COMMISSION STAFF WORKING PAPER SCOREBOARD FOR THE SURVEILLANCE OF MACROECONOMIC IMBALANCES: ENVISAGED INITIAL DESIGN

EUROPEAN COMMISSION Brussels, 8.11.2011 SEC(2011) 1361 final COMMISSION STAFF WORKING PAPER SCOREBOARD FOR THE SURVEILLANCE OF MACROECONOMIC IMBALANCES: ENVISAGED INITIAL DESIGN EN EN

Inroducion Agreemen has been reached beween he European Parliamen and he Council on he legislaive proposals o enhance economic governance in he EU, including on he regulaions o esablish a surveillance procedure o preven and correc macroeconomic imbalances. These regulaions are expeced o ener ino force by he end of 2011, in ime for he sar of he nex European semeser. The new regulaions provide for he drawing up of a scoreboard o provide an early-warning signalling device of poenially harmful macroeconomic imbalances in Member Saes. As oulined in he relevan recial of he regulaion, he scoreboard will be se up in close cooperaion wih he Council and he Parliamen. In order o allow he swif applicaion of he new regulaions on macroeconomic imbalances as soon as hey ener ino force, his Commission Saff Working Paper presens he envisaged design of he iniial scoreboard of early-warning indicaors which is fully in line wih he relevan aricle of he regulaion on he prevenion and correcion of macroeconomic imbalances. In accordance wih he relevan recial of he regulaion, he envisaged design of he scoreboard will be presened o he compeen commiees of he European Parliamen and he Council. However, boh insiuions have already been involved in he echnical discussions on he design of he scoreboard as a par of he rilogue negoiaions, and on a echnical basis in he Economic Policy Commiee and he Economic and Financial Commiee. Afer he discussion wih Parliamen and Council, he scoreboard design will be finalised. A scoreboard o faciliae he early idenificaion of poenial macroeconomic imbalances In he new excessive imbalances regulaion, he scoreboard indicaor resuls are subjec o an economic reading by he Commission in is Aler Mechanism Repor (AMR). The economic reading will be based on he resuls of he scoreboard bu here is no auomaiciy. The overall number of breaches of hresholds, he severiy of individual breaches as well as he combinaion of breaches, poenially signalling broad based problems, will all be aken ino accoun. Moreover, he AMR will draw, when appropriae, on oher relevan economic and financial indicaors wih a view o ge a more complee picure. Based on he economic reading, he Commission will idenify he Member Saes i considers may be affeced by, or may be a risk of being affeced by imbalances. Taking due accoun of discussions in he Council and he Euro group on he AMR, he Commission will deermine he lis of Member Saes where furher in-deph reviews are needed. I is hese in-deph reviews, and no he AMR, which provide he basis for any recommendaions o be addressed o counries under he prevenive or correcive arm of he new excessive imbalances regulaion. Principles underpinning he design of he scoreboard Firs, he choice of indicaors should focus on he mos relevan dimensions of macroeconomic imbalances and compeiiveness losses, wih a paricular focus on he smooh funcioning of he euro area. For his reason, i is envisaged ha he scoreboard consiss of indicaors which can monior exernal balances, compeiiveness posiions and inernal imbalances, and encompass variables where boh he economic lieraure and recen experiences sugges associaions wih economic crises. Secondly, he scoreboard (indicaors and hresholds) need o provide reliable signalling device for poenially harmful imbalances and compeiiveness losses a an early sage of EN 2 EN

