Chapter 1 – summary
As technology advances and economies develop, the structure of countries’ production changes. Over the past three decades, many economies in the EBRD regions have managed to reinvent themselves to a significant extent, substantially changing their export structures and moving up the value chain as they have benefited from trade liberalisation and closer integration with the European Union. Exporting a diverse mix of relatively complex products to a larger number of destinations has supported economic growth.
The rapid increase in global demand for artificial intelligence, higher spending on defence, and technological advances in new areas, such as the space economy, could further catalyse the development of new export capabilities. In recent years, demand for AI-related products has boosted global growth, while spending on defence has been on the rise. The markets for AI and defence goods are large, and the goods in question tend to be complex and highly connected to other goods in production networks. This makes these markets attractive in principle. Yet, as globalisation stalls and export markets become more concentrated, capitalising on the AI- and defence-related demand shocks will be challenging for economies in the EBRD regions. Economies’ export structures tend to be persistent, and developing new export specialisations is often difficult for complex products.
Building new capabilities involves trial and error and tends to be more successful where economies already have established specialisations in related upstream or downstream products. It also requires active policy support, including the promotion of investment, the negotiation of free-trade agreements granting preferential access to large markets, worker training, tax incentives, the selective use of special economic zones and other well-targeted industrial policies, and measures promoting the rule of law and ensuring macroeconomic stability.
Large exporters can play a key role in developing and marketing new complex products, as in the case of AI and military goods, generating spillovers to the broader ecosystem of firms. However, such capabilities remain concentrated in a limited number of countries: while some economies in the EBRD regions specialise in specific AI-related and military goods, average specialisation in such goods remains relatively low.
In 2024, medicaments accounted for just 1.7% of world trade, but 34% of Slovenia’s exports
In the median EBRD economy, products that have an RCA today but did not 25 years ago currently account for about 35% of total RCA exports
Products with an RCA today that were not exported at all 25 years ago currently account for about 6% of total RCA exports in the median EBRD economy
The median EBRD economy now exports around 28% of all product categories, up from around 13% in 1995
Chapter 2 – summary
After the end of the Cold War, governments worldwide reduced the share of public resources allocated to defence. In Europe, the resulting “peace dividend” averaged about 0.9 per cent of gross domestic product (GDP) a year. Rising geopolitical tensions have brought this peace dividend to an end. Following Russia’s full-scale military invasion of Ukraine in 2022, governments across Europe and in the United States of America have committed to substantial and sustained increases in defence spending. Several economies in the EBRD regions are at the forefront of this shift.
Defence budgets comprise personnel and operating expenses, military infrastructure, procurement of equipment, and military research and development (R&D). While higher defence expenditure can raise debt levels and draw resources away from welfare spending, it can also increase demand for metals, electronics and the output of other civilian industries, and boost R&D. On average, it is estimated that a €1 increase in defence spending in the EBRD regions is associated with a subsequent €1.47 increase in GDP.
While increases in defence spending are determined primarily by security considerations, comparable security objectives can be met and financed in different ways, with varying impacts on innovation, the output of domestic civilian industries and citizens’ welfare. Most economies in the EBRD regions remain highly dependent on imports of defence-related goods, with the notable exceptions of Poland and Türkiye, as defence production and R&D remain highly concentrated globally. Eight countries are estimated to supply more than 80 per cent of the world’s arms exports. The aerospace and defence sector in the EBRD regions is disproportionately state owned, and these state-owned firms tend to be less productive than private counterparts. Defence technology startups in EBRD economies rarely raise capital beyond the earliest funding round.
Inefficiencies associated with the state ownership of military equipment manufacturers, the limited scale of domestic markets, a deficit of skills and weak access to capital all constrain innovation and the domestic production of military hardware and software in the EBRD regions. For economies with a strong industrial base, retooling, joint procurement and the licensing of technologies developed elsewhere can help to expand domestic production.
