Development of GDP

Situation in the euro area and the EU

In the fourth quarter of 2025, seasonally adjusted GDP increased, up 0.2% quarter on quarter in both the euro area and the EU. For comparison, an increase of 0.3% had also been observed in Q3 2025 in the euro area and 0.4% in the EU.

Situation in the EU countries

Based on the latest quarter-on-quarter rates of change, GDP increased in Q4 2025 in 22 countries, was unchanged in 1 and decreased in 4. Malta (up 2.1%) and Lithuania (up 1.9%) recorded the largest increases of GDP, while the decreases were observed in Ireland (down 3.8%), Romania (down 1.8%), Estonia and Luxembourg (both down 0.1%).

Situation in non-EU countries

Compared with the previous quarter, GDP expanded in the fourth quarter of 2025 in China (up 1.2%), Japan (up 0.3%), the United States (up 0.1%) and Switzerland (up 0.1%), while it contracted in Norway (down 0.3%) and Iceland (down 0.8%).

(*) See latest estimates for more data.

Industrial production

In February 2026, seasonally adjusted ’‘’industrial production’’’ increased month on month by 0.4% in both the euro area and the EU. In the previous month, there had been a decrease of 0.8% in the euro area and 0.9% in the EU.

Production in construction

In February 2026, seasonally adjusted production in construction decreased 0.2% in the euro area and increased 0.1% in the EU compared with the previous month. In the previous month, it had decreased 1.3% in the euro area and 2.1% in the EU.

Retail trade

In February 2026, the seasonally adjusted volume of sales in retail trade decreased compared with the previous month in the euro area (down 0.2%) and the EU (down 0.3%), following on from no change (0.0%) in the previous month in both the euro area and the EU.

Inflation

Situation in the euro area and the EU

According to the latest data from Eurostat, the annual inflation rate (based on the HICP) in the euro area was 2.6% in March 2026, up 0.7 percentage points from the previous month. In the EU, the annual inflation rate was 2.8% in March 2026, also up 0.7 percentage points from the previous month.

Situation in the EU countries

The highest annual inflation rates (based on the HICP) in March 2026 were recorded in Romania (9.0%), Croatia (4.6%) and Lithuania (4.4%), while the lowest rates were recorded in Denmark (1.0%), Czechia, Cyprus, Sweden (all 1.5%) and Italy (1.6%).

Situation in non-EU countries

In March 2026, the annual inflation rate was 5.0% in Iceland, 3.6% in Norway, 3.3% in the United States, 1.0% in China and 0.6% in Switzerland. In February 2026, the annual inflation rate was 1.3% in Japan.

Total unemployment

Situation in the euro area and the EU

In February 2026, the seasonally adjusted unemployment rate was 6.2% in the euro area, up from 6.1% in the previous month. In the EU, the rate was 5.9%, unchanged from the previous month.

Situation in the EU countries

In February 2026, the lowest unemployment rates were recorded in Bulgaria, Czechia, Poland (all 3.2%) and Malta (3.5%). The highest rates were in Finland (10.6%) and Spain (9.8%).

Situation in non-EU countries

The unemployment rate in February 2026 was 6.0% in Iceland, 4.9% in Norway, 4.4% in the United States and 2.6% in Japan. The unemployment rate was 5.1% in Switzerland in December 2025.

Youth unemployment

In February 2026, the youth unemployment rate (for people aged 15 to 24 years) was 14.9% in the euro area, unchanged from the previous month. In the EU, the rate was 15.3%, up from 15.2% in the previous month.

Economic sentiment indicator

Situation in the euro area and the EU

In March 2026, the economic sentiment indicator decreased, down 1.6 [[Glossary:Percentage_point|percentage points]] in the euro area and 1.5 percentage points in the EU: the index level was 96.6 percentage points in the euro area and 96.7 percentage points in the EU. The decrease in the euro area resulted from decreases in confidence among retail trade and services managers and consumers outweighing slight increases among industrial and construction managers.

Situation in the EU countries

In March 2026, the economic sentiment indicator showed more negative than positive developments among the EU countries, increasing in 6, unchanged in 1 and decreasing in 20. The strongest increase was in Denmark (up 2.8 percentage points), followed by Malta (up 2.6 percentage points). The strongest decreases were in Cyprus (down 6.3 percentage points) and France (down 3.7 percentage points). Apart from these 4 countries, the latest month-on-month changes ranged between 1.1 and -2.5 percentage points. The index level of the economic sentiment indicator (with a long-term average = 100) generally varied from 89.5 percentage points in Belgium to 106.8 percentage points in both Greece and Croatia; the index in Malta (114.8 percentage points) was above this range.

Employment expectations indicator

In March 2026, the employment expectations indicator decreased (down 1.4 percentage points compared with the previous month) to 96.4 in the euro area; expectations were down 1.3 percentage points in the EU to 97.3. The decrease in the euro area in March 2026 was due to marginally less pessimistic plans among industrial managers being outweighed by marginally less optimistic employment plans among construction managers as well as more pessimistic plans among retail trade and services managers.

€-coin

The €-coin decreased to 0.48 in March 2026, down from 0.54 in the previous month. The performance of the indicator was driven by worsening industrial production data and falling stock prices (according to the latest release of the National Bank of Italy on 2 April 2026).

Business climate indicator

The business climate indicator for the euro area increased to -0.27 points in March 2026 (up from -0.36 points in the previous month).

