lag and are subject to fre-quent andmajor revisions. TheOeNB�sshort-term economic indicator (web-link: www.oenb.at/de/geldp_volksw/prognosen/prognosen.jsp), which hasbeen published on a quarterly basissince 2003, is based on this method(Fenz et al., 2005; Schneider and Spit-zer, 2004). Second, factor analysis of-fers a novel approach of estimatingnonobservable time series such as thecore inflation rate.

Many institutions such as theOECD have been publishing compo-site indicators for decades in order toassess economic trends early on. Inthe last few years, however, a number

of new composite indicators of thistype have been established. It is ex-tremely difficult to compare the vari-ous approaches in terms of their infor-mative quality and reliability, as theirspecific methods are very different.For instance, they refer to different ref-erence series (GDP growth, growth inindustrial production, prediction ofturning points) in various presenta-tions (annual, quarterly or annualizedgrowth rate), have different leadingproperties and are released as an indexvalue or explicitly as a growth forecast(point estimator or range). The largenumber of data series that are embed-ded in composite indicators comprise:— Survey data: Consumer or indus-

trial confidence, construction in-dustry surveys, purchasing manag-ers� index, financial investor sur-veys;

— Real economic data: Industrial pro-duction, building permits, labormarket indicators, car sales, U.S.and Asian economic data;

— Price data:Consumer and producerprice data, core inflation rate, oiland other commodity prices;

— Financial data: Interest rates, inter-est rate spread, exchange rates,equity indices, international inter-est rate gap;

— Monetary aggregates: M1, M2, M3.The variables are included in the in-

dicator calculationwith a time lag/leadof varying length. Furthermore, manycomposite indicators also include dataon errors in previous publications,which makes them into �self-correct-ing� indicators. Table 2 systematicallypresents these methodological differ-ences. The box �Examples of Compo-site Indicators� (see p. 79), provides ad-ditional information on sampling, thecalculation method, the input seriesand the relevant web links.




Examples of Composite Indicators

The OECD has been publishing its Composite Leading Indicator (CLI), which is considered to be a goodindicator for turning points since the 1980s (www.oecd.org/std/cli). It is now ascertained for 23 countriesand 7 economic areas. The CLI for the euro area (weighted mean of individual countries� CLIs) has beenpublished since 1999. The CLI�s calculation includes 5 to 10 series per country. For the euro area, this in-volves 75 series in all. Theweight of sentiment indicators is almost 50%. The CLI has a comparatively weakcorrelation with the industrial production reference series, but it has a relatively long lead of six months onaverage. However, the CLI is also only published six weeks after the end of each month.

Handelsblatt, Germany�s financial daily, has been publishing an indicator for turning points in westernGermany since 1992, for eastern Germany since 1995 and, finally, for the euro area since 1999 (www.han-delsblatt.com). The Euro Area Economic Indicator is �self-correcting� and is improved regularly as a result,the lead could be extended from one quarter to up to three quarters, according to Handelsblatt sources.The reference parameter is seasonally adjusted, annualized growth of real GDP. The calculation includesfive individual series, with the weight of sentiment indicators having been reduced from an initial 50% to acurrent 30%.

The Financial Times has been publishing the Euro Growth Indicator since 2000. This is calculated bythe Euroframe group (nine research institutes fromGermany, France, Italy, the United Kingdom, the Neth-erlands, Finland, Ireland and Austria), (www.euro-frame.org). The indicator aims to predict annual realGDP growth two quarters ahead of official statistics. The forecast considers eight data series in all, threeof which are sentiment indicators surveyed by the European Commission. Factor analysis is used to extractthe key factor from individual questions on industry, retail and construction. To arrive at a short-term fore-cast, these factors need to be predicted (Charpin et al., 2000).

The Centre for Economic Policy Research (CEPR) has been publishing the EuroCOIN Indicator for theeuro area and for Germany, France, Italy, Spain, the Netherlands and Belgium since 2002 (www.cepr.org).Its reference parameter is seasonally adjusted, quarterly growth of real GDP. The EuroCOIN provides anestimate of the cyclical component of GDP, adjusted for measurement errors and idiosyncratic regionaland sectoral shocks. The calculation, which is based on factor analysis, includes around 1,000 monthlyseries from the six largest countries in the euro area. These series are adjusted by filters for measurementerrors and short-term fluctuations.

