OeNB Report 2026/19: CESEE-8 government debt and risks in 2025

Zoltan Walko, Oesterreichische Nationalbank, Central, Eastern and Southeastern Europe, zoltan.walko@oenb.at

Publication: 13 July 2026

This report takes stock of 2025 government debt ratios in eight EU member states in Central, Eastern and Southeastern Europe (CESEE-8). It offers comparisons to 2024 and analyses the major drivers behind debt developments, such as: primary balances, interest payments, the impact of inflation, real GDP growth and stock-flow adjustments. In addition, the report gives an overview of the risk profile of government debt at the end of 2025. Finally, it presents the most recent available data on government contingent liabilities and on national fiscal governance frameworks (2024 data).

Highlights

1 Introduction

This report takes stock of government debt ratios in eight Central, Eastern and Southeastern European EU member states (CESEE-8) ¹ at the end of 2025, with comparisons to 2024. It analyses the major drivers of debt dynamics, such as primary balances, interest payments, the impact of inflation, real GDP growth and stock-flow adjustments (i.e. differences between the change in debt and the fiscal balance). In addition, the report gives an overview of the risk profile of government debt at the end of 2025, including major changes over the past few years. Compared to the previous edition of February 2026, which presented data for the end of 2024 ² , the analytical scope of this report has been broadened to include the most recent data (for 2024) on government contingent liabilities and on national fiscal governance frameworks.

The current report is based on the first release of deficit and debt data for 2025 contained in the April 2026 fiscal notifications by EU member states (as published in the AMECO and Eurostat databases).

2 Government debt ratios rose to new highs in several countries in 2025

In 2025, the general government debt ratio (as a percentage of GDP) increased in six of the CESEE-8 countries discussed in this report (chart 1). These six countries were the same which had seen a rise in their debt ratios in 2024 (i.e. Bulgaria, Czechia, Hungary, Poland, Romania and Slovakia). While in Czechia, Poland, Romania and Slovakia the increase in the debt ratio in 2025 was smaller than in 2024, the ratio rose more strongly in Bulgaria and Hungary. The strongest increase was recorded in Bulgaria (+6 percentage points), followed by Poland and Romania (+4.8 and +4.5 percentage points, respectively). In Croatia and Slovenia the debt ratio continued to decline in 2025, but it did so at a slower pace than in 2024 or in 2023 (chart 1).

As a result, at the end of 2025, the debt ratios in most countries remained well above their 2019 levels (i.e. pre-COVID-19 levels), with the notable exceptions of Croatia, where the ratio fell substantially (‑14.6 percentage points), and Slovenia, where it fell to slightly below its 2019 level. In Bulgaria, Czechia, Poland, Romania and Slovakia, the ratio reached its highest level since the start of the global financial crisis.

Chart 1

Chart 2

Primary balances contributed to the increase in the debt ratio in all CESEE-8 countries (chart 2), which is a modest deterioration compared to 2024, when Slovenia recorded a small primary surplus. The largest primary deficits were recorded in Romania, Poland, Slovakia and Bulgaria. In these countries, primary deficits were sizeable also in structural terms ³ . This also holds true for Croatia and Slovenia, but to a smaller extent. Compared to 2024, both primary balances and structural primary balances deteriorated in six of the CESEE-8 (Bulgaria, Czechia, Croatia, Hungary, Poland and Slovenia).

Interest payments had a major adverse impact on the debt ratio in Hungary (+3.8 percentage points). However, this impact was smaller than in 2024 (by around 1 percentage points) due to a significant decrease in the interest rates (and despite the increase in the size of debt). In Romania and Poland, interest payments made a relatively strong contribution to the rise in the debt ratio as well, but this effect was smaller than in Hungary due to the lower levels of the debt ratio in both countries and a somewhat lower implicit interest rate in Poland. Still, their contribution was bigger than in 2024 in both countries.

Nominal GDP growth put a brake on debt ratios across the CESEE-8. With the exception of Poland, most countries experienced favourable nominal GDP effects stemming from inflation, as measured by the GDP deflator, although its impact was smaller than in 2024 in the majority of countries (except in Bulgaria and Slovakia). Real GDP growth had a decreasing effect on the debt ratios in all CESEE-8 countries, with the effect being stronger in Czechia and Poland than in 2024.

Overall, in 2025, the snowball effect (i.e. the net impact of interest rates, inflation and real GDP growth on the debt ratio) continued to slow the increase in debt ratios (or contributed to their decline) across the region, particularly in Croatia. However, this effect was smaller than in 2024 in all CESEE-8 except Czechia and Hungary.

