Call for applications: Klaus Liebscher ­ Economic Research Scholarship

The scholarship program targets Austrian and international experts with a ­proven research record in economics and finance, and postdoctoral research experience. Applicants need to be in active employment and should be interested in broadening their research experience and expanding their personal research networks. Given the OeNB’s strategic research focus on Central, Eastern and South­eastern Europe, the analysis of economic developments in this region will be a key field of research in this context.

The OeNB offers a stimulating and professional research environment in close proximity to the policymaking process. The selected scholarship recipients will be expected to collaborate with the OeNB’s research staff on a prespecified topic and are invited to participate actively in the department’s internal seminars and other research activities. Their research output may be published in one of the department’s publication outlets or as an OeNB Working Paper. As a rule, the consultancy services under the scholarship will be provided over a period of two to three months. As far as possible, an adequate accommodation for the stay in Vienna will be provided. ¹

Applicants must provide the following documents and information:

• a letter of motivation, including an indication of the time period envisaged for the consultancy
• a detailed consultancy proposal
• a description of current research topics and activities
• an academic curriculum vitae
• an up-to-date list of publications (or an extract therefrom)
• the names of two references that the OeNB may contact to obtain further information about the applicant
• evidence of basic income during the term of the scholarship (employment contract with the applicant’s home institution)
• written confirmation by the home institution that the provision of consultancy services by the applicant is not in violation of the applicant’s employment ­contract with the home institution

Applicants will be notified of the jury’s decision by end-November.

Developments in selected CESEE countries

Coronavirus overruns the region ² , ³ , ⁴

1 Regional overview

This report gives an overview of the economic situation prior to the corona­virus pandemic and sheds light on macroeconomic strengths and weaknesses and the associated policy space for the CESEE region. It then sketches the development of the coronavirus crisis in CESEE and the policy measures implemented so far.

Pre-coronavirus: moderating though still broadly robust economic growth amid often slowly rising price pressures

GDP figures for the second half of 2019 were broadly in line with the trends outlined above (see table 1). Real GDP growth in Turkey accelerated to 1.9% (quarter on quarter) in the fourth quarter of 2019 – by far the strongest reading in the CESEE region. Decelerating economic momentum was reported for the ­CESEE EU Member States, where average GDP growth came in at 0.7% (quarter on quarter) in the final quarter of 2019. This was a rather weak reading compared to the dynamism of the past three years. GDP growth in Russia amounted to 0.6% (quarter on quarter) in the fourth quarter of 2019 and was broadly in line with CESEE EU Member States’ figures.

Growth remained firmly rooted in domestic demand. Private consumption made the strongest contribution to GDP growth in seven of the ten countries ­under observation. It continued to benefit from benign labor market conditions and swift wage growth. Furthermore, consumer sentiment was robust until most recently and consumer credit expanded swiftly.

Unemployment rates have been falling consistently in recent years, from an average level of around 10% in early 2013 to 3.4% in February 2020. This ­represents the lowest reading since the start of transition. Furthermore, employment kept expanding throughout most of the region up to the outbreak of the pandemic, ­contributing to a convergence of employment rates toward euro area levels.

Investment also supported growth throughout most of CESEE in the second half of 2019. Capital spending was bolstered by high capacity utilization, favorable financing conditions, robust construction activity and the availability of EU ­funding in many countries. In some countries, however, poor export prospects already led companies to postpone or scale down investment. This was especially the case in Slovenia, where growth in capital formation dipped into the red. A notable ­deceleration of investment growth was also observed in Croatia, Hungary and ­Poland, coming down from very high levels, however.

Export growth continued to soften in the review period, mirroring declining world trade amid lower international demand. Given the high import content of CESEE’s export production, import growth was also somewhat lower. On ­balance, the external sector often contributed negatively to GDP dynamics, especially in Russia and Turkey.

Despite somewhat lower growth rates, means of production – especially labor – continued to be utilized in full in the second half of 2019. High GDP growth, ­emigration and a lack of skilled workers had led to labor shortages in the past years. This translated into high wage growth rates that reached up to 12% annually. The situation has relaxed in recent months and both wage growth and labor shortages came down somewhat (e.g. annual nominal hourly wage growth declined to around 9% by the end of 2019). Nevertheless, the region’s economies mostly ran above full potential. The positive output gap for the CESEE EU Member States in 2019 was estimated at an average 2% of GDP by the European Commission.

This translated into higher inflation rates. Average HICP inflation came in at 3.7% in January and 3.6% in February 2020 in the CESEE EU Member States, the highest level since late 2012. While this in part reflects higher energy prices up until February, core inflation was on the rise, too. In February 2020, it reached an average of 3.3%. Several central banks (e.g. in the Czech Republic, Hungary, ­Poland and Romania) have missed their inflation targets in recent months (at least temporarily).