The European Commission has been publishing the Euro Area GDP Indicator for quarterly real GDPgrowth in the euro area since 2002. It is released for the two quarters ahead, for which neither preliminaryGDP data nor flash estimates have been released, in the form of a range (95% confidence interval based onthe standard error of the regression). The calculation includes four real variables and two financial dataseries (Grasman and Keereman, 2001). Furthermore, the European Commission has been publishingthe Business Climate Indicator (BCI) for the euro area on a monthly basis since 2000. The common com-ponent and information specific to every individual question are extracted from five individual questions onindustrial confidence (euro area aggregate) using factor analysis. Information on the driving forces behindthe business cycle can be derived from the specific components. The weblink for the Euro Area GDP In-dicator and the BCI is europa.eu.int/comm/economy_finance/indicators_en.htm.

Studies by the publishing institu-tions that compare composite indica-tors with the reference series oftenshow an impressive correspondancewith very high correlation coefficients.However, it should not be forgottenthat, at the time a new index value isbeing established, some input seriesare yet to be released and will only beadded with a time lag, or must evenbe forecast and will be substituted at

a later date. Or the input series willbe revised retrospectively, e.g. indus-trial production. Composite indicesare therefore often themselves subjectto major revisions (unlike sentimentindicators, see section 2). The correla-tion of the first release of an indexvalue can sometimes bewell below thatwhich is calculated ex post on the basisof the definitive value, and it is pre-cisely the latest indicator values that




are associated with greater uncertainty.This problem is illustrated in the box�Forecast Revisions Due to New orRevised Input Data of the Euro AreaGDP Indicator� (see p. 81). Dieboldand Rudebusch (1991) carried out aformal analysis of the forecasting qual-ity of a well-known U.S. compositeindicator (Census Bureau�s Index ofLeading Economic Indicators). Compar-ing the results of an in-sample forecastwith those of the out-of-sample forecastfor both final indicators and firstreleases shows that the inclusion ofthe composite index can reduce fore-casting errors in the first two instanceswhereas this is only partly the case forthe out-of-sample forecast based on thefirst releases. This proves that only theanalysis of real-time data can throwlight on the indicator quality of a com-posite index.

Many of the aforementioned indi-cators were created and flourished atthe end of the last century. However,evidence after 9/11 has highlightedtheir limitations. At a time when opin-ions about the reactions of both mar-kets and consumers to the attacks di-verged, signals issued by composite in-dicators were paid particular attention.Many composite indicators issued falsesignals, just as the Ifo or ZEW confi-dence indicators did, primarily be-cause of the high input weight attachedto the confidence indicators. Sincethen, the composition of composite in-dicators has received greater attention

in interpreting signals issued. Institu-tions that publish composite indicatorshave also taken action and, in severalcases, reduced the weight of the con-stituent sentiment indicators.

In a summary article on compositeindicators the ECB (2001) concludesare useful as an additional tool but can-not replace extensive coverage in eco-nomic analysis. The relationship be-tween indicators and the business cycleis frequently not stable. This is why es-pecially the latest indicator values havea limited informative quality. Althoughturning points (at least based on defin-itive indicator values) were often indi-cated early on in the past, this does notpermit conclusions to be drawn aboutthe exact date or feature of future turn-ing points especially since the lengthsof leads fluctuate strongly and false sig-nals are issued. The added value ofcomposite indicators for short-termforecasting is considered to be verylimited.

To sum up, composite indicatorsoffer an attractive tool for drawing con-clusions from different, often diver-gent signals. However, the informativequality of the latest relevant index val-ues can be reduced by input series thatare includedwith a time lag and subjectto revisions. In any case, composite in-dicators cannot replace the analyses ofindividual data series, as only these per-mit conclusions to be drawn about thedriving forces behind a trend.




Forecast Revisions Due to New or Revised Input Data

of the Euro Area GDP Indicator

The European Commission�s Euro Area GDP Indicator is published for the two quarters ahead, forwhich GDP growth data have yet to be released. It is updated on a monthly basis with the latest data avail-able. For every quarter, there are therefore six sequentially published range estimates for real GDP quar-terly growth.

A systematic record of the experience of the 14 quarters since the launch of the GDP indicator inJanuary 2002 allows the following conclusions to be reached: The current range is 0.4 percentage point(until mid-2003: 0.3 percentage point) and is therefore relatively wide. The range shifted by as muchas 0.5 percentage point within the six publications for a specific quarter. Whereas the relevant last twopublications of the range include subsequent actual GDP growth in 86% of cases, this was true for only55% of cases on average for the first three publications and this despite the relatively broad range. Allin all, the Euro Area GDP Indicator�s accuracy should be seen as being only moderate — especially forthe indicator�s initial releases.