Stock-flow adjustments (charts A1-A3 in the annex) drove up debt ratios in the CESEE-8 except in Slovenia. The biggest positive impact was recorded in Bulgaria, where the government beefed up its currency and deposit holdings and injected capital into the national development bank (in the magnitude of 2% of GDP each). Stock-flow adjustments contributed to the increase in the debt ratio to a comparatively large extent also in Czechia (+1.3% of GDP), Hungary, Romania and Poland (by around +1% of GDP). In Czechia, this resulted mainly from the accumulation of currency and deposit holdings (+1.1% of GDP), in Hungary mainly in connection with prepayments for EU-transfers (+1.1% of GDP), valuation effects ⁴ and the impact of the appreciation of the currency on foreign currency debt (-1.7% of GDP). In Poland, loans extended by the general government in connection with long-term prepayments for military equipment contributed to the increase in the debt ratio (+1.1% of GDP), while the appreciation of the zloty had a dampening effect on the debt ratio (-0.4% of GDP). Various elements were also at play in Romania: accumulation of currency and deposit assets, extended loans, net acquisition of other accounts receivable/payable and valuation effects.

3 High foreign currency share in non-euro area CESEE-8 government debt

In assessing fiscal risks, the structure of government debt matters in addition to its size. The currency denomination of debt may be an important source of vulnerability. As of end-2025, the government debt of the CESEE-8 countries that had adopted the euro at that time (Croatia, Slovenia and Slovakia) was almost exclusively denominated in their national currency ⁵ (chart 3). In Bulgaria, which introduced the euro at the beginning of 2026, government debt was almost exclusively denominated in euro (71%) and the Bulgarian lev (25%) at the end of 2025. Czechia also held a high share of its government debt in its national currency. In the other CESEE-8 countries that have not yet adopted the euro, foreign currency debt played a more important role. This was particularly the case in Romania, where foreign currency debt accounted for 53% of total government debt. Hungary and Poland had shares of 32% and 26%, respectively. While the elevated share of foreign currency debt combined with floating exchange rates in these countries represents a source of vulnerability, this was mitigated by the fact that foreign currency debt in Czechia and Hungary was almost exclusively denominated in euro (after hedging activities). In Romania and, to a smaller extent, in Poland, other foreign currencies played a somewhat more important role ⁶ .

Chart 3

Chart 4

There have been no major shifts in the currency structure of government debt over the past few years in most of the countries under review. Not surprisingly, the share of domestic currency debt rose sharply in Croatia when the country adopted the euro at the beginning of 2023, and a similar shift is expected for Bulgaria in the data for end-2026. The share of domestic currency debt rose modestly in Czechia as well. By contrast, the share of foreign currency debt in Hungary increased strongly after 2021 (from 23% as of end-2021 to 33% as of end-2024) and remained above the 2015–2019 average at end-2025 (32%). The foreign currency share has also risen modestly in Poland in recent years but remained lower at end-2025 (26%) than during 2015–2019 (32%). Favourably though, the foreign currency share in these four non-euro area countries declined modestly at end-2025 compared to end-2024. By contrast, in Romania, where the share of foreign currencies had been hovering slightly above 50% in recent years, it rose slightly during the past year.

The rise in the foreign currency share of the outstanding stock of government debt in Hungary largely reflected the substantially increased net issuance of debt in foreign currencies. During 2022–2025, half of the net incurrence of government debt was denominated in foreign currencies (i.e. the share of foreign currencies was substantially higher in net issuance than in the outstanding stock). By contrast, during 2015–2019, foreign currency debt was being repaid on a net transaction basis, and during 2020–2021, net issuance in foreign currencies was much smaller than in local currency. Another factor contributing to the increase was the depreciation of the forint. In Poland and Romania, foreign currency debt accounted for roughly the same share in net issuance as in the outstanding stock of government debt in recent years.

It should be noted, however, that the numbers above take into account hedging activities, and governments may in fact incur more debt in non-euro foreign currencies than indicated above. For example, statistics about the currency structure of debt securities issued by the general government (chart 4) show that the US dollar accounted for nearly 15% of total outstanding government debt securities in Hungary at end-2025 ⁷ if hedging activities are disregarded (although according to the statistics on the hedged currency structure of government debt, foreign currency debt was exclusively denominated in euro).