In Turkey, inflation came down from around 25% in late 2018 to a three-year low of 8.6% in October 2019. However, price growth subsequently reaccelerated to 12.4% in February 2020, owing in part to unfavorable base effects. Among the HICP components, it was especially industrial goods prices that fueled the ­increase.

Russia was the only country with a clear downward trend in inflation in recent months. In February 2020, price growth declined to 2.3% (from 4% in September 2019), well below the central bank’s target of 4%. The most important building block for this development was a base effect from a value-added tax increase in January 2019. Other disinflationary factors included a decline in prices of food products and non-food goods.

Regional central banks reacted to inflation developments. The Czech central bank increased its policy rate by 25 basis points to 2.25% in February 2020 after having missed its inflation target for three months in a row.

The Turkish central bank decreased its policy rate in four steps, from 16.5% in early October 2019 to 10.75% in late February 2020, reversing most of its ­crisis-induced rate hikes carried out after the turbulences in mid-2018. It argued that the improvement in macroeconomic indicators supported the fall in the ­country’s risk premium. The exchange rate of the Turkish lira vis-à-vis the euro remained broadly stable from October 2019 until the end of February 2020.

The Russian central bank cut its key rate in three steps, from 7% in early ­October 2019 to 6% in late February 2020, citing disinflationary pressures and – in its February move – rising risks of a substantial global economic slowdown.

Coronavirus crisis: macroeconomic strengths and weaknesses and policy space for the CESEE region

The period before the great financial crisis was characterized by twin deficits in the external and public sector accounts in many countries of the CESEE region. At the outbreak of the coronavirus crisis, the region reported a combined current and capital account surplus of 2.2% of GDP (end-2019). Among the individual ­CESEE countries, only Romania recorded a notable current account deficit in 2019 (see chart 1). The external adjustment of the region was driven predominantly by better outcomes in the goods and services balances. The momentum in the trade balance initially rested on strongly reduced domestic demand at the height of the global financial crisis that led to a substantial decline of imports. Trade in goods, however, also benefited from the close integration of the CESEE (EU) region in international supply chains as the associated rise in competitiveness bolstered ­export growth after the acute crisis years. In recent quarters, pronounced (­productivity-adjusted) wage increases in the manufacturing sector have already shaved off some of the competitive edge of the region, however.

Yet, some CESEE countries belong to the world’s most tightly integrated in terms of global value chains (see chart 2), which implies a high vulnerability in the current crisis. All countries under observation are more strongly integrated than the OECD average, and the degree of integration in countries like the Czech ­Republic, Hungary and Slovakia is even five to six times as high. Global production networks could be an important transmission mechanism for shocks throughout the global economy and have already been disrupted at an early stage by the ­stoppage of production in China. Furthermore, global value chains not only ­transmit shocks in supply, they also transmit shocks in demand. As an example, the automobile industry in CESEE will not only be impaired by a lack of Chinese ­inputs into domestic production, but also by a decline in Chinese auto demand via falling sales of international (especially German) car producers that usually obtain inputs from CESEE.

International trade also acts as a transmission mechanism for coronavirus-­related shocks. Chart 3 shows that the degree of openness has mostly increased since the global financial crisis and is especially high in the Central European EU Member States (up to three times as high as in the EU-15). Even though the ­geographical distribution of CESEE countries’ trade remains very much centered on the EU-15, the weight of China and the U.S.A. as the largest and second-largest global economies (in terms of purchasing power parity) in final demand for CESEE products must not be ignored given the strong integration in international value chains.

Further, tourism will be one of the sectors most strongly affected by the ­coronavirus recession; this warrants a closer look at the importance of this sector for CESEE (see chart 4). In terms of dependence on tourism, the CESEE countries can be broken down into roughly three groups. The first group includes Poland, Russia, Romania, Slovakia, the Czech Republic and Hungary. These countries ­reported a share of tourism in total GDP of between 5% to 8%. The second group includes Bulgaria, Turkey and Slovenia, with a share of tourism of around 11% to 12% in total GDP. Finally, Croatia is in a league of its own, with a share of tourism in GDP and employment of around 25%, making it the country most dependent on tourism in the whole region (and in fact also throughout much of Europe).