4 PeculiaritiesThis section presents a few somewhatpeculiar indicators that are repeatedlyreferred to in the media as leading indi-cators. An example of this is the R-Word Indicator, which uses a U.S.model to measure the frequency withwhich the word �recession� appearsin the financial media. According to astudy by Bayerische Hypo-VereinsbankAG based on articles in Handelsblattand Frankfurter Allgemeine Zeitung,this measure is also a useful leading in-dicator for signaling an imminentdownturn in Germany. The R-Word

Indicator delivers a correct signal intwo out of three cases. However, the In-dicator�s causality is not entirely clear,aswriting about a recession can in itselfpush recessions.

Another indicator is based on theobservation that growing consumptionand investment are reflected early on inthe freight costs incurred by the trans-portation of raw materials and inter-mediate goods. The Baltic Dry Index(BDI), an index of freight costs onthe world�s most important shippingroutes, is considered to be a good lead-ing indicator not only for global indus-

Table 2

Comparison of Composite Indicators

Publishiginstitution

Publishedsince

Currently releasedfor the followingcountries

Frequencyof publica-tion

Number ofinput series

Revi-sions1)

Reference series Type ofpublication2)

OECD-CLI OECD 1980 23 OECD countriesand 7 economicregions

monthly 75 J Industrial production I

Euro Area EconomicIndicator

Handelsblatt 1992 Germany, euro area monthly 5 J Annualized annualGDP growth

P

Euro Growth Indicator Euroframe 2000 Germany, euro area monthly 8 J Annual GDP growth P

EuroCOIN CEPR 2002 Euro area and6 individual euro areacountries

monthly 1,000 J Quarterly GDPgrowth

Z

Euro AreaGDP Indicator

EuropeanCommission

2002 Euro area monthly 6 J Quarterly GDPgrowth

B

BCI EuropeanCommission

2000 Euro aera monthly 5 N Annual growth inindustrial production

I

1) Y = systematically retrospective revisions due to delayed publication or revisions of the input series, N = no retrospective revisions of the input series (the BCI�s historicalindex values may, however, be revised whenever the factor analysis is carried out again using the latest industrial confidence values).

2) I = publication as an index value, F = publication as a forecast for the following quarters, E= estimate of the cyclical component of GDP, R= publication as a range forecastfor the following quarters.




trial demand but also for German ex-ports. The BDI already indicated sharpswings to an extent never seen beforewhen the raw material-intensive eco-nomic boom in China and its repercus-sions for the global economy only justbegan to be discussed in the media.The BDI is calculated by the Baltic Ex-change in London, a global freight mar-ket place. Similarly, indicators of flightprices could also be mentioned here.

A final example is the LuxuryGoods Index, which is a leading indica-tor of the global economic cycle basedon the equity prices of leading luxurygoods manufacturers worldwide. Theunderlying idea is that the manufactureand sales of luxury goods are particu-larly sensitive to economic develop-ments. However, a certain qualificationneeds to be made. This indicator can-not accurately predict the strength ofan upturn, as equity prices can be influ-enced by many factors that are not nec-essarily cyclical. Hypo-Vereinsbanknonetheless uses such a luxury goodsindicator for its economic forecastsfor Germany. The indicator�s lead isabout six months, according to banksources. Of particular interest is thefact that the Luxury Goods Index — un-like the Ifo or ZEW Index — did not is-sue a false signal in 2002, indicating anupturn.

5 Economic Indicators forthe New EU MemberStates

The ten countries that joined the EUon May 1, 2004, were faced with newdata provision requirements upon ac-cession. In many cases, the preparatorygroundwork had been carried out inthe runup to EU accession. This ledto the timely release of comparabledata. In other cases, governments weregranted a period of grace, so that satis-factory data series can only be ex-

pected in the years to come. This frag-mentary availability of data and/or ina-vailability of qualitatively satisfactorydata is tapping interest in alternativeeconomic indicators. In countrieswhere the economy is still undergoinga period of radical structural change,growth is also particularly difficult toforecast.