4 Non-residents hold substantial government debt across the CESEE-8

Non-residents are important investors in government debt across the CESEE-8 countries ⁸ . In Bulgaria, Romania, Slovenia and Slovakia, they held more than half of the total outstanding debt of the general government as of end-2025. Non-residents accounted for about 36% to 38% of total outstanding government debt in Hungary and Poland and for 25% to 30% in Czechia and Croatia (chart 5). According to data available for five countries (chart 6), the share of non-residents was particularly high (55% to 70%) in governments’ loan liabilities at end-2025, but their share was also substantial in outstanding government debt securities (20% to 60%). However, given that debt securities are by far the largest component of total government debt, non-residents hold substantially more debt securities than loans (except in Croatia).

Trends in non-residents’ share of government debt have varied across the region in recent years. Comparing the 2015–2019 average and end-2025, we see that their share declined sharply in Czechia (albeit gradually) and Poland (primarily until end-2020) and to a smaller extent in Croatia, Slovenia and Hungary. ⁹ By contrast, non-residents’ share increased in Bulgaria, Slovakia and Romania if we compare the average of 2015–2019 and end-2025 (with a rise also during 2025 in the latter two countries).

In euro area countries, significant non-resident ownership can be interpreted as a result of financial market integration within the euro area. In the other countries of the sample, the amount of government debt owned by non-residents relative to total government debt is more likely the result of various factors, such as the currency composition of government debt and the place of issuance (i.e. issuance on international versus domestic markets), the interest rate premium versus foreign markets in combination with perceived exchange rate risks, market liquidity (especially on the domestic government securities market), the role of lending by international financial organisations, etc.

There is more cross-country heterogeneity in the shares of resident investor sectors. One striking observation is the large (but in 2025 slightly decreasing) share of Hungarian government debt held by households (21.5% at end-2025). This has been attributable to the government’s deliberate strategy of seeking households as investors to reduce dependence on non-residents. For that purpose, special bonds have been sold to households on very favourable terms (e.g. attractive coupon compared to other government bonds, tax exemption, nearly cost-free redemption at face value before maturity). Elsewhere in the region, non-financial sectors (i.e. households and non-profit institutions serving households (NPISH) and non-financial corporations) only hold comparatively small shares of government debt (up to 10%), but these shares have been slowly rising in recent years.

A second major feature is that central banks hold a comparatively high share of outstanding Slovenian and Slovak government debt. This has been the result of the Eurosystem’s public sector purchase programme (PSPP) ¹⁰ , which was launched in March 2015. Consequently, central bank holdings in Slovenia and Slovakia rose in a first step during 2015–2017 and in a second step during 2020–2022. Croatia, Hungary and Poland have much smaller but still notable central bank holdings of government debt; these countries had also implemented quantitative easing programmes in response to the COVID-19 crisis. Central bank holdings have been declining in these countries in recent years not only as a percentage of total outstanding government debt but also in absolute terms.

Chart 5

Chart 6

Resident banks (i.e. “other MFIs”) are also important holders of government debt across the region. They held about 40% to 45% of government debt in Bulgaria, Czechia and Poland and about 25% to 30% in Croatia and Hungary. Their holdings were smaller in Romania, Slovenia and Slovakia (roughly 10% to 22%). At the same time, claims vis-à-vis their respective sovereigns account for a comparatively large share of total banks assets in most CESEE-8 countries. Compared to other EU member states, interlinkages between banks and the government (“sovereign-bank nexus”) seem to be relatively large in most CESEE-8 countries (Slovenia and Slovakia being exceptions). Czechia, Poland and Romania stand out in particular (chart 7). In Czechia, where banks held almost half of government debt (vertical axis in chart 7), claims on the government represented 15% of total bank assets (horizontal axis in chart 7). While Polish banks accounted for nearly 40% of Polish government debt at end-2025, claims on the Polish government made up 23% of Polish banks’ total assets. Relatively strong linkages were also present in Romania, where claims on the sovereign accounted for 28% of banks’ total assets, with these banks holding about 22% of total outstanding government debt. A strong sovereign-bank nexus may give rise to macroprudential stability concerns ¹¹ owing to potential adverse feedback loops between the two sectors in case of financial tensions. At the same time, it is important to note that these interlinkages in part reflect structural features of the CESEE-8, notably underdeveloped domestic capital markets with limited availability of both alternative long‑term domestic investors for government debt (e.g. investment or pension funds) and alternative long-term investment opportunities for banks (e.g. corporate bonds or traded shares). Also, macroprudential risks seem to be currently contained across the region, with banks benefiting from comparatively strong capitalisation and solid profitability, supporting their loss‑absorption capacity and lending resilience.

