The 2025 EU-wide stress test introduced a granular sectoral dimension that is often overlooked in headline coverage. Alongside the aggregate GDP and capital figures, the ESRB and ECB published a detailed breakdown of real gross value added (GVA) impacts across 16 NACE economic sectors for each of the 27 EU member states. This data reveals which parts of the economy would be hit hardest — and where the greatest cross-country variation lies.
For risk officers and credit teams in financial institutions, this sectoral data is directly relevant: it underpins the credit risk modelling that banks are required to perform in the stress test, and it signals where portfolio concentrations could generate outsized losses under stress.
What Is the Sectoral GVA Breakdown?
Gross value added measures the economic output of an industry after subtracting intermediate inputs. In the stress test context, the ESRB/ECB scenario provides projected GVA growth paths — both baseline and adverse — for 16 NACE sectors in every EU country. Banks use this data to calibrate sector-specific probability of default (PD) and loss given default (LGD) parameters when modelling credit risk across their loan portfolios.
The 16 sectors span the entire economy: agriculture (A), mining (B), manufacturing split into low/medium-tech and high-tech (C), energy and utilities (D), water and waste (E), construction (F), wholesale and retail trade (G), transportation (H), hospitality (I), ICT (J), financial and insurance services (K), real estate (L), professional services (MN), public administration (OPQ), and arts and entertainment (RSTU).
The Most and Least Affected Sectors at EU Level
Under the adverse scenario, the cumulative GVA impact over 2025–2027 varies dramatically across sectors. The table below ranks all 16 sectors from most to least affected at the EU aggregate level.
Why Manufacturing and Transport Are Hit Hardest
The adverse scenario’s narrative — centred on geopolitical escalation, trade fragmentation, and commodity price surges — explains the sectoral distribution of losses. High-tech manufacturing (NACE C-high) faces the steepest cumulative decline at −14.5% because it is highly exposed to global supply chains, trade tariffs, and export demand. The scenario assumes severe disruptions to these channels, compounded by higher energy and raw material costs.
Transportation and storage (NACE H) ranks second at −11.1% for similar reasons: the assumed collapse in global trade volumes directly reduces freight activity, while energy price surges increase operating costs across the logistics chain.
Key insight: The gap between the most and least affected sectors is substantial — a 14.4 percentage-point spread between high-tech manufacturing (−14.5%) and public administration (−0.1%). For banks with concentrated sectoral exposures, this range translates into very different credit loss trajectories under the same macroeconomic scenario.
Country-Level Variation: Where Sectoral Risk Concentrates
The EU aggregate figures mask significant cross-country differences. For the worst-hit sector — high-tech manufacturing — the cumulative adverse GVA impact ranges from −6.0% in Slovenia to −20.1% in Sweden. Countries with large export-oriented manufacturing bases, greater exposure to global value chains, or higher dependence on energy imports tend to face steeper declines.
Sweden: −20.1% cumulative in high-tech manufacturing. Also faces −19.5% in mining and −17.5% in low-tech manufacturing.
Finland: −19.8% in high-tech manufacturing. The mining sector (−13.8%) and transportation (−14.8%) are also severely stressed.
Greece: −19.6% in high-tech manufacturing, with transportation at −15.1% and mining at −16.2%.
Ireland: Construction actually grows (+3.2% cumulative) under the adverse scenario. High-tech manufacturing impact (−9.7%) is below the EU average.
Lithuania: Construction at +2.4% and public administration at +3.4%. High-tech manufacturing at −9.4% — well below the EU average.
Bulgaria: Construction at +0.3%. Several service sectors show relatively limited damage.
This variation matters because EU banking supervisors use the country-specific sectoral data to assess each bank’s credit risk exposure. A bank with a large loan book concentrated in Swedish manufacturing faces a fundamentally different stress test outcome than one focused on Irish services.
The Year-by-Year Pattern: Deepest Impact in 2026
Across most sectors and countries, the adverse scenario follows a consistent temporal pattern. The initial shock hits in 2025, with negative GVA growth across virtually all sectors. However, the deepest contraction typically comes in 2026 — reflecting the lagged effects of the trade and confidence shocks that drive the scenario narrative. By 2027, most sectors show a partial recovery, with some returning to weak positive growth, though GVA levels remain far below their baseline trajectories.
At the EU level, high-tech manufacturing illustrates this pattern clearly: adverse GVA growth of −4.7% in 2025 deepens to −8.3% in 2026 before recovering to −2.1% in 2027. This 2026 trough reflects the full propagation of supply chain disruptions, export demand collapse, and tightened financial conditions through the industrial economy. The same pattern holds for transportation (−3.7% in 2025, −6.5% in 2026, −1.3% in 2027), where the assumed reduction in global trade volumes takes time to feed through to freight and logistics activity.
For construction and public services, by contrast, the pattern is milder and flatter. Construction at the EU level shows −1.0% in 2025, −2.3% in 2026, and +0.4% in 2027 — a relatively modest swing compared to the manufacturing sectors. Public administration barely registers the shock at all, declining just −0.5% in 2025 before recovering to +1.6% in 2027, reflecting the “no policy change” convention built into the scenario design.