This section looks at the availabilityof short-term economic indicators forthe new EU Member States. Althoughcertain national indicators look back toa longer history, these are not exam-ined here. Instead, an overview is pre-sented on which of the established indi-cators for the EU cover the new EUMember States in conformity with astandardized methodology. After all,a few new EUMember States will soonjoin the euro area. However, only ahandful of the established institutionsthat have long been determining lead-ing indicators for European countrieshave so far focused on this region. Eventhe CLI calculated by the OECD, ofwhich Hungary, Poland, the Czech Re-public and the Slovak Republic aremembers, has yet to extend its scopeto this group of countries. Similarly,the PMI has so far been surveying onlyPolish and Czech data using a compara-ble methodology.

Of the best-known indicators, onlytwo can be accordingly cited as positiveexamples: the European Commission�sconfidence indicator and the Consen-sus forecast. Consensus Economicshas been surveyingmore than 140 fore-casters in Central and Eastern Europeevery two months since May 1998and deriving themean values of 19 indi-vidual countries. Consensus Econom-ics therefore surveys all the new EUMember States (except for Malta)and countries to be joining the EUshortly such as Bulgaria and Romania,as well as candidate countries such as




Croatia and Turkey, not to mentionRussia and some of the former SovietFederal Republics.

The European Commission�s sur-vey on economic confidence also pro-vides full coverage of all EU MemberStates (except for Malta). The newEU Member States have been partici-pating in these surveys not only sinceMay 2004 — some of them have beendoing so since the mid-1990s. Bulgariaand Romania have also been taking partin the survey for years. On the basis ofthese data gathered by the EuropeanCommission, the question is examinedas to whether sentiment indicators inthe new EUMember States have a sim-ilar reliability and forecasting quality asindicators published for countries thathave been covered for a longer periodof time. After all, a certain experiencewith this type of survey is required byboth forecasting institutions and re-spondents. Likewise, the populationand businesses could lack the experi-ence of assessing the current and futuredevelopment of their economy cor-rectly. As regards the following analy-sis, however, it should be emphasizedfrom the outset that, owing to the re-cent availability of data in the newEU Member States (data on growthin industrial production has generallybeen released only from 1999 onward),the results should be interpreted withcaution.

The results presented in table 1help select the ESI�s subcomponentsfor the purposes of further analysis.Those ESI components that are themost heavily weighted in the ESI com-posite indicator are also found to havethe highest correlation coefficients: in-dustry, consumers and the service sec-tor. With the exception of the con-sumer confidence indicator, these indi-cators also have a relatively small lag,which means that they can be expected

to provide added value. Production ex-pectations in industry, a subcompo-nent of the industrial confidence indi-cator, are highly related and show aslight lead. The indicators in the con-struction and retail sectors, which eachhave a weight of only 5% in the ESIcomposite indicator, show both a verysmall correlation and long lags and donot signal some turning points in indus-trial production at all. The analysis be-low therefore includes the ESI, indus-trial confidence, consumer confidenceand production expectations in indus-try. The service sector is not included,as survey data in the service sector haveonly been released for the new EUMember States since 2002 — a fact thatdoes not ensure reliable analysis.

A panel data regression of growthin seasonally adjusted industrial pro-duction on sentiment indicators isnow carried out, using these data foreach of the nine new EU MemberStates (NMS) surveyed and for the re-maining 15 long-standing EU MemberStates (EU-15). The model

ðIPi;t � IPi;t�1=IPi;t�1 ¼�i þ �Ii;tþj þ �i;t;

is specifically estimated, where �i

is a country-specific constant and j awhole number at an interval of �12that is calculated as the degree of leador lag for which the fit of the model(expressed by the adjusted correlationcoefficientR2

adj) is maximized. In table3, R2

adj is marked for each indicator inthe first line, as is the degree of lead orlag of the indicator series in bracketsfor which the fit of the model is maxi-mized. In the second line, the esti-mated coefficient is indicated. For theEU-15, the table presents the resultsfor both the entire sample and the sam-ple restricted to the 1999 to 2005 pe-riod in order to take account of the factthat confidence indicators and indus-




trial production data have been re-leased for most NMS only from thisdate onward.

Finally, the forecasting quality isevaluated for each indicator and foreach of the samples, whereby it is spe-cifically analyzedwhat added value sen-timent indicators can offer for fore-casts of growth in industrial produc-tion. The methodology used is theout-of-sample approach outlined insection 2. For a specific start montha few years following the start of therelevant sample (long sample: January1995, short sample: January 2002), apanel model, which includes the previ-ous values of both industrial produc-tion growth and the indicator with dif-ferent lag structures, is estimated. Theoptimal model is the one for which thegoodness of fit (R2

adj) is maximized.This model is used to forecast growthin industrial production with a forecasthorizon of one, three, six and twelvemonths. The sample is then extendedby a month and the exercise recom-mences. This procedure is repeated un-til the full-sample estimate is reached.Whereas the long sample generates123 forecasts (with a horizon of one,three, six and twelve months, respec-tively), the short sample generates only39 forecasts.