Compared to 2024, the sovereign-bank nexus strengthened particularly in Croatia and Poland, and modestly in the other CESEE-8. This was mainly attributable to the rise in the share of government debt held by banks (except in Hungary, Poland and Romania; in Romania the share of debt held by banks decreased notably). The share of claims on the general government in total bank assets increased particularly in Poland and Croatia.

Compared to the pre-pandemic average of 2015–2019 (chart 8), the sovereign-bank nexus seems to have weakened by 2025 in the majority of the CESEE-8: In Bulgaria, Hungary, Slovenia and Slovakia this was attributable to both components of the nexus. In Romania, the large decline in the share of government debt held by banks (vertical axis in chart 8) was nearly paralleled by a substantial rise in the share of claims on the general government in total bank assets (horizontal axis in chart 8). The opposite goes for Croatia, though the decline was less pronounced. The sovereign-bank nexus strengthened in this longer comparison in Czechia and to a smaller extent in Poland, particularly on account of the rise in the share of government debt held by banks.

Chart 7

Chart 8

Finally, resident non-monetary financial corporations (e.g. investment funds, pension funds, insurance corporations, financial auxiliaries, etc.) ¹² held a comparatively large share of government debt in Croatia and Czechia (25% and 22%, respectively) at end-2025. The high share in Croatia is likely supported by the relatively large size of occupational pensions and personal pensions linked to employment (second- and third-pillar schemes) as a percentage of GDP. The share of resident non-monetary financial corporations was also relatively large in Romania (19%), followed by Poland, Hungary (about 12% to 13%) and Bulgaria (9%). By contrast, the share of this sector was negligible in Slovenia and Slovakia, the two euro area countries in the sample with the biggest shares of outstanding government debt ¹³ held by non-resident investors and central banks.

5 Government debt mostly long term in the CESEE-8

Detailed data about the maturity structure of government debt are relatively scarce for the CESEE-8. Data by original maturity (chart 9) suggest that governments in the CESEE-8 cover their financing needs almost exclusively with long-term debt (i.e. debt instruments with maturities of over one year). Available data for the four CESEE-8 countries with more detailed information ¹⁴ show that debt with original maturities between 10 and 30 years is the preferred choice, followed by debt with maturities of between 5 and 10 years. Debt with original maturities of over 30 years is relatively uncommon, as are maturities of between 1 and 5 years. The composition of debt by original maturity has changed somewhat over the past few years. In Croatia, there has been a notable shift away from medium maturities (5–10 years) to both shorter maturities (up to 5 years) and longer maturities (10–30 years). By contrast, Czechia has seen more issuance in the medium segment at the cost of both shorter and longer maturities. The share of longer maturities (10–30 years) also decreased in Romania and Slovakia, mainly to the benefit of the medium maturity segments (5–10 years).

Chart 9

Chart 10

Most governments in the CESEE-8 lengthened the average remaining maturity of government debt over the past decade (chart 10), thereby reducing rollover risks. Given the high share of fixed-rate government debt (see below), this contributed to a reduction in interest rate risk, while at the same time also reducing the flexibility to take advantage of periods with low interest rates. The average maturity ranged between around 6 and 9 years at the end of 2025 (with the exception of Bulgaria and Hungary: 4.4 and 5.3 years, respectively), compared to between around 4 and 8.5 years in 2015.

Chart 11

More detailed data on government debt by remaining maturity (chart 11) show that, at the end of 2025, short-term debt with a remaining maturity of up to one year accounted for about 10% of government debt in most CESEE-8 countries. Major exceptions were Croatia, Hungary and Romania with a share of around 16% to 17%. The share of short-term debt increased modestly in 2025 in several CESEE-8 countries (Bulgaria, Czechia, Hungary, Romania and Slovenia).

Debt with remaining maturities of at least five years accounted for half or more of total government debt in most of the CESEE-8 at end-2025 (except Hungary and Poland). The share reached around two thirds in Bulgaria and Slovakia. According to more granular data for four countries, remaining maturities of between five and ten years are the most common (30% to 35% of the total), followed by maturities of 10–30 years. Very long remaining maturities (30 years or more) continue to be extremely rare. In Hungary and Poland, instruments with remaining maturities of at least five years accounted for slightly less than 40% of government debt, with most of the government debt (around 47% to 50% of the total) maturing in one to five years in these two countries.