heir emergence. This would jusify having a combinaion of sock and flow indicaors which can capure boh shorer-erm rapid deerioraions as well as he longer erm gradual accumulaion of imbalances. Moreover, indicaive hresholds should be se a pruden levels, which on he one hand avoid excessive numbers of 'false alarms', bu which on he oher hand are no se so sringenly ha hey only idenify problems once hey are enrenched. To his end, hresholds should be esablished wih a saisical approach based on he disribuions of he indicaors' values idenifying he hresholds as he lower and/or upper quariles of he disribuions: such hresholds are consisen wih he values found in some empirical sudies in he available economic lieraure. Thirdly, scoreboard has an imporan communicaion role. For his purpose, he scoreboard should consis of a limied number of indicaors. Moreover, he choice of indicaors and ransformaions should be kep as simple and sraighforward as possible. Daa ransformaion should be ransparen and racable so ha hey can be replicaed by hird paries. The choice of indicaors should complemen indicaors/arges used in oher EU surveillance exercises. The scoreboard indicaors will be made publicly available on he Commission's websie. Fourh, he indicaors should be of high saisical qualiy in erms of imeliness and comparabiliy across counries. To his end, hey should be derived from daa compiled according o he principles of he European Saisics Code of Pracice of he European Saisical Sysem (ESS). Where available, EUROSTAT sources should be used so ha he daa comparabiliy and saisical qualiy can be assured. Oherwise, when EUROSTAT daa are no available, he highes qualiy alernaive daa source should be chosen (e.g. he ECB). Envisaged design of he scoreboard 3 year backward moving average of he curren accoun balance as a percen of, wih he a hreshold of +6% of and - 4% of ; ne inernaional invesmen posiion as a percen of, wih a hreshold of -35% of ; 5 years percenage change of expor marke shares measured in values, wih a hreshold of -6%; 3 years percenage change in nominal uni labour cos, wih hresholds of +9% for euroarea counries and +12% for non-euro-area counries. 3 years percenage change of he real effecive exchange raes based on HICP/CPI deflaors, relaive o 35 oher indusrial counries, wih hresholds of -/+5% for euro-area counries and -/+11% for non-euro-area counries; privae secor deb in % of wih a hreshold of 160%; privae secor credi flow in % of wih a hreshold of 15%; year-on-year changes in house prices relaive o a Eurosa consumpion deflaor, wih a hreshold of 6%; general governmen secor deb in % of wih a hreshold of 60%; 3 year backward moving average of unemploymen rae, wih he hreshold of 10%. The economic reading of he scoreboard All indicaors should be examined for heir conribuion o problemaic macroeconomic imbalances. In paricular, surveillance of public deb under he new regulaion will be limied o is conribuion o macroeconomic imbalances as he susainabiliy of public finances is already covered by he SGP. The indicaor on he evoluion of unemploymen shall be read in EN 3 EN

conjuncion wih oher more forward-looking scoreboard indicaors and be used o beer undersand he poenial severiy of macroeconomic imbalances in erms of heir likely persisence and he capaciy of he economy o adjus. The criical imporance of aking due accoun of counry specific circumsances and insiuions should be recognised when underaking an economic reading of he scoreboard indicaors. In addiion o indicaors in he scoreboard, addiional indicaors and consideraions should be aken ino accoun in line wih he regulaion. This will include he need o caer for nominal and real convergence inside and ouside he euro area. In paricular, he ne exernal deb as well as he share of foreign direc invesmen should be aken ino accoun as imporan addiional informaion in he economic reading of he scoreboard. Also, he capial accoun balances will be considered for Member Saes where hey are imporan. Similarly, oher elemens will be aken ino accoun such as alernaive indicaors of privae secor deb based on consolidaed daa, where available, or alernaive indicaors providing more insighs ino poenial imbalances accumulaing in housing or financial markes. Also broader indicaors of developmens in he real economy such as differen measures of produciviy developmens should be considered as hey reflec he poenial for he developmen of imbalances as well as he adjusmen capaciy of an economy. The imporance of he role of financial markes in he curren crisis is recognised in he scoreboard. Some indicaors, noably on credi developmens and house prices, reflec his. I is envisaged o develop a wider indicaor of he banking/financial secor by he end of 2012 and in ime for he subsequen European semeser. EN 4 EN