By comparing these forecast valueswith the realized values, the root meansquared error (RMSE) can be sepa-rately calculated for each forecast hori-zon and each country, with lowerRMSE values signifying better forecast-ing quality. For each forecast horizon,the average RMSE across each groupof countries can be compared withthe corresponding RMSE of a bench-

mark model. A simple panel autore-gressive process, which forecastsgrowth in industrial production basedexclusively on its own previous values,is used here as a benchmark. For eachforecast horizon (denoted in this in-stance as � —1, � —3 etc.), the improve-ment in the RMSE between the indica-tor model and the benchmarkmodel inpercentage terms is thus a measure ofthe added value sentiment indicatorsoffer for forecasting. This figure is com-parable between groups of countriesand individual indicators.9

Table 3 presents interesting infor-mation on indicator properties, theirstability over time and a comparisonof the groups of countries. The degreeof lead or lag for which the goodness offit is maximized confirms the conclu-sion drawn from table 1, according towhich only the production expecta-tions for the EU-15 are actually a lead-ing indicator. However, even if the ESIand the industrial confidence indicatorare coincident or lagging slightly, theycan provide additional information onaccount of their date of publication,which is six weeks earlier. By contrast,the quality of the consumer confidenceindicator is heavily dependent on thesample, which mostly has a strong lag.It is clear from the goodness of fit thatthe degree of fit is maximized for in-dustrial confidence, followed by ESIand production expectations. How-ever, which indicator is actually bestsuited to forecasting growth in indus-trial production can be assessed onlyon the basis of the aforementionedcomparison of forecasting quality withthat of its benchmark model. Thisstudy confirms across all the samples

9 Diebold and Mariano (1995) propose a test to check whether RMSE values actually differ from each other signifi-cantly. However, the benefit of this test is doubtful since, in practice, one always falls back on the model with thelower RMSE — even for non-significant differences. See Kunst (2003) for a critical view of tests investigating rel-ative forecasting quality.




that the forecasting quality of industrialconfidence indicators for the forecastoutstrips that of ESI production ex-pectations rank in third place.10 Fore-casts based on consumer confidenceperform worst and are frequently out-performed even by simple autoregres-sive forecasts.

The two samples for the EU-15 il-lustrate the extent to which indicatorproperties have changed over time. Ta-ble 3 shows that, for the first two indi-cators, the lag has disappeared in theshorter sample. Likewise, the con-sumer confidence indicator now has aslight lead. This could suggest an im-provement in indicator properties.For the shorter sample, however, thecoefficient of determination tends tobe lower. This outcome is shown inchart 3, in which the coefficient of de-termination for the panel regression ofgrowth in industrial production on the

coincident ESI value is shown in a mov-ing five-year window. According tothis, R2

adj attained a trough in the five-year window from 1994 to 1998 andstagnated thereafter for two and halfyears at a low level before subsequentlystabilizing again at a higher level. Table3 also shows that the added forecastingvalue of the indicators is higher in thelonger sample. This is likely to be con-nected with the fact that economicfluctuations in 2002 and 2003 were his-torically particularly poorly signaledby indicators in view of 9/11 and thatthis period is prominently representedin the short sample. The aforemen-tioned pecking order of the indicatorsaccording to their forecasting qualityremains stable over time.

The comparison of both groups ofcountries (EU-15 and NMS) with acomparable sample length shows thatcoefficients and coefficients of deter-

10 Although these and the following statements apply to forecasts with a horizon of one, three and six months, they donot always hold for the twelve-month forecast. However, the forecasting errors for these long-term forecasts are veryhigh. This could be connected with seasonal patterns that have not been entirely remedied or the fact that indicatorsonly contain information for a shorter period ahead.