Changes in the remaining maturity structure of government debt were very limited in 2025. They mainly concerned a modest decline in the already comparatively low share of maturities of at least five years in Hungary and Poland, and a modest decline in the relatively high share of this segment in Slovenia (in Hungary and Poland primarily to the benefit of the 1- to 5-year segment, in Slovenia to the benefit of the short-term segment).

6 Implicit interest rate on government debt much higher since 2021

Based on data for seven CESEE-8 countries, variable rate debt accounted for a very modest share of government debt with remaining maturities of more than one year ¹⁵ in Bulgaria, Romania, Slovenia and Slovakia, representing less than 5% of total outstanding debt in this maturity segment at end-2025 (chart 12). This share was bigger in Czechia and Croatia (around 12% to 13%) and particularly high in Poland (22%). The numbers are very similar for debt with original maturities of more than one year. The most notable development over the past few years has been the sharp decline in the share of variable rate debt in Bulgaria.

Chart 12

Chart 13

More granular data are available for the biggest part of government debt, i.e. government debt securities (chart 13): Fixed-rate government bonds with remaining maturities of more than one year accounted for the bulk of government securities at end-2025 in all CESEE-8 countries (80% to 95%, except in Poland and Hungary at around 60% and 67%, respectively). Longer-dated bonds (with maturities of more than five years) with fixed interest rates had a relatively large share (around 60% to 70%) in Bulgaria, Slovenia and Slovakia, while their share was lowest in Poland (24%) and Hungary (34%). In the latter two countries, variable rate government bonds accounted for about 25% to 30% of outstanding debt securities, a relatively high level. In Czechia, this share was lower at around 12%. Zero coupon instruments played a subordinate role, with only Croatia using them to any notable extent (14%; with remaining maturities of up to one year). The most remarkable changes in this structure over the past few years have been the notable rise in the share of fixed-rate bonds in Bulgaria and, to a smaller extent, in Slovakia.

Chart 14

The implicit interest rate on government debt ¹⁶ decreased gradually and continuously across the CESEE-8 prior to the COVID-19 crisis (chart 14). The trend ended in 2021–2022 as monetary tightening to contain recovery-induced inflationary pressures after the pandemic led to increases in the general interest rate level. The following rise in the implicit interest rate was largest in Hungary and Poland, followed by Romania and Czechia. A likely contributing factor to this development was the relatively large share of instruments with variable interest rates and/or short remaining maturities in Poland, Hungary and Czechia when interest rates started to rise. In Hungary and Romania, concerns about fiscal risks and sustainability, as reflected in sharp increases in interest premiums, may also have played a role. The implicit interest rate fell notably in Hungary in 2025 compared to 2024 (in part owing to the decrease in inflation-linked coupons), while it remained broadly unchanged elsewhere in the region. The main exception was Bulgaria, where the implicit interest rate rose by 1.3 percentage points, reflecting the rise in the average coupon rate of outstanding government bonds compared to 2024 amid accelerating inflation).

7 Sizeable contingent liabilities add to fiscal risks in some of the CESEE-8

The Eurostat database contains data on four types of contingent liabilities: guarantees assumed by the government; liabilities of government-controlled entities classified outside the general government; government liabilities related to public-private partnerships (PPPs) recorded off balance sheet ¹⁷ and non-performing loans (NPLs) provided by the general government. The latest available data on contingent liabilities are from 2024.

Guarantees assumed by the general government represent one type of contingent liabilities. These may add to fiscal vulnerabilities if they are called. According to Eurostat data, the stock of guarantees was particularly large in Hungary at the end of 2024, amounting to 13% of GDP (chart 15). This represented a modest increase compared to 2023 but a doubling compared to the average of 2015–2019. The stock of guarantees was significantly lower but still sizeable and rising in Poland (around 5% of GDP), followed by Romania and Slovenia (both slightly above 3% of GDP). In Hungary, Poland and Slovenia one-off guarantees, i.e. those based on individual decisions to support specific entities (such as development banks, guarantee institutions) or projects, accounted for the bulk (at least 70%) of the total. Standardised guarantees, i.e. those linked to government schemes that support certain sectors or activities and cover many smaller contracts (e.g. lending schemes to households and corporations), dominated in Romania and they were also comparatively large as a percentage of GDP in Hungary.