Table 1. Envisaged indicaors and indicaive hresholds (*) Indicaor Daa source Indicaive hresholds Period for calculaing hresholds Some addiional indicaors o be used in economic reading Noes: 3 year average of curren accoun balance as a % of Balance of Paymens saisics EUROSTAT. Exernal imbalances and compeiiveness Ne Inernaional Invesmen Posiion as a % of Balance of Paymens Saisics, EUROSTAT. +6/-4% -35% Lower quarile 1970-2007 Ne lending/borrow ing vis-à-vis ROW (Capial Accoun + Curren Accoun as % of ) Firs available year (mid- 1990s)-2007 Ne Exernal Deb as % % change (3 years) of Real Effecive Exchange Rae, HICP deflaors relaive o 35 indusrial counries (a) DG ECFIN indicaor daa base on Price and Cos compeiiveness. +/-5% for A +/-11% non A Lower and Upper Quariles of EA - /+ s.d. of EA % change (5 years) in expor marke shares Balance of Paymens saisics, EUROSTAT. -6% Lower quarile % change (3 years) in nominal uni labour cos (b) EUROSTAT +9% A +12% non- A Upper Quarile A3 p.p y-o-y % change in deflaed house prices (c) Harmonised house price index by EUROSTAT, compleed wih ECB, OECD and BIS daa. +6% Upper quarile privae secor credi flow as % of (d), (e) Transacions AFA, EUROSTAT for annual daa and QSA, ECB for quarerly daa. Inernal imbalances privae secor deb as % of (d), (e) Balance Shee AFA, EUROSTAT for annual daa and QSA, ECB for quarerly daa. general governmen deb as % of (f) EUROSTAT (EDP reay definiion). 3 year average of unemploymen rae EUROSTAT LFS daa. EN 5 EN +15% Upper Quarile 160% Upper Quarile +60% +10% 1995-2007 1995-2007 1995-2007 1995-2007 1994-2007 1994-2007 REER vis-à-vis res of he euro area Expor marke shares based on volumes of goods; Labour produciviy; Trend TFP growh Nominal ULCs (changes over 1, 5, 10 years); Effecive ULC relaive o res of euro-area Oher measures of produciviy Real house price (changes over 3 years); Nominal house price (changes over 1 and 3 years) Residenial consrucion Indicaor on change in financial liabiliies of he nonconsolidaed financial secor and he deb over equiy raio Privae secor deb based on consolidaed daa (a) for EU rading parners HICP is used while for non-eu rading parners, he deflaor is based on a CPI close o he HICP in mehodology; (b) index providing raio of nominal compensaion per employee o real per person employed; (c) changes in house prices relaive o he consumpion deflaor of EUROSTAT; (d) privae secor is defined as non-financial corporaions; households and non-profi insiuions serving households; (e) sum of Loans, and Securiies oher han shares; liabiliies, non consolidaed; (f) he susainabiliy of public finances will no be assessed in he conex of he EIP given ha his issue is already covered by he SGP. However his indicaor is par of he scoreboard because public indebedness conribues o oal indebedness of he counry and herefore o he overall vulnerabiliy of he counry. (*) I is envisaged o develop a wider indicaor of he banking/financial secor by he end of 2012.

Formulas for he indicaors' ransformaions Indicaors Formulas for daa ransformaion 3 year backward moving average of CA CA CA + + CURRENT ACCOUNT BALANCE 1 2 as a % of * 100 3 NET INTERNATIONAL INVESTMENT POSITION as a % of NIIP % change (3 years) of REAL EFFECTIVE EXCHANGE RATE wih HICP deflaors relaive o 35 oher indusrial counries (a) ( REER _ HICP _ 35) ( REER _ HICP _ 35) ( REER _ HICP _ 35) 3 3 Noes: % change (5 years) in EXPORT MARKET SHARES % change (3 years) in NOMINAL UNIT LABOUR COST (b) y-o-y % change in DEFLATED HOUSE PRICES (c) PRIVATE SECTOR CREDIT FLOW as % of (d), (e) PRIVATE SECTOR DEBT as % of (d), (e) GENERAL GOVERNMENT DEBT as % of 3 year backward moving average of UNEMPLOYMENT RATE EXPc EXPc EXPworld EXP EXPc EXPworld 5 ( ULC) ( ULC) ( ULC) 3 HPI HPI DEFL DEFL HPI DEFL 1 PSCF PSD 1 GGD world 3 1 1 5 + + 1 2 ( UR) ( UR) ( UR) 3 a) For EU rading parners HICP is used while for non-eu rading parners, he deflaor will be based on a CPI close o he HICP in mehodology ; (b) index providing raio of nominal compensaion per employee o real per person employed; (c) changes in house prices relaive o he consumpion deflaor of EUROSTAT; (d) privae secor is defined as non-financial corporaions; households and non-profi insiuions serving households; (e) sum of Loans, and Securiies oher han shares ; liabiliies, non consolidaed. EN 6 EN

