Frequently asked questions

 

Work in progress - this page is still under construction, it will soon be completed

What is Eurocoin?

 

The Banca d'Italia and CEPR start (from September 2007) the publication of a new economic activity indicator for the Euro area €-coin. This new indicator replaces the old Eurocoin index, which as been published on a monthly basis by CEPR since 2002.

€-coin is a monthly estimate of the underlying growth rate of the euro area GDP and provides a comprehensive and reliable assessment of the euro area economic outlook,

€-coin is published at the end of each month and refers to the same month. So it is a very timely indicator of growth.

€-coin  is constructed weighing the flow of new data that becomes available as time goes by and translating it into an updated estimate of GDP growth free from measurement errors, seasonal and other short-run fluctuations.

€-coin is a unique indicator since, unlike other widely and commonly used indexes (IFO, PMI, ESI), it neither refers to a single country or sector, nor does it reflect the opinions of a specific group of economic agents. Instead it tracks GDP growth, anticipating the release official GDP releases by several months.

€-coin  is obtained by extracting, in a statistically optimal fashion, from a large set of data on the euro area (that includes the indexes mentioned above), the information that is most relevant to forecast future GDP. The resulting index is a “smooth version” of GDP growth, that:

  • gives an early estimate of Euro area growth performance in terms of quarter-on-quarter changes in GDP;
  • sheds light on the underlying trend in GDP, because it removes short-run fluctuations and measurement errors (in this respect is not only a forecast, but also a clearer indicator of the true growth momentum in the Euro area).

 

Why do you need an indicator? Doesn't looking at GDP tell you everything you need to know about the economic outlook?

 

The short answer is “no”. In fact, while GDP is a comprehensive measure of economic activity:

1. GDP figures are only available at a quarterly frequency and they are published with a rather long delay;
2. GDP figures can be misleading since in any given month GDP growth may be high or low depending on seasonal effects and measurement errors;
3. GDP is subject to large revisions;
4. Euro area GDP can be influenced by factors affecting only a particular country. These factors are of no importance for assessing the health of the euro area economy as a whole.

Ideally one would like to have a comprehensive measure of economic activity, like GDP, but free from the problems and shortcomings listed above.

Removing seasonal effect and short run noise can be easily done, for example by applying a band pass filter to the GDP growth series. This statistical technique, however, still presents the same problems in terms of frequency and timeliness and, which is worse, produces estimates that are highly unreliable at the end of the sample.

We propose an indicator that produces a monthly estimate of GDP growth, free from seasonal and other short run fluctuations, from local and sector-specific shocks and from errors in the measurement of GDP and highly reliable at the end of the sample (i.e. the last figure is not subject to major revisions as new data become available).

€-coin clearly shows the underlying direction in which economy is moving.


How do I read €-coin?

 

If the graph of the indicator has a positive slope, the underlying growth rate of GDP is increasing, a negative slope indicates that the rate of growth is decreasing. The actual numbers can be used as a reference for future releases of GDP. As shown in the table below using Eurocoin as a forecast of future GDP produces rather reliable results

 

RMSFE  

   GDP q-o-q

GDP y-o-y   

GDP BP

0.20  

    0.15  

   0.13

 

Over the 1999-2006 period the root mean squared forecast error in real time with respect to the quarterly growth rate of GDP is 0.20, which is in line with best practice bridge models. As stated above, though, €-coin is not constructed to track GDP growth as such, rather it is an estimate of the underlying growth rate. In technical terms the latter can be defined as the cycle-trend component of GDP growth, which can be approximated with a moving average of quarterly growth rates. A better forecast in terms of GDP growth is obtained, in fact, for the year on year growth rate one quarter ahead, which is a 4 quarters moving average of quarterly rates. The third column in the table shows the real time forecasting performance of €-coin with respect to a measure of the "trend-cycle GDP growth" obtained in the middle of the sample by a band pass bilateral filter on GDP growth. This is as close as we can go to the "target" of €-coin, and, as should be expected, the forecast error with respect to this smooth measure of growth is very small.