Chart 15

Chart 16

Liabilities of entities controlled by, but classified outside, the general government represent additional contingent liabilities. These were particularly large in Poland (more than 40% of GDP) at end-2024, followed by Slovenia, Hungary (14% to 18% of GDP), Croatia and Czechia (11% to 12%). In Poland and Croatia, these were mostly liabilities of state-owned banks, in Hungary and Czechia the liabilities of non-financial state-owned enterprises (SOEs) dominated. In Slovenia both groups had a roughly equal share. The size of these liabilities remained stable over the past few years in most CESEE-8 countries, with the notable exception of a decrease in Slovenia (inter alia related to the sale of the state’s stakes in two major banks) and Bulgaria. When considering these liabilities, it should be borne in mind that these SOEs also have assets that are not captured by the statistics. Moreover, according to the statistics, the bulk of the liabilities of non-financial SOEs were owed by profitable entities in the majority of countries: the liabilities of loss-making non-financial SOEs were comparatively bigger (as a percentage of GDP) only in Bulgaria and Poland, at 2.5 and 3% of GDP, respectively.

The liabilities of off-balance public-private partnerships represent a third group of contingent liabilities for the general government. These liabilities are mostly in connection with long-term construction contracts with assets recorded outside government accounts (e.g. motorway projects). These liabilities were generally low across the CESEE-8 at end-2024, with somewhat higher but decreasing values only in Slovakia and to a smaller extent in Hungary.

Finally, the stock of non-performing loans provided by the general government was also generally low across the region at the end of 2024. The highest value was recorded in Croatia at 0.8% of GDP. Favourably, in those countries where NPLs provided by the general government had previously been comparatively large (Czechia, Croatia, Slovenia), the share has considerably declined by 2024.

Chart 17

Chart 18

8 Room to improve fiscal governance

Sound fiscal governance can contribute to mitigating fiscal risks by ensuring a transparent, predictable and sustainable conduct of fiscal policy. According to the European Commission’s fiscal governance database, which provides a quantitative assessment of (1) medium-term budgetary frameworks, (2) independent fiscal institutions and (3) fiscal rules, the CESEE-8 countries show a mixed picture.

The CESEE-8 seem to score comparatively well in terms of the fiscal rule index, which assesses the strength of individual national fiscal rules based on their coverage, legal strength, monitoring and corrective procedures and flexibility (chart 19). According to the latest available data for 2024, five of the CESEE-8 countries perform better than the median on the basis of this indicator. By contrast, Slovenia and Hungary are the weakest performers in the EU-27.

In the case of Hungary, the main constraints are: the lack of real-time monitoring of fiscal developments; the lack of an independent body providing or endorsing the official macroeconomic or budgetary forecasts on which the annual budget is prepared; insufficient correction mechanisms in case of deviation from the rules; the absence of clearly defined escape clauses which are in line with the Stability and Growth Pack; the lack of budgetary margins defined in relation to the rule or safety margins linked to the medium-term budgetary objective (MTO) which are enshrined in national legislation; and the lack of targets defined in cyclically adjusted terms or any other ways which account for the cyclical position of the economy.

In Slovenia, the main impediments are: the lack of real-time monitoring; the absence of budgetary margins defined in relation to the rule or safety margins linked to the MTO which are enshrined in national legislation; and the lack of exclusions from the rule in the form of items that fall outside authorities’ control at least in the short term (e.g. interest payments, unemployment benefits).

However, this assessment needs to be treated with some caution since the current methodology for calculating of the fiscal rule index puts countries with more fiscal rules at an advantage. The overall index score is calculated as the sum of the weighted scores ¹⁸ for individual rules. Hence, country A with more fiscal rules can achieve a higher overall index score than country B with fewer fiscal rules, even if the scores for the individual rules of country A are worse than for country B. This is exactly the case for Hungary (three fiscal rules) and Slovenia (one fiscal rule, with a better score than for any of the three scores for Hungary) in 2024, but this issue may affect the whole ranking of the EU-27 countries. Similarly, the overall fiscal rule score of a country can improve upon the introduction of an additional rule, even if this new rule has a lower score than the previously existing one(s).

The Medium-Term Budgetary Framework Index evaluates the strength of national medium-term budgetary frameworks based on their coverage, bindingness over the annual budget, involvement of key stakeholders and level of details. On this indicator, Romania, Slovakia, Bulgaria and Croatia are mid-range performers among the EU-27, while Czechia and particularly Poland and Hungary rank last (chart 20). The index values of the two weakest performers, Hungary and Poland, are dragged down by no, or only weak, alignment between medium-term fiscal plans and annual budgets; the lack of objectives/ceilings included in the national medium-term budgetary plan that are fixed in advance for a number of years and are not changed; the lack of well-defined actions to be taken in case of deviations from the medium-term budgetary plan; the lack of involvement of an independent fiscal institution in the preparation of the medium-term budgetary plan; and the lack of detailed explanations on the budgetary impact of alternative macroeconomic scenarios in the medium-term budgetary plan. In addition, in the case of Hungary, missing coordination between sublevels of government when setting the multi-annual targets and no involvement of the national parliament in the preparation of the medium-term budgetary plan represent a weakness. In Poland, the lack of a detailed breakdown of the total expenditure and revenue projections based on unchanged policies; the lack of an explanation of revenue and expenditure projections and their components; and the lack of a quantification of the impact of reforms over the time span of the plan represent additional shortcomings.