Finally, we take a look at foreign capital flows and external debt to get an idea of the sustainability of the external position of the CESEE region. For the region as a whole, external liabilities in the international investment position (IIP) ­increased by some 20 percentage points of GDP between 2008 and 2019 (see chart 5), mostly on account of FDI inflows (+16.3 percentage points of GDP) but also on account of higher portfolio liabilities (+10.7 percentage points of GDP). Other ­investments were the only component of IIP liabilities that posted a notable decline (–6.4 percentage points of GDP). The latter mainly reflected two things: The ­deleveraging of international banking groups in the region and the change in the refinancing structure of CESEE banking sectors away from external liabilities to domestic deposits. The outlined dynamics in IIP liabilities translated into a somewhat higher gross external debt of the CESEE region (+4.9 percentage points of GDP, reaching 48.7% of GDP in 2019; see chart 6). At the same time – supported by current account surpluses – reserves (including gold) also trended higher (+7.4 percentage points of GDP) and covered more than 50% of total external debt by the end of 2019.

Developments at the country level were heterogeneous, however. Between 2008 and 2019, a large degree of external deleveraging was observed for Hungary and Bulgaria (and to a lesser extent also for Croatia). In all three countries, the ­decline in IIP liabilities was almost exclusively driven by other investments, leading to a corresponding decline in gross external debt. The Czech Republic and ­Slovakia, by contrast, reported an increase in their external liabilities of more than 50 percentage points of GDP. Strongly rising portfolio and other investment ­liabilities – amid an also substantial upward trend in FDI – pushed up external debt in the two countries. In Turkey and Poland, the increase in IIP liabilities amounted to some 25 percentage points of GDP in the past decade. While, in the case of ­Poland, the increase rested on FDI and portfolio flows only, all IIP components contributed to the rise in Turkey. Hence, external debt in Turkey also posted a relatively large increase. Comparatively little change could be observed in ­Slovenia, Romania and Russia, where increases in FDI and decreases in other investments broadly leveled out.

Summing up, the region entered the current downturn with its external ­accounts in broadly solid shape. Combined current and capital accounts were mostly in surplus and an increase in external debt was buffered by an even larger increase in reserves. On the country level, Romania was the only country to ­report a notable current account deficit in 2019. Most of the shortfall, however, was ­covered by capital (mostly FDI) inflows. Given the high degree of openness and the strong integration in global value chains, Central European countries (especially Hungary, the Czech Republic and Slovakia) will most likely be more affected by a slowdown in global demand. The sudden stop in international mobility will particularly dampen the Croatian economy, as the country is highly dependent on tourism. Finally, Turkey experienced a rather large increase in external debt amid broadly unchanged international reserves, leaving the coverage of external debt by foreign exchange reserves among the lowest of the region. Furthermore, exchange rate risks are high given the Turkish lira’s strong volatility in recent years and the country is also rather susceptible to the effects of a sudden stop in tourism.

A central question on the eve of the recession is whether fiscal policy has enough room for maneuver to weather the fallout from the coronavirus crisis. ­Ideally, policymakers would want to rely on automatic fiscal stabilizers, discretionary fiscal stimuli and favorable financing conditions to shield companies, workers and households from the most severe consequences of the coronavirus-induced economic slump. The ability to do so crucially depends on a country’s fiscal space.

A look at general government net lending in the region reveals that most ­CESEE countries managed to substantially bring down headline deficits as well as ­cyclically adjusted budget deficits from the heights after the global financial crisis (see chart 7). Measures to improve revenue collection combined with strong economic growth in recent years have underpinned a notable improvement in public finances. In fact, most countries reported notably lower deficits in 2019 than they did before the outbreak of the global financial crisis in 2008.

The reform momentum, however, stalled somewhat in 2019. Six of the ten countries under observation reported higher headline deficits (lower surpluses) than in 2018, despite continuing robust economic momentum. An expansionary fiscal stance was also documented by the deterioration in the (projected) cyclically adjusted deficits of most CESEE EU Member States.

In Romania, the headline deficit increased to 4.3% of GDP, 1.3 percentage points of GDP above the Maastricht threshold. Romania has been in a significant deviation procedure (SDP) since 2017. In November 2019, the European ­Commission commented that the Romanian authorities do not intend to act upon the recommendations issued within the SDP and that the new pension law poses a significant upward risk to the public deficit in 2020 and beyond. Against this background, an excessive deficit procedure (EDP) was launched in spring 2020. ­Hungary – the second country subject to an SDP – was on a better trajectory ­before the coronavirus hit the region. In its November 2019 report, the European Commission noted that the Hungarian economy was experiencing good times, with an improvement in the overall fiscal situation.