Annex 1 A descripion of each indicaor Curren accoun balance as a percen of The scoreboard should include an indicaor on curren accoun balance as a percen of. This is a widely known and undersood indicaor, which has been exensively explored in economic lieraure. I is envisaged ha he ransformaion for he scoreboard is calculaed as he hree years backward moving average of he curren accoun balance as a percen of so as o conrol for shor-erm flucuaions of he annual figures. In addiion, he laes available daa should be used in he economic reading of he scoreboard. Daa on he curren accoun balance are derived from he Balance of Paymens saisics repored by EUROSTAT. The simple saisical disribuion analysis provides defici hresholds of -4%. This value is also broadly in line wih he evidence from he empirical lieraure on balance of paymen crises and susainabiliy of curren accoun imbalances. The upper value of he hreshold is envisaged o be se a +6% of. In he economic reading of he scoreboard, addiional informaion should be aken ino accoun ha migh have bearing on correc inerpreaion of he indicaor, in paricular he specificiies of caching up economies. In his respec, he sum of curren accoun and capial accoun should be explicily considered for Member Saes for which his informaion is relevan 1. Ne inernaional invesmen posiion (NIIP) as a percen of To complemen he flow indicaor on curren accoun balance wih a sock indicaor, he inernaional invesmen posiion (NIIP) as a percen of should be included in he scoreboard. This is an imporan indicaor of he ne asse posiion of he domesic secors of he economy vis-à-vis he res of he world. Daa on he ne inernaional invesmen posiion are derived from he Balance of Paymens saisics repored by EUROSTAT, i.e. he same daa source used for he curren accoun balance. The saisical analysis gives -35% of as an indicaive hreshold. Also in his case, accoun should be aken of addiional relevan informaion in he economic reading of he scoreboard. In order o caer for he specificiies of exernal posiions of caching up Member Saes and in paricular he FDI inflows, oher appropriae indicaors will be explicily considered, in paricular he Ne Exernal Deb (NED) for counries for which his informaion is available 2. Real effecive exchange rae (REER) based on he harmonised index of consumer prices deflaors (HICP) As a measure of persisen changes in price compeiiveness relaive o he major rading parners of he respecive counry, an indicaor on he real effecive exchange rae (REER) based on he harmonised index of consumer prices deflaors (HICP/CPI) should be included 3. This indicaor 1 2 3 Concepually, he sum of curren accoun and capial accoun deermines he ne lending/borrowing of a counry and is hus he flow counerpar of he ne foreign financial asse posiion/ne inernaional invesmen posiion. While ypically small, capial accoun balance can be non-negligible in a number of caching up Member Saes as a par of srucural/cohesion funds is recorded here. By focusing on exernal deb liabiliies, i.e. hose ha require paymens of principal and/or ineres, i is argued ha NED beer reflecs he riskiness of a counry's exernal asse posiion. For EU rading parners HICP is used while for non-eu rading parners, he deflaor is based on a CPI close o he HICP in mehodology. EN 7 EN