Concerning other CESEE-8 countries, important areas requiring improvements are: insufficient alignment of medium-term fiscal plans and annual budgets, the lack of objectives/ceilings included in the national medium-term budgetary plan that are fixed in advance for a number of years and are not changed, and the lack of well-defined actions to be taken in case of deviations from the medium-term budgetary plan.

Chart 19

Chart 20

The third component of the fiscal governance database is the Independent Fiscal Institutions Index, which assesses the scope of the legal mandate and activities of national independent fiscal institutions (IFIs). This scope may range from monitoring fiscal rule compliance to producing macroeconomic and budgetary forecasts and assessing long-term fiscal sustainability of fiscal policy. In the case of this indicator, six of the eight countries under review score among the worst performers across the EU-27 (Croatia and Poland are the weakest performers in the EU-27). Only Romania and Slovenia are better than the EU-27 median.

For Poland, the main shortcomings identified were: suboptimal monitoring of compliance with numerical fiscal rules including the lack of involvement of the IFI in triggering a correction mechanism; insufficient involvement of the IFI in macroeconomic forecasting including the lack of a reconciliation procedure in case of diverging views between the government and the IFI; insufficient involvement of the IFI in budgetary forecasting and policy costing (i.e. the estimate of the financial impact of policy measures); the lack of long-run sustainability analyses; and the lack of the active promotion of fiscal transparency.

More generally in the region, the functioning of IFIs is being affected by these major shortcomings: the lack of proper involvement of IFIs in macroeconomic forecasting including the lack of a reconciliation procedure in case of diverging views between government and the IFI; the lack of ex post evaluations; and the lack of quantitative policy costing and of long-run sustainability analyses.

Chart 21

9 Conclusions

This report reviewed general government debt levels and structure across eight EU member states in Central, Eastern and Southeastern Europe (CESEE-8) during 2025. It provided some insight into selected sources of vulnerability associated with the structure of government debt and with contingent liabilities. The report also briefly reviews the main characteristics of fiscal governance frameworks in the region.

In 2025, the general government debt ratio (as a percentage of GDP) increased in six of the CESEE-8 countries discussed in this report. These six countries were the same which had seen a rise in their debt ratios in 2024. Primary balances contributed to the higher debt ratios in all CESEE-8 countries, which represents a modest deterioration compared to 2024. Moreover, compared to 2024, both primary balances and structural primary balances deteriorated in six of the CESEE-8. Interest payments had a major adverse impact on the debt ratios of several countries, most notably Hungary, followed by Romania and Poland. However, in Hungary this adverse impact was smaller than in 2024. Nominal GDP growth put a brake on debt ratios across the CESEE-8. With the exception of Poland, most countries experienced favourable nominal GDP effects stemming from inflation, as measured by the GDP deflator, although its impact was smaller than in 2024 in the majority of countries. Real GDP growth had a favourable impact on the debt ratios in all CESEE-8 countries, particularly in Czechia and Poland, where the effect was greater than in 2024. Stock-flow adjustments drove up debt ratios in the CESEE-8 except in Slovenia.

As of end-2025, the government debt of the CESEE-8 countries that have adopted the euro was almost exclusively denominated in their national currency, the euro. In Bulgaria, which introduced the euro at the beginning of 2026, government debt was almost exclusively denominated in euro and the Bulgarian lev, and also Czechia held a high share of its government debt denominated in its national currency. In the other CESEE-8 countries that have not yet adopted the euro, foreign currency debt played a more important role, especially in Romania. Although foreign currency debt in combination with variable exchange rates can be a source of vulnerability, this risk seems to be mitigated by the fact that the euro accounts for the bulk of foreign currency debt in these countries.