In EBRD economies in the EU, defence spending has risen from 1.3% of GDP in 2013 to 2.7% in 2025
Türkiye produces about 90% of its arms imports domestically under licence
Only 18% of defence firms in emerging Europe that raise funding at pre-seed or seed stage go on to a Series A funding round or beyond
EBRD economies in the EU import almost 60% of the inputs needed for electronics production, compared with about 30% in advanced Europe
Chapter 3 – summary
Unlike previous waves of automation, generative AI primarily affects knowledge-intensive work, such as writing, software development (coding) and data analysis. AI capabilities continue to improve rapidly, giving firms and workers less time to adjust to innovation than in the past. In the United States, workplace adoption of generative AI has been faster than that of the personal computer (relative to their respective mass-market launches), while its overall adoption has outpaced the rollout of the internet.
The capacity to build AI systems remains heavily concentrated: a handful of economies and firms dominate the markets for advanced semiconductor chips, data infrastructure and the development of frontier AI models. Economies in the EBRD regions play a limited role in these upstream segments of the AI supply chain, accounting for just 0.8 per cent of global exports of related products (but around 8 per cent of total global exports across all goods). The EBRD regions’ potential is stronger in the application layer, where AI models are turned into products and services for end users. The barriers to entry are lower in this market segment, and firms in the EBRD regions are already active in many areas relating to AI applications.
The biggest opportunity lies in AI adoption. Under a medium adoption scenario (modelled on the diffusion of the internet), AI could add around 0.03-0.7 percentage points to annual labour productivity growth across the EBRD regions over the coming decade. Economies with strong digital infrastructure and the skills needed to adopt and use new technologies effectively are best positioned to benefit from AI. Those starting from a lower level of digital economy development, including many economies in the southern and eastern Mediterranean region and sub-Saharan Africa, risk falling further behind, with AI contributing to income divergence.
Realising the potential gains from AI will depend first and foremost on the ability of firms, workers and governments to use AI effectively, and on policy choices that support digital infrastructure, govern data use, facilitate the upgrading of skills and foster international cooperation.
TSMC in Taipei China accounts for around 70% of the global foundry market
AI-related goods accounted for an estimated 6.4% of world merchandise trade in 2024
Data centres consumed an estimated 485 TWh of electricity in 2025, about 1.5% of global electricity demand
The United States and China are expected to account for 78% of the global increase in data-centre capacity
Chapter 4 – summary
Entrepreneurship is the engine of business dynamism and economic growth. While firms in the EBRD regions have continued to catch up with their advanced-economy peers in terms of average productivity levels, young, high-growth firms’ share of total firms has fallen steadily over the past two decades in EBRD economies in the EU. Nonetheless, the EBRD regions remain home to many growth-oriented startups.
As in advanced European economies, the majority of startup founders in the EBRD regions have previously studied or worked abroad. They tend to focus on products that are outside their economies’ established export baskets, often in information technology, professional or financial services. Serial founders – entrepreneurs who have established more than one firm – typically account for 12-19 per cent of total founders in EBRD economies, with most serial founders tending to target a different sector when they launch a new venture. In many cases, the new venture is also in a different economy to the founder’s previous one.
However, relatively few growth-oriented startups in the EBRD regions manage to scale up their operations successfully. Founders whose firms succeed in reaching scale are more likely to have (i) studied or worked abroad before founding the firm, (ii) had a senior role at another firm before launching the startup, (iii) started their firm with at least one co-founder and (iv) obtained funding from at least one foreign investor. Meanwhile, venture-capital funding remains thin and tends to be concentrated in a single metropolitan area in each economy, with a focus on finance and other services. Startups in the EBRD regions tend to raise less venture capital per firm than their counterparts in advanced Europe, and most of what they do raise comes from abroad, predominantly the United States, the United Kingdom and advanced Europe.
Founders in the EBRD regions are more likely to raise funds from investors headquartered in economies where they have previously studied or worked. This reliance on foreign money intensifies as startups grow and mature: domestic and mixed syndicates dominate early funding rounds, but founders in the EBRD regions are far more dependent on foreign funders than their advanced European peers at later stages. Consequently, deepening venture-capital markets, connecting founders to investors and widening the geographical spread of finance will all be central to turning more startups into large productive firms.