will capure global price-compeiiveness developmens. I is no mean o rigger discussions on he euro exchange rae policy. I is defined as he 3-year percenage change of he REER relaive o 35 oher indusrial counries and is calculaed by DG ECFIN as par of he daa base on price and cos compeiiveness indicaors 4. The advanage of such a daa ransformaion is ha i is no heavily biased by he rend appreciaion ha some caching-up Member Saes have experienced in he laes decade. In line wih he Task Force repor, a differeniaion in he indicaive hresholds beween euro area and non-euro area counries is envisaged o reflec greaer nominal exchange rae variabiliy in he laer group. The hresholds corresponding o he lower and upper quariles of he disribuion over he sample including euro-area counries are -5% and +5% respecively. To accoun for he nominal exchange rae variabiliy in non-euro area counries, he indicaive hresholds for he euro-area counries were exended by 6 p.p. which corresponds o one sandard deviaion (6%) of he disribuion wihin he euro-area. This approach gives hresholds of -11% and +11%. For euro area counries, an addiional indicaor measuring REER wih respec o he oher 16 euro area counries should be considered in he economic reading of he scoreboard. Addiionally, he possibiliies o monior REER developmens in a broader group of compeiors ha would also include emerging economies such as China, India and Brazil will be explored ogeher wih he ECB. Expor marke shares The scoreboard should include an indicaor on he percenage change in expor marke shares o measure slow and persisen losses in compeiiveness. This is an oucome indicaor ha in a synheic way capures componens of compeiiveness such as non-price compeiiveness or he abiliy o exploi new sales opporuniies due o increased demand of emerging economies. To capure he srucural losses in compeiiveness ha can accumulae over longer ime periods, he indicaor is calculaed as he percen variaion over 5 years of he value of expors of goods and services of each counry as share of he world expors of goods and services. The indicaor is based on he Balance of Paymens saisics from Eurosa. The saisical disribuion of he indicaor gives a lower hreshold of -6%. The proposal o use value of expors (as opposed o volume of expors) is based on a careful consideraion of pros and cons of boh measures. Using expor marke shares in volumes has an advanage ha i excludes changes in expor marke shares due only o relaive expor prices developmens. On he oher hand, his daa is only direcly available for goods bu no services (due o limied daa availabiliy of world's service expors in real erms). In addiion, also changes in expor marke shares for goods in volumes should be invesigaed as par of is background analysis o suppor he economic reading of he scoreboard. When a reliable indicaor on expor marke shares in volumes becomes available, i could replace he exising indicaor in he scoreboard. Uni labour coss An indicaor should be included on percenage change in nominal uni labour cos (ULC) as a measure of cos compeiiveness. ULC measures he average cos of labour per uni of oupu. I is a widely used indicaor which provides a direc link beween coss and produciviy. The envisaged indicaor is calculaed as he 3-year percenage change of he raio of nominal 4 hp://ec.europa.eu/economy_finance/db_indicaors/compeiiveness/index_en.hm EN 8 EN

compensaion per employee o real per person employed. The original daa on nominal compensaion per employee, and employmen derive from EUROSTAT. The hree-year change conrols for he cyclical behaviour of his indicaor and keeps memory of he compeiiveness losses building up more han he year-on-year variaion. The hreshold corresponding o he upper quarile of he saisical disribuion over he sample of euro-area counries is 9%. For non-euro area counries, a hreshold of 12% is obained by adding 3 p.p. o he euro-area hreshold. This differeniaion is jusified by he fac ha he majoriy of non-euro area counries have experienced a major rade liberalisaion in he period covered by our daa (since 1995) which enails a naural process of facor price equalisaion owards he levels of he rade parners 5. In he economic reading of he scoreboard, developmens over shorer (1 year) and longer ime periods (5 or 10 years) should be analysed. In addiion, divergences in nominal uni labour coss developmens among euro-area counries as measured by he effecive ULC relaive o he res of euro-area counries will be considered. The House Price Index (HPI) The scoreboard should include an indicaor of house prices developmens as housing marke have figured prominenly in many of he previous financial crises and proved o be an imporan source of macroeconomic imbalances. They have a poenial impac on growh and oher macroeconomic developmens hrough wealh effecs, invesmen decisions, iner-secoral subsiuion effecs and he correlaion wih credi flucuaions. Agains his background, he envisaged indicaor is he year-on-year change in real house prices, more specifically he house price indicaor relaive o a Eurosa consumpion deflaor. The Eurosa house price index (HPI) will be used as he daa source as soon as i is available for all Member Saes (a subsanial number of Member Saes already provide HPIs and full coverage is expeced a he laes as from he hird quarer of 2012). Meanwhile, he ECB Residenial Propery Prices (RPP) will be used o complemen he missing daa. An indicaive hreshold of house price increases of 6% will be used, which is derived from boh a saisical mehod and he economic lieraure. As par of he economic reading of he scoreboard, he real house price increases over longer ime periods (e.g. 3 years) should be considered o complemen he shor-run indicaor. Moreover, due aenion will be paid o oher measures ha can help beer inerpre he developmens in real house prices such as nominal house price indexes, volume indicaors, in paricular he indicaors of residenial consrucion and value-added in consrucion as a share of, or price-o-income and price-o-ren raios. Privae secor credi flow (ransacions) To complemen he privae deb indicaor (see below), an indicaor on credi flows o he privae secor should be included. High credi flows appear o be one of he bes indicaors o predic crises incidence early on. Large credi flucuaions can be very ofen associaed wih boom and bus cycles in asse markes, poenial banking sysem vulnerabiliies, house price bubbles and also curren accoun imbalances. The scoreboard indicaor will capure privae secor credi flows (ransacions) as % of, and i includes loans and securiies oher han shares (as in he case of privae secor deb, he subcaegory "oher accouns receivable/payable" is no included). The daa sources will be he same as in he case of privae deb indicaor and he indicaor will for he 5 Therefore adding a 1% margin for each year o non-euro-area Members Saes should accoun for such issues. EN 9 EN