Non-residents are important investors in government debt across the CESEE-8 countries. In Bulgaria, Romania, Slovenia and Slovakia, they held more than half of the total outstanding debt of the general government as of end-2025. Concerning other counterpart sectors, households are important investors in Hungary, which is the result of a deliberate strategy pursued by the country’s debt management agency. A second major feature is the high but gradually decreasing share of outstanding Slovenian and Slovak government debt held by central banks. Resident banks are also important holders of government debt across the region. This has given rise to some concern about the sovereign-bank nexus, which, however, in part reflects structural features of the CESEE-8. Currently, related risks appear to be mitigated by solid banking sectors.

Most governments in the CESEE-8 have lengthened the average remaining maturity of government debt over the past decade, thus reducing their rollover risks. At the end of 2025, short-term debt with remaining maturities of up to one year accounted for only about 10% of government debt across the CESEE-8. The major exceptions were Croatia, Hungary and Romania with shares of around 16% to 17%. Debt with remaining maturities of at least five years accounted for half or more of total government debt in most of the CESEE-8 countries except Hungary and Poland, but very long-term maturities (i.e. 30 years or more) are still very rare in the region.

Debt instruments carrying a variable interest rate make up a very modest part of government debt in Bulgaria, Romania, Slovenia and Slovakia. However, their share is larger in Czechia and Croatia, and particularly high in Poland. According to more granular data on government securities (the largest component of government debt), fixed-rate bonds with remaining maturities of more than one year accounted for the bulk of government securities at end-2025 in all CESEE-8 countries. Longer-dated bonds (with maturities of more than five years) with fixed interest rates had a relatively large share (around 60% to 70%) in Bulgaria, Slovenia and Slovakia. Their share was lowest in Poland (24%) and Hungary (34%), where variable rate government bonds accounted for about 25% to 30% of outstanding debt securities. Zero coupon instruments play a subordinate role, with only Croatia using them to any notable extent.

The implicit interest rate on government debt had increased from 2021 to 2024 in all CESEE-8 countries, with the biggest increases observed in Hungary and Poland, followed by Romania and Czechia. The implicit interest rate fell notably in Hungary in 2025 compared to 2024 (in part owing to the decrease in inflation-linked coupons), while it remained broadly unchanged elsewhere in the region, except for Bulgaria, where it rose substantially (reflecting the rise in the average coupon rate of outstanding government bonds compared to 2024 amid accelerating inflation).

Sizeable contingent liabilities add to fiscal vulnerabilities in some of the CESEE-8 countries. The stock of government guarantees was particularly large in Hungary at the end of 2024, having increased from 2023 and doubled compared to the average of 2015–2019. It was significantly lower but still sizeable and rising in Poland, followed by Romania and Slovenia. Liabilities of entities controlled by, but classified outside, the general government were particularly large in Poland at end-2024, followed by Slovenia, Hungary, Croatia and Czechia. However, it should be borne in mind that these state-owned entities also have assets that are not captured by the statistics. The liabilities of off-balance public-private partnerships, which mostly occur in connection with long-term construction contracts, were generally low across the CESEE-8 at end-2024, as was the stock of non-performing loans provided by the general government.

Concerning their fiscal governance frameworks, the CESEE-8 present a mixed picture. They score comparatively well in terms of the Fiscal Rule Index: Five of the CESEE-8 countries perform better than the median in the EU-27. By contrast, Slovenia and Hungary are the weakest performers in the EU-27. However, this assessment needs to be treated with some caution since the current methodology of the calculation of the Fiscal Rule Index puts countries with more fiscal rules at an advantage. With regard to the Medium-Term Budgetary Framework Index, Romania, Slovakia, Bulgaria and Croatia are among the mid-range performers in the EU-27, while Czechia and particularly Poland and Hungary rank last. The following areas need work in the majority of the CESEE-8 countries: insufficient alignment of medium-term fiscal plans and annual budgets, the lack of objectives/ceilings included in the national medium-term budgetary plan that are fixed in advance for a number of years and are not changed; and the lack of well-defined actions to be taken in case of deviations from the medium-term budgetary plan are important areas where the majority of the CESEE-8 countries would need improvements. As for the Independent Fiscal Institutions Index, six of the eight countries under review score among the worst performers across the EU-27 and only Romania and Slovenia are above the EU-27 median.

Major shortcomings concerning the functioning of IFIs in most of the CESEE-8:

• the lack of proper involvement of independent fiscal institutions (IFIs) in macroeconomic forecasting, including the lack of a reconciliation procedure in case of diverging views between the government and the IFI;

• the lack of ex post evaluations; and finally:

• the lack of quantitative policy costing and of long-run sustainability analyses.

10 Annex

Chart A1

Chart A2

Chart A3