Sector–Country Combinations to Watch
The most informative view of the sectoral data is not at the aggregate level but at the intersection of specific sectors and specific countries. Several sector–country combinations stand out as particularly severe under the adverse scenario.
Sweden’s high-tech manufacturing decline of −20.1% cumulative is the single worst sector–country combination outside very small economies. This reflects Sweden’s heavy dependence on export-oriented advanced manufacturing — including automotive, electronics, and precision engineering — all of which are acutely sensitive to the scenario’s trade fragmentation assumptions. Combined with Sweden’s mining sector at −19.5% and low-tech manufacturing at −17.5%, the data implies a broad-based industrial contraction that would pose significant challenges for Swedish banks with concentrated corporate lending books.
Finland presents a similar picture, with high-tech manufacturing at −19.8%, transportation at −14.8%, and mining at −13.8%. Finland’s relatively small, open economy and its exposure to northern European and Russian trade routes make it particularly vulnerable to the scenario’s geopolitical assumptions.
At the other end of the spectrum, several smaller economies show surprising resilience in specific sectors. Ireland’s construction sector actually records +3.2% cumulative growth even under the adverse scenario — a reflection of Ireland’s ongoing housing supply deficit and structural demand. Greece’s construction sector similarly shows +2.0%, supported by EU-funded infrastructure investment programmes that are assumed to continue under the “no policy change” convention.
For banks and risk teams, these sector–country intersections are where the stress test’s sectoral dimension adds the most analytical value. Aggregate EU-level figures mask the wide dispersion in outcomes that determine individual bank results.
Implications for Banks and Risk Teams
The sectoral GVA data has several practical implications for financial institutions and their risk management frameworks. First, it provides a transparent basis for benchmarking internal credit risk models. Banks should be able to explain how their projected sector-specific losses relate to the GVA shocks prescribed by the scenario — and supervisors will scrutinise deviations closely during the quality assurance process.
Second, the data highlights concentration risk in bank lending portfolios. Banks with disproportionate exposure to manufacturing, mining, or transportation should expect more challenging stress test outcomes and potentially higher Pillar 2 Guidance (P2G) requirements as a result. The wide dispersion in sectoral outcomes — ranging from −14.5% for high-tech manufacturing to −0.1% for public administration — means that portfolio composition has a material impact on projected capital depletion.
Third, the EBA noted in its 2025 results report that banks have made progress in differentiating sector-level impacts compared with earlier exercises, but that modelling capabilities still require further refinement. In particular, only around 15 banks per sector (out of 64 in the sample) used statistical models for probability of default projections across more than half of their exposures. This suggests that sectoral credit risk modelling will be an area of increasing supervisory focus in future exercises — and that banks investing in more granular, sector-aware models will be better positioned to demonstrate the resilience of their portfolios.
Finally, for non-bank financial institutions, asset managers, and corporate treasurers, the sectoral GVA data offers a valuable lens for assessing counterparty risk, evaluating industry-specific credit exposures, and understanding how different parts of the economy respond to the type of severe macroeconomic shock that regulators consider plausible.
To explore the full sectoral GVA data for any country and sector combination, use our free Sectoral GVA Impact Analyser, which visualises baseline and adverse paths across all 16 NACE sectors and 27 EU member states. For the broader macroeconomic scenario data — GDP, unemployment, inflation, and real estate prices — see our EBA 2025 Stress Test Scenario Explorer.
Explore Sectoral GVA DataFrequently Asked Questions
What is GVA in the context of the EBA stress test?
Gross value added (GVA) measures the economic output of an industry minus intermediate consumption. In the stress test, the ESRB and ECB provide projected GVA growth paths for 16 NACE sectors under both baseline and adverse scenarios. Banks use this data to model sector-specific credit risk — calibrating how default probabilities and loss severities change across different parts of the economy under stress.
Which sector is most affected under the 2025 adverse scenario?
High-tech manufacturing (NACE C-high) is the most severely affected sector at the EU aggregate level, with a cumulative adverse GVA decline of −14.5% over 2025–2027. This reflects its high exposure to global trade disruptions, supply chain fragmentation, and rising input costs — all core elements of the adverse scenario narrative.
How many sectors are covered in the stress test?
The stress test covers 16 economic sectors based on the NACE classification system. Manufacturing is split into two sub-categories: low and medium-tech (C-low) and high-tech (C-high). The sectors span the entire economy from agriculture to public administration, enabling banks to model credit risk granularly across their loan portfolios.
Why does the sectoral GVA data matter for banks?
Banks use the sector-specific GVA projections to calibrate probability of default and loss given default parameters when modelling credit risk. The data determines how much capital banks need to hold against losses in different parts of their loan portfolios. Banks with concentrated exposures to the most affected sectors — such as manufacturing or transportation — will face higher projected losses and potentially greater capital depletion under the adverse scenario.
Which countries face the worst sectoral impact?
Countries with large export-oriented manufacturing bases tend to face the steepest sectoral declines. Sweden (−20.1% in high-tech manufacturing), Finland (−19.8%), and Greece (−19.6%) are among the most affected. Countries with more diversified, service-oriented economies, such as Ireland and Lithuania, generally show milder sectoral impacts.