Table 3

Regression Comparison of EU-15 and New EU Member States (NMS)

Coefficient EU-15 (1985) EU-15 (1999) NMS (1999)

ESI R2adj1) 0.31 (+1) 0.27 (0) 0.26 (0)Coefficient2) 0.55* 0.52* 0.52*

�-1/ �-3/ �-6/ �-123) 12.5 / 12.8 / 14.0 / 14.7 3.4 / 10.0 / 11.2 / 5.0 �1.4 / 0.7 / 2.1 / 11.2

Manufacturing R2adj1) 0.34 (+1) 0.31 (0) 0.30 (0)Coefficient2) 0.58* 0.56* 0.55*

�-1/ �-3/ �-6/ �-123) 15.3 / 16.3 / 17.4 / 18.4 10.2 / 11.1 / 11.8 / 10.1 2.5 / 5.5 / 5.2 / 10.9

Consumer R2adj1) 0.12 (+3) 0.12 (�1) 0.08 (+7)Coefficient2) 0.35* 0.36* 0.30*

�-1/ �-3/ �-6/ �-123) 4.8 /�0.5 / 0.8 / 2.9 �5.3 /�1.6 / 0.6 /�6.1 �5.9 /�12.9 /�12.6 /�4.1

Production expectations R2adj1) 0.30 (�1) 0.25 (�1) 0.10 (0)Coefficient2) 0.56* 0.50* 0.34*

�-1/ �-3/ �-6/ �-123) 9.6 / 11.0 / 12.1 / 13.5 0.4 / 7.0 / 7.9 / 6.0 �2.5 /�5.8 /�1.1 / 5.0

1) Maximum coefficient of determination of the panel regression of growth in industrial production on the indicator. The degree of the time lead/lag (in months) of theindicator relative to the reference series, for which themaximum coefficient of determination is reached, is indicated in brackets. A negative value implies a time leadof the indicator. All series are standardized in a way such that they have mean 0 and standard deviation 1.

2) * represents significance at the 1% level.3) Improvement of the RMSE of a forecastmodel including the relevant indicator compared with the RMSE of the panel autoregressive process, in percentage terms, foreach forecast horizon of one, three, six and twelvemonths. A negative value stands for a forecastmodel that is outperformed by the panel autoregressivemodel. Startmonth for the forecast analysis is for the long sample January 1995 and for the short sample January 2002.




mination are for the most part not sig-nificantly different, but that the leadingproperties in the NMS are worse forproduction expectations and that thelag is longer for consumer confidence.In just the same way, the added valueof the indicators is lower. Frequently,forecasts are even outperformed bythe simple panel autoregressivemodel.However, the pecking order of the indi-vidual indicators is also maintained inthe NMS.

An explanation for the lower addedvalue of NMS indicators could be thatboth respondents and forecasting insti-tutions in the NMS still partly lack ex-perience with such surveys owing tothe indicators� more recent history.This may reflect initial technical sur-veying and processing difficulties, aswell as the fact that both people andbusinesses in the NMS still lack experi-ence in assessing the current and futuredevelopment of their economy owingto the radical transition in the past15 years. In addition, many of the fore-casting institutions emerged from for-mer government institutions and, inmany cases, are still not considered asfully independent of government influ-

ence. In contrast, the shorter lead ofproduction expectations could be dueto the fact that, in the NMS, businessesthat are highly specialized in exportingintermediate products to their westernEuropean parent companies respondparticularly quickly and flexibly tochanges in demand conditions. As a re-sult, the lead is shorter than is the casefor traditional trade cooperations.

To sum up, the results indicate thatthe industrial confidence indicatorthroughout has the highest forecastingquality for growth in industrial produc-tion, followed by ESI. In both cases, theindicators are coincident or slightly lag-ged but do provide added value on ac-count of their earlier date of release.The analysis of the regional compari-son should be interpreted with cautionin the NMS due to their relatively re-cent history of data gathering.Whereasstatements on the pecking order of in-dicator properties also apply to thisgroup of countries, the forecast analy-sis suggests that the added value ofNMS indicator forecasting is not equalto that of countries that have been con-structing sentiment indicators formany years.