Coincident indicators

The growth cycle coincident indicator (GCCI) was 0.06 in January 2026, the business cycle coincident indicator (BCCI) was 0.00 and the acceleration cycle coincident indicator (ACCI) was 0.24, indicating a phase of expansion with accelerating growth for the euro area. Note that no new observations are available to Eurostat for this series since the December 2025 data were published and the extension of the time series has been paused.

The GCCI, BCCI and ACCI are synthetic indicators which are experimental in nature and calculated for Eurostat’s business cycle clock. The GCCI shows the probability of a slowdown in the economy and signals the peaks and troughs of the growth cycle, the BCCI shows the probability of a recession and signals the peaks and troughs of the business cycle and the ACCI shows the probability of a deceleration in the growth rate and signals the peaks and troughs of the growth rate cycle. The probability is given on a scale from 0 to 1.

Read more about the methodology in the Statistics Explained article about the business cycle clock.

DZ BANK’s Euro-Indikator

The DZ BANK’s Euro-Indikator decreased 0.46% month on month to 98.13 points in March 2026, following on from a smaller increase (up 0.19%) in the previous month. When compared with the same month in the previous year, the Euro-Indikator was 0.66% higher in March 2026.

Trend-cycle estimates

This graph shows the quarterly GDP in current prices for the euro area, together with trend-cycle decompositions.

A trend-cycle decomposition method aims at further decomposing seasonally adjusted data into a trend component and a cyclical component (expressed as the deviation from the trend). Note that no new observations are available to Eurostat for the trend-cycle decomposition and the extension of the time series has been paused.

The trend provides information on longer-term movements in the seasonally adjusted series over several years; the cycle is a sequence of smoother fluctuations around the longer-term trend, in part characterised by alternating periods of expansion and contraction.

You can choose to show the estimates 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 non-parametrically.
• The Harvey and Trimbur decomposition, which is probably the most well-known model within the Unobserved Components (UC) framework.

Read more in the methodological note.

Trend-cycle estimates

This graph shows the monthly industrial production index for the euro area, together with trend-cycle decompositions.

A trend-cycle decomposition method aims at further decomposing seasonally adjusted data into a trend component and a cyclical component (expressed as the deviation from the trend). Note that no new observations are available to Eurostat for the trend-cycle decomposition and the extension of the time series has been paused.

The trend provides information on longer-term movements in the seasonally adjusted series over several years; the cycle is a sequence of smoother fluctuations around the longer-term trend, in part characterised by alternating periods of expansion and contraction.

You can choose to show the estimates 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 non-parametrically.
• The Harvey and Trimbur decomposition, which is probably the most well-known model within the Unobserved Components (UC) framework.

Read more in the methodological note.

Trend-cycle estimates

This graph shows the quarterly employment in thousand people for the euro area, together with trend-cycle decompositions.

A trend-cycle decomposition method aims at further decomposing seasonally adjusted data into a trend component and a cyclical component (expressed as the deviation from the trend). Note that no new observations are available to Eurostat for the trend-cycle decomposition and the extension of the time series has been paused.

The trend provides information on longer-term movements in the seasonally adjusted series over several years; the cycle is a sequence of smoother fluctuations around the longer-term trend, in part characterised by alternating periods of expansion and contraction.

You can choose to show the estimates 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 non-parametrically.
• The Harvey and Trimbur decomposition, which is probably the most well-known model within the Unobserved Components (UC) framework.

Read more in the methodological note.

Trend-cycle estimates

This graph shows the quarterly GDP in current prices for the EU, together with trend-cycle decompositions.

A trend-cycle decomposition method aims at further decomposing seasonally adjusted data into a trend component and a cyclical component (expressed as the deviation from the trend). Note that no new observations are available to Eurostat for the trend-cycle decomposition and the extension of the time series has been paused.

The trend provides information on longer-term movements in the seasonally adjusted series over several years; the cycle is a sequence of smoother fluctuations around the longer-term trend, in part characterised by alternating periods of expansion and contraction.

You can choose to show the estimates 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 non-parametrically.
• The Harvey and Trimbur decomposition, which is probably the most well-known model within the Unobserved Components (UC) framework.

Read more in the methodological note.

Trend-cycle estimates

This graph shows the monthly industrial production index for the EU, together with trend-cycle decompositions.

A trend-cycle decomposition method aims at further decomposing seasonally adjusted data into a trend component and a cyclical component (expressed as the deviation from the trend). Note that no new observations are available to Eurostat for the trend-cycle decomposition and the extension of the time series has been paused.

The trend provides information on longer-term movements in the seasonally adjusted series over several years; the cycle is a sequence of smoother fluctuations around the longer-term trend, in part characterised by alternating periods of expansion and contraction.

You can choose to show the estimates 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 non-parametrically.
• The Harvey and Trimbur decomposition, which is probably the most well-known model within the Unobserved Components (UC) framework.

Read more in the methodological note.

Trend-cycle estimates

This graph shows the quarterly employment in thousand people for the EU, together with trend-cycle decompositions.

A trend-cycle decomposition method aims at further decomposing seasonally adjusted data into a trend component and a cyclical component (expressed as the deviation from the trend). Note that no new observations are available to Eurostat for the trend-cycle decomposition and the extension of the time series has been paused.

The trend provides information on longer-term movements in the seasonally adjusted series over several years; the cycle is a sequence of smoother fluctuations around the longer-term trend, in part characterised by alternating periods of expansion and contraction.

You can choose to show the estimates 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 non-parametrically.
• The Harvey and Trimbur decomposition, which is probably the most well-known model within the Unobserved Components (UC) framework.

Read more in the methodological note.