ime being be based on non-consolidaed daa. The hreshold of privae secor credi flow is 15% of which corresponds o he upper quarile of he disribuion of he indicaor. In he economic reading of he scoreboard, i could be useful o look a wider indicaors of he financial marke developmens, including he change in financial liabiliies of he nonconsolidaed financial secor and he deb over equiy raio. Also he change in privae secor deb sock should be analysed in he economic reading of he scoreboard. Given ha boh price and volumes effecs deermine variaions in deb socks, Eurosa could be invied o look furher ino he possible disenangling of price and volume effecs, given ha he quanificaion of price effecs may provide useful informaion. Privae secor deb An indicaor on privae secor deb should be included as excessively high levels of deb imply significan risks for growh and financial sabiliy and increase overall vulnerabiliy of a counry. The envisaged indicaor measures privae secor 6 deb level as a % of and is calculaed as he sum of loans and securiies oher han shares; he subcaegory "oher accouns receivable/payable" is no included. The sources of daa are he annual financial accouns and balance shees (AFA) colleced by Eurosa and he quarerly financial accouns (QFA) colleced by he ECB. The hreshold of privae secor deb is 160% of derived from he saisical disribuion of he indicaor. The envisaged indicaor is based on non-consolidaed daa, i.e. including inra-secor liabiliies such as inra-enerprise loans. Once availabiliy of consolidaed daa is improved in he fuure, he relaive meris of consolidaed versus non-consolidaed daa can be reconsidered. Neverheless, o accoun for economically relevan naional specificiies, consolidaed daa should be considered, where available, as par of he economic reading of he scoreboard. In paricular, he reasons for large differences beween consolidaed and nonconsolidaed daa should be examined. General governmen (public) secor deb The scoreboard should include an indicaor in he scoreboard on general governmen secor deb as a percenage of o consider he poenial conribuion of public deb o macroeconomic imbalances. I should be emphasised ha he indicaor for public secor deb is no included in he scoreboard wih a view o monior risks of unsusainable public finances. This aspec is clearly covered by he SGP. Insead i should be considered as a complemen o he indicaor on privae deb offering a broader picure for Member Saes' indebedness and herefore heir overall vulnerabiliy. The definiion of general governmen consolidaed gross deb is he one used in he EDP. Daa source is Eurosa. The Treay reference value of 60% will be considered as he indicaive hreshold for he public secor deb. Unemploymen rae An indicaor measuring he evoluion of unemploymen should be included. This indicaor is inended o monior high and persisen raes of unemploymen. This indicaor poins owards a poenial misallocaion of resources (mismach) and general lack of furher adjusmen capaciy in he economy. I should hus be read in conjuncion wih oher more forward-looking scoreboard 6 Privae secor is defined as non-financial corporaions and households and non-profi insiuions serving households. The secor non-financial corporaions include boh privae and public corporaions classified ouside he general governmen deb secor. EN 10 EN

indicaors and be used o beer undersand he poenial severiy of macroeconomic imbalances in erms of heir likely persisence and he capaciy of he economy o adjus. The (backward) average over 3 years of he unemploymen rae based on he Labour Force Surveys annual daa from he EUROSTAT should be used. The saisical approach delivers an indicaive upper hreshold of 10%. EN 11 EN