��1�$��)�$��=��.��!�� %�4;

�!+��!��!�� !��!�� '

��

��

��

��

��

��

��

��""��"��

��

��

��

+��"��"�&�9��@��@

4�� 4��

4�� 4��

4�� 4��=

4�� 4��>

4�� 4��

4�� 4��

4�� 4��

4�� 4��

4�� 4��

4�� 4��

4��




6 SummaryThis contribution provides an over-view of the most common short-termindicators of economic developmentin the euro area. First, indicators arepresented that are compiled from (in-ternational or national) surveys of busi-nesses, households, financial analystsand forecasters. The advantages of indi-cators obtained from such surveys aretheir early date of release, theirmonthly basis of publication, the factthat they are largely not subject to revi-sions, and their provision of detailed in-formation on each sector. Most surveyindicators are well-established and area popular component of economicanalysis due to their simple structure.Second, a broad spectrum of compo-site indicators, which are calculatedby combining a variety of measuresinto a single indicator with the helpof regression and factor analysis, waspresented. They offer an attractive toolfor drawing conclusions from different,often divergent signals. However, theirlimitation lies in the substantial revi-sions to which they are sometimes sub-ject due to input series being includedwith time lags. As a result, the informa-tive value of the latest relevant signals islimited.

Even the most reliable economicindicators, however, can only be inter-preted as constituent elements of com-prehensive economic analyses. Afterall, post-9/11 experience has shownnot least that economic indicatorssometimes issue false signals at thevery moment when uncertainty aboutfuture economic development is at apeak. Recently, the debate aboutwhether the service sector, which is

steadily growing in weight, is reducingthe informative value of leading indica-tors, is becoming increasingly vehe-ment. As a result, indicators, whichare primarily based on manufacturingindustry data, could lose their repre-sentativeness. The selection of leadingindicators should take account of theirlead quality, date of publication, the oc-currence of retrospective revisions towhich they are subject and the breadthof their economic basis.

Only a handful of these economicindicators also survey the countriesthat have become EU Member Statessince May 2004. Only Consensus Eco-nomics and the European Commis-sion�s Economic Sentiment Indicatoroffer full coverage. This is not leastlikely to be due to the fact that, inmanyof these countries, experienced fore-casting institutions that could conductthese surveys are still scarce. Thisstudy also shows that the NMS surveyresults released by the European Com-mission are not fully comparable withthose of the EU-15. For instance, theadditional forecasting quality of senti-ment indicators in the NMS is lower,and some of the lead properties differbetween the NMS and the EU-15. Thiscould have something to do with theNMS� relatively recent experiencewith business and consumer surveys.After all, forecasting institutions andrespondents require experience incompiling reliable and stable indica-tors. This is what makes the appeal toestablished institutions to extend theirsurveys early on to cover the NMS — ofwhich a few could soon be joining theeuro area — all the more important.




ReferencesBatchelor, R. A. 2001. How Useful Are the Forecasts of Intergovernmental Agencies? The IMF and

OECD versus the Consensus. In: Applied Economics 33. 225—235.

Bewley, R. and D. G. Fiebig. 2002. On the Herding Instinct of Interest Rate Forecasters. In: Empirical

Economics 27(3). 403—426.

Blix, M., J. Wadefjord, U. Wienecke and M. Aß dahl. 2001. How Good Is the Forecasting

Performance of Major Institutions? In: Sveriges Riksbank Economic Review 2001:3. 38—68.

Burns, A. F. and M. C. Mitchell. 1946.Measuring Business Cycles. In: NBER Studies in Business Cycles 2.

Charpin, F., H. Pe«le«raux and P. Sigogne. 2000. A New Simpler EMU Indicator. OFCE Analysis and

Forecast Department (mimeo).

Deroose, S., P. Mills and B. Saint Aubin. 2001. Business Climate Indicator for the Euro Area. Paper

prepared for CEPR Banca d�Italia Conference on �Monitoring the euro area business cycle�. Rome.

6 to 7 September 2001 (mimeo).

Diebold, F. X. and G. D. Rudebusch. 1991. Forecasting Output with the Composite Leading Index:

A Real-time Analysis. In: Journal of the American Statistical Association 86. 603—610.

Diebold, F. X. and R. S. Mariano. 1995. Comparing Predictive Accuracy. In: Journal of Business and

Economic Statistics 13. 253—263.

Dreger, C. and C. Schumacher. 2005. Out-of-sample Performance of Leading Indicators for the Ger-

man Business Cycle. Single vs Combined Forecasts. In: Journal of Business Cycle Measurement and

Analysis 2(1). 71—88.

Emerson, R. A. and D. F. Hendry. 1996. An Evaluation of Forecasting using Leading indicators. In: Jour-

nal of Forecasting 15. 271—291.

European Commission. 2000. Business Climate Indicator for the Euro Area (Presentation Paper).

