EUROINDICATORS WORKING GROUP

Transcription

1 EUROINDICATORS WORKING GROUP 1 TH MEETING 3 & 4 DECEMBER 27 EUROSTAT D1 DOC 26A/7 IMPROVEMENTS OF THE EUROSTATISTICS PUBLICATION ANNEX 1: NEW SECTION 1 ITEM 5.1 ON THE AGENDA OF THE MEETING OF THE WORKING GROUP ON EUROINDICATORS DOC n 26a/7

2 Comparison of GDP growth forecasts for the euro area (14 November 27 update) Percentage change Over previous quarter (q/q-1) Over previous year (q/q-4) 27Q2 27Q3 27Q4 28Q1 27Q2 27Q3 27Q4 28Q1 EUROSTAT release Consensus forecast October EUROFRAME HANDELSBLATT EUROSTAT-OFCE DG-ECFIN EuroCOIN IRC Quantitative indicators (GDP) EUROFRAME: calculated by the OFCE (Paris) for Euroframe consortium which consists of: CPB (Den Haag), DIW (Berlin), ESRI (Dublin), ETLA (Helsinki), IFW (Kiel), NIESR (London), PROMETEIA (Bologna), WIFO (Vienna). The purpose is to forecast euro area GDP growth rates two quarters ahead of official statistics. The indicator considers surveys from industry, retail and construction sectors, ISM index of US industry, real euro/dollar exchange rate, real short-term interest rate and raw material price index (excluding oil and unprocessed food). The indicator is based on ordinary least squares (OLS) estimation of the GDP growth rate, and released on a monthly basis. HANDELSBLAT: calculated quarterly, is conceived to forecast 4-quarter GDP growth two quarters ahead of official publications. This is a composite indicator of the European Commission industrial and consumer confidence indicators, monthly growth of industrial production excluding construction, growth in the monetary aggregate M1 and growth of new orders in industry. The series are weighted and aggregated in the composite indicator with various lags. The resulting series is related to the 4-quarter smoothed growth rate of GDP. EUROSTAT-OFCE: monitored by the OFCE (Paris) for Eurostat. This indicator is a 2-quarter leading indicator of quarterly GDP growth. It comprises an industry survey, household survey, real euro/dollar exchange rate, 1-year/3-month interest rate spread, and real oil price in euro. The indicator is based on OLS estimation, and released on a monthly basis. DG ECFIN: monitored by the European Commission. This indicator is a dynamic factor model designed to summarise information contained in large datasets, and uses these to make forecasts of euro area GDP growth in quarterly terms. The forecasts are presented as ranges with confidence intervals rather than as point estimates. EuroCOIN: real-time monthly estimate of euro area GDP growth, computed each month by the Banca d Italia. It is obtained by collecting a large set of statistics and extracting from it information relevant to forecast future GDP. It tracks GDP growth anticipating official GDP releases by several months, by giving each month an early estimate of euro area growth in terms of quarter-on-quarter changes in GDP. Moreover, it sheds light on the underlying trend by removing short-run fluctuations and measurement errors from the growth rate; in this respect it is not only a forecast, but also an indicator of the true growth momentum in the euro area. IRC: coincident indicator of the business cycle available in real time. Through the use of temporal disaggregation method, using EU business surveys to extract a high frequency series (monthly GDP) from a low frequency series (quarterly GDP), the indicator provides monthly estimates of growth of euro area GDP. The indicator, monitored by the French COE-REXECODE institute, is converted in quarterly terms by averaging monthly values.