The least productive firms in EBRD economies in the EU have raised their productivity by 75%, compared with 45% for the most productive firms
Young, fast-growing firms’ share of total firms in EBRD economies in the EU has more than halved in the past two decades
Around 70% of serial founders’ ventures are in a different sector to their previous venture
In the Western Balkans, EEC, and SEMED and Türkiye, between a third and two-fifths of serial founders’ ventures are in a different economy to the previous one
Chapter 5 – summary
This chapter presents the EBRD’s latest assessment of transition challenges in its regions and selected comparator economies, tracking progress on structural reforms. The assessment focuses on six key qualities of a sustainable market economy, looking at whether economies are competitive, well governed, green, inclusive, resilient and integrated. For each quality, progress is assessed on a scale of 1 to 10, where 10 corresponds to the standards of a sustainable market economy. Those scores are based on a wide range of external and internal data sources.
This year’s assessment also introduces the EBRD’s SME Competitiveness Index. Small and medium-sized enterprises (SMEs) are major sources of employment and make an important contribution to innovation and growth. For each economy, the SME Competitiveness Index aggregates numerous performance indicators in five specific areas: bank financing; non-bank financing; legal frameworks; business skills and standards; and the business environment.
Compared with the Bank’s broader measure of economic competitiveness, the SME Competitiveness Index puts greater emphasis on sources of bank and non-bank financing and the business environment. This is because SME competitiveness revolves around those firms’ ability to grow in the early stages of their development, access finance, innovate, expand into new markets and compete successfully with domestic and international firms.
Assessments of SME competitiveness and the competitiveness of the broader economy are closely, yet imperfectly, aligned. Between 2016 and 2026, increases in overall competitiveness have tended to outpace improvements in SME competitiveness, particularly in central Europe and the Baltic states. This underlines the importance of fine-tuning dedicated SME policies across the EBRD regions, with a focus on access to finance, skills and legal frameworks for small firms.
Some economies have seen declines in their relative standings for integration as a result of restrictions on trade and FDI, higher mobile broadband costs and weaker trade
Well governed scores have declined, on average, in the EBRD regions, widening the existing gap to advanced economies
Improvements in financial resilience have been driven by declines in NPLs, stronger capital adequacy ratios and higher liquidity
Between 2016 and 2026, increases in overall competitiveness have tended to outpace improvements in SME competitiveness
Chapter 5 methodological notes – summary
The transition indicators reflect the judgement of the EBRD’s Office of the Chief Economist, the Impact and Partnerships department, and the Policy Strategy and Delivery department on the transition progress in the economies where the EBRD invests. According to this approach, a sustainable market economy is characterised by six qualities: competitive, well governed, green, inclusive, resilient and integrated.
This approach measures the state of each quality and its components in a given economy, as compared with the other economies in the EBRD regions and a few select developed economies,1 against a frontier. The frontier is set either by the best performance in this group of economies or by an unobserved theoretical value, and provides a common benchmark against which all economies are assessed consistently and comparably. The same frontier values are also applied across the years to ensure that computed scores are comparable and capture changes in underlying indicators through time.
1 The group of comparator developed economies currently includes Canada, Cyprus, France, Germany, Japan, Sweden, the United Kingdom and the United States of America.
Read methodological notes in full
PDF format / 1.35 MBTransition Report 2026-27
PDF format / 12.74 MBForeword
PDF format / 0.66 MBExecutive summary
PDF format / 1.05 MBChapter 1: Creating comparative advantage
PDF format / 2.16 MBChapter 2: The economics of defence spending
PDF format / 1.96 MBChapter 3: Building the AI economy
PDF format / 2.10 MBChapter 4: Entrepreneurship at scale
PDF format / 2.03 MBChapter 5: Structural reform
PDF format / 1.11 MBAcknowledgements
PDF format / 0.13 MBCentral Europe and the Baltic states
PDF format / 1.70 MBCroatia
PDF format / 0.23 MBEstonia
PDF format / 0.25 MBHungary
PDF format / 0.22 MBLatvia
PDF format / 0.24 MBLithuania
PDF format / 0.27 MBPoland
PDF format / 0.22 MBSlovak Republic
PDF format / 0.26 MBSlovenia
PDF format / 0.26 MBSouth-eastern Europe
PDF format / 1.88 MBAlbania
PDF format / 0.25 MBBosnia and Herzegovina
PDF format / 0.28 MBBulgaria
PDF format / 0.24 MBKosovo
PDF format / 0.26 MBMontenegro
PDF format / 0.28 MBNorth Macedonia
PDF format / 0.30 MBRomania
PDF format / 0.25 MBSerbia
PDF format / 0.26 MBTürkiye
PDF format / 0.23 MB