November: europa.eu.int/comm/economy_finance/indicators_en.htm

European Commission. 2004a. The Joint Harmonised EU Programme of Business and Consumer Sur-

veys — User Guide. European Commission. Directorate General Economic and Financial Affairs. May:

europa.eu.int/comm/economy_finance/indicators_en.htm

European Commission. 2004b. Methodological Changes of the Economic Sentiment Indicator:

Inclusion of the Services Survey and Overall Streamlining of its Calculation. European Commission.

Directorate General Economic and Financial Affairs. May: europa.eu.int/comm/economy_finance/

indicators_en.htm

ECB. 2001. Economic Indicators� Information Content of the Business Cycle in the Euro Area. In: ECB

Monthly Report. November. 45—57.

ECB. 2004a. Surveys on Business, Price and Labor Market Trends in the Euro Area — Features and

Benefits. In: ECB Monthly Report. January. 61—73.

ECB. 2004b. The Monetary Policy of the ECB. Frankfurt am Main: European Central Bank.

Fenz, G., M. Schneider and M. Spitzer. 2005. The Economic Indicator of the OeNB: Methods and

Forecasting Performance. Proceedings of OeNB Workshops. Macroeconomic Models and Forecasts

for Austria. 11 to 12 November 2004. 126—147.

Forni, M. and L. Reichlin. 1998. Let�s Get Real: A Factor Analytical Approach to Disaggregated Business

Cycle Dynamics. In: Review of Economic Studies 65. 453—473.

Forni, M. and M. Lippi. 1999. The Generalized Dynamic Factor Model: Representation Theory. Univer-

site« Libre de Bruxelles. Working Paper 132.

Forni, M., M. Hallin, M. Lippi and L. Reichlin. 2000. The Generalized Dynamic Factor Model: Iden-

tification and Estimation. In: The Review of Economics and Statistics 82(4). 540—554.

Garcia, J. A. 2003. An Introduction to the ECB�s Survey of Professional Forecasters. ECB Occasional

Paper Series 8. September.




Geweke, J. 1977. The Dynamic Factor Analysis of Economic Time Series. In: Aigner, D. and A. Goldberger

(eds.). Latent Variables in Socio-economic Models. Amsterdam: North Holland. 365—383.

Grasmann, P. and F. Keereman. 2001. An Indicator-based Short-term Forecast for Quarterly GDP in

the Euro Area. European Communities Economic Paper 154. June.

Hu‹ fner, F. P. and M. Schro‹der. 2002. Forecasting Economic Activity in Germany — How Useful Are

Sentiment Indicators? ZEW Discussion Paper 02—56.

Inoue, A. and L. Kilian. 2004. In-sample Or Out-of-sample Tests of Predictability: Which One Should

We Use? In: Econometric Reviews 23(4). 371—402.

Kunkel, A. 2003. Zur Prognosefa‹higkeit des ifo Gescha‹ftsklimas und seiner Komponenten sowie die U‹ ber-

pru‹fung der ªDreimal-Regel�. Ifo Diskussionsbeitra‹ge 80. March.

Kunst, R. 2003. Testing for Relative Predictive Accuracy: A Critical Viewpoint. Institute of Advanced Stud-

ies. Economic Series 130. May.

Marcellino, M. 2006. Leading Indicators: What Have We Learned? In: Elliott, G., C. W. J. Granger and

A. Timmermann (eds.). Handbook of Economic Forecasting. North Holland (forthcoming).

NTC Research. 2002. Early, Accurate, Monthly GPD Estimates — A PMI-based Model of Gross Domes-

tic Product. Economic Research Monograph 3. NTC Research. May.

Sargent, T. J. and C. A. Sims. 1977. Business Cycle Modeling Without Pretending to Have Too Much A

Priori Economic Theory. In: C. A. Sims (eds.). New Methods in Business Cycle Research. Minneapolis:

Federal Reserve Bank of Minneapolis.

Schneider, M. and M. Spitzer. 2004. Forecasting Austrian GDP Using the Generalized Dynamic Factor

Model. OeNB Working Paper 89. August.

Vanhaelen, J., L. Dresse and J. De Mulder. 2000. The Belgian Industrial Confidence Indicator:

Leading Indicator of Economic Activity in the Euro Area? National Bank of Belgium Working

Paper 12. November.

Zarnowitz, V. 1984. The Accuracy of Individual and Group Forecasts from Business Outlook Surveys.

In: Journal of Forecasting 3(1). January.




an overview of european economic indicators ...5caf75ee-6277-4f29-92ab-013a6c80867b/mop... · this...

Documents