3 Quarter-over-quarter GDP growth indicators (q/q-1) 1.2 % change over previous quarter Euroframe 1..8 EuroCoin DG-ECFIN range.6 IRC.4.2 GDP release Eurostat- OFCE. Handelsblatt 6q1 6q2 6q3 6q4 7q1 7q2 7q3 7q4 8q1 GDP release Handelsblatt Euroframe IRC EuroCoin Eurostat-OFCE Year-over-year GDP growth indicators (y/y-1) 3.4 % change over previous EuroCoin 3. IRC Euroframe Handelsblatt GDP release 1.8 EuroStat- OFCE DG-ECFIN range 6q1 6q2 6q3 6q4 7q1 7q2 7q3 7q4 8q1 GDP release EuroCoin Handelsblatt Eurostat-OFCE Euroframe IRC

4 Comparison of indicators of turning points in the business cycle for the euro area (14 November 27 update) Q4 27Q1 27Q2 27Q3 27m5 27m6 27m7 27m8 27m9 OECD-CLI Index IFO Index DG ECFIN Business Climate Indicator Points of stddev SZ-Euroland Index DZ-Euroland indicator Index IARC Probability IESR Probability Qualitative indicators (turning points and business climate) OECD: Composite Leading Indicator (CLI) designed to predict cyclical turning points in industrial production as a proxy of the business growth cycle. It comprises a set of component series from a wide range of economic indicators. Data are presented in trend restored form. IFO: Economic Climate Indicator, designed to give an accurate picture of the economic situation and forecasts for economies on a quarterly basis. It consists of qualitative information: appraisals and expectations of economic experts. For the euro area the trend of the indicator correlates well with the actual business-cycle trend. DG ECFIN: Business Climate Indicator (monthly), designed to deliver a clear and early assessment of the cyclical situation of the euro area. The indicator uses, as input, five balances of opinions from DG ECFIN's Business and Consumer Surveys. The resulting common factor may be read as a survey result: the higher the level, the healthier the cyclical situation, and a rise in the indicator points to an upswing in activity and an improvement in business climate. SZ-Euroland-Indicator: designed by Dekabank as a reference series reflecting movements in growth in monthly industrial production. It is currently based on five series: business survey in industry, Purchasing Manager Index in manufacturing, new orders in manufacturing, an interest rate spread, and the MSCI stock index for the EMU. It is monitored monthly to predict turning points of the economy with 1-month to 5-month leads. DZ-Euroland-Indicator: calculated by DZ Bank (monthly), aims at predicting recessions and has also proved to reflect movements in GDP growth rate. It is built on concepts implemented by the US Conference Board for monitoring cyclical indicators. Nine series for euro area or for the main countries are selected for their leading properties, and aggregated to give a composite index which can be compared with the level of GDP in order to anticipate recessions. IARC: prepared by the Centre d Observation Economique (Paris). It is a monthly indicator aimed at forecasting turning points in the business cycle. Each month, the probability that each selected series has crossed a turning point is computed. Probabilities are aggregated by a weighting method to give an overall signal of a possible downswing in the next nine months (a value of the indicator higher than 6 and lower than 8), or three months (a value higher than 8). IESR: conceived by the Centre d Observation Economique. It is a coincident indicator able to detect in real time peaks and troughs of the business cycle. It is based on the Markov-Switching model applied to various economic time series. The filtered probabilities are combined by taking into account the risks of false signals in order to provide a recession probability (a value of the indicator higher than.5).

5 DZ-Euroland and IFO Business Climate Indicators 6 % change over previous Index 1995= IFO Indicator (right-hand scale) Y-O-Y GDP growth rate (left-hand scale) 6-2 DZ-Euroland-Indicator (left-hand scale) SZ-Euroland and DG-ECFIN Business Climate Indicators 6 In % 4 Y-O-Y growth rate of the industrial production index (excl. construction) SZ-Euroland-Indicator 2-2 DG-ECFIN business climate indicator

6 The OECD Leading Indicator 1 T P T P T Jan Mar Feb Dec Jun T:trough / P:peak Lead in months mont annualized growth rate of the OECD indicator -4 Detrended industrial production index Latest trend cycle estimates for the euro area and EU Methodology The trend cycle is the component in a time series that represents variations of low frequency, the high frequency fluctuations having been filtered out. Identifying and estimating trend and cyclical components of key economic indicators is invaluable when making decisions and informing policy-makers. The distinction between the two components lets the analyst ascertain whether any change to the indicator is temporary or permanent. However, there exist many alternative methods to identify and estimate trends and (business or deviation) cycles. The estimates below for three indicators GDP, industrial production, and employment are produced by the National Institute of Economic and Social Research in London using three different filters: Hodrick-Prescott (HP): a filter widely used in macroeconomics to fit a smooth curve through a set of points. Christiano-Fitzgerald (CF): a well-known approximation to an ideal band pass filter, estimated nonparametrically. unobserved components (UC): a parametric approach which involves estimating a statistical model and decomposing the series into components, including the trend and cycle.

7 GDP trend cycle Based on GDP at market prices, in millions of euro, chain-linked volumes, seasonally adjusted and working day adjusted, reference year 2 (at 2 exchange rates). Source: NIESR. GDP trend cycle EA q1 1996q1 1997q1 1998q1 1999q1 2q1 21q1 22q1 23q1 24q1 25q1 26q1 27q1 GDP trend cycle EU q1 1996q1 1997q1 1998q1 1999q1 2q1 21q1 22q1 23q1 24q1 25q1 26q1 27q1

8 Industrial production trend cycle Based on monthly industrial production index (excluding construction), seasonally-adjusted volumes. Source: NIESR. Industrial production trend cycle EA Industrial production trend cycle EU

9 Employment trend cycle Based on total employment (domestic concept) series, seasonally-adjusted and working day-adjusted. Source: NIESR. Employment trend cycle EA q1 1992q1 1993q1 1994q1 1995q1 1996q1 1997q1 1998q1 1999q1 2q1 21q1 22q1 23q1 24q1 25q1 26q1 27q1 Employment trend cycle EU q1 1996q1 1997q1 1998q1 1999q1 2q1 21q1 22q1 23q1 24q1 25q1 26q1 27q1

10 Turning-point chronology The need for a cycle turning-point chronology is widely recognised by experts and practitioners of economic analysis. The cycles measure fluctuations around the long-term trend in aggregate economic activity. ADD DEFINITIONS OF GROWTH CYCLE, BUSINESS CYCLE AND ACCELERATION CYCLE HERE Euro area growth cycle Q1 197 Q Q Q Q1 198 Q Q Q Q Q4 2 Q Q Q Q Q Q Q Peak Q Q Q Q Q3 198 Q Q Q Q Q Q Q2 23 Trough Q4 21 OECD Anas Ferrara OECD: January 27, based on industrial production index series. NOTE MORE UP-TO-DATE OECD DATA ARE AVAILABLE SHOWING A PROVISIONAL TROUGH IN MARCH 25 (PRESS RELEASE 9 NOV). THE WEB SITE ALSO SHOWS MORE RECENT DATA WITH A TROUGH IN NOV 21. Anas Ferrara: 26, BBK method, based on total GDP series. THERE ARE NO OTHER RECENT STUDIES PROVIDED (OTHERS ARE FROM 21 AND 22) Euro area business cycle Q Q1 198 Q Q Q1 198 Q Q Q Peak Q Q3 198 Q Q Q Q4 198 Q Q Trough Anas and Ferrara Benalal and Ali

11 Anas Ferrara: univariate BB method, based on total GDP series. Date of estimation June 26. Benalal and Ali: BBQ (???) method, May 26. NOTE THERE ARE 2 ESTIMATIONS MADE IN 24 BY BENGOECHEA AND PEREZ-QUIROS WHICH SHOW LATER PEAKS AND TROUGHS. WE COULD ADD ONE OF THESE TO THE GRAPH. Euro area acceleration cycle M M3 198 M M M M M M Peak M M M M M M M M11 21 M M M M M M M Trough Eurocoin Harding Eurocoin: univariate method, factor analysis on 1 series. Date of estimation May 26. Harding: univariate mehod, BBK to growth rate, based on GDP series. Date of estimation November 22.