After the start of transition, CESEE public sectors had been known for their low indebtedness for a long time. The global financial crisis thoroughly altered this situation, however (see chart 8). Discretionary fiscal spending and large-scale banking sector support in several countries drove up general government debt to more “normal” European levels after 2008. Strong economic dynamics in recent years reversed some of these increases, but debt levels mostly remain substantially higher than ten years ago. In Turkey, government debt even displayed a clear ­upward trend in the past two years as the country went through a recession. Russia reports the by far lowest public debt and the country’s National Welfare Fund holds about USD 150 billion in liquid assets (9% of GDP).

For fiscal sustainability, it is not only the level of debt that matters but also its composition. With respect to the creditor structure, a large share of foreign ­investors can drive up risk premiums and impede access to market funding and favorable financing conditions in turbulent times. Domestic investors (such as ­pension funds), by contrast, are usually more long-term oriented and less prone to swings in sentiment. The share of domestic investors is relatively low in CESEE, partly because comparatively shallow capital markets make it hard to issue ­sufficient portions of sovereign debt domestically. Chart 9 shows that nonresidents on ­average held close to 50% of total debt in CESEE in 2019. In Slovakia and Slovenia, the shares reached close to 70% and displayed a notable upward trend.

If we look at the currency denomination of public debt, we find that the share of foreign currency-denominated debt in total debt is rather high in many CESEE countries given the strong presence of nonresident investors (see chart 10). ­Exchange rate risks, however, are mitigated by two factors: The euro is the most important foreign currency in the CESEE EU Member States, given their close trade, financial and business cycle linkages with the euro area and the associated hedging possibilities. Second, the stock of euro-denominated debt is ­especially high in those countries that have pegged their currencies to the euro or whose ­exchange rate policy is strongly focused on the euro (Bulgaria, Romania, Croatia). Among the non-EU countries, a high share of (mostly U.S. dollar-­denominated) foreign currency debt can be observed in Turkey. In Russia, the share of foreign currency-­denominated government debt has decreased notably in recent years as sanctions have made tapping international markets increasingly ­difficult for the country.

Regarding the ratings for long-term foreign currency sovereign debt, Slovakia, Slovenia, the Czech Republic and Poland currently show the strongest ratings (see chart 11). Bulgarian, Hungarian, Romanian and Russian government bonds are assessed as moderately risky. While Standard & Poor’s applies the moderate risk category also to Croatia, Moody’s sees a substantial credit risk for the country. Turkey is the lowest-rated country in the CESEE region, with credit risk being deemed high by both rating agencies. Comparing 2008 and today, current ratings by Moody’s are often somewhat weaker than before the global financial crisis, while no major changes can be observed in the country risk assessments released by ­Standard & Poor’s.

Short-term debt issuance is usually associated with lower funding costs but – at the same time – it is also associated with higher interest and rollover risks ­especially in times of turbulence. Since the global financial crisis, the share of short-term debt in total debt has fallen substantially in all CESEE countries except Russia (see chart 12). In some cases, it has reached long-term lows. This development was ­related to favorable market conditions in an environment of prolonged monetary accommodation that favored the issuance of long-term debt.

Summing up, despite some recent deterioration, CESEE countries’ budget deficits are generally lower than in 2008, and three countries even reported ­surpluses in 2019. However, public debt levels have increased substantially throughout most of the region. The structure of public debt remains skewed toward nonresident investors and – at least in some countries – foreign currency-denominated debt. On a positive note, refinancing risks have been reduced considerably due to a lengthening of debt maturities. Therefore, the recent increase in government bond spreads (see below) does not pose an immediate threat.

On the country level, Romania and Turkey seem to be exposed most strongly to public finance risks. Romania’s public finances have deteriorated throughout the past years and the country is currently subject to an excessive deficit procedure (the only ongoing EDP in the EU). Its debt level has nearly tripled since 2008. ­Furthermore, it scores relatively high in terms of the share of nonresident investors and foreign currency public debt. Turkey has by far the weakest country rating in the CESEE region and reports a relatively large (and rising) share of U.S. dollar-­denominated debt amid pronounced exchange rate volatility in recent years.

In recent years, credit growth accelerated in an environment of strong GDP growth, ample liquidity and low interest rates in most CESEE countries. Some credit segments already displayed first signs of overheating. This applies in particular to housing loans, which grew swiftly given strong housing demand and ­ever-increasing housing prices. Several CESEE countries (especially EU Member States) have introduced macroprudential measures and/or recommendations to put a brake on this development. Furthermore, countercyclical capital buffers were activated in Bulgaria, the Czech Republic and Slovakia.

Compared to 2008, banking sector risks in CESEE have moderated. For ­instance, the share of foreign currency-denominated loans in total loans to the ­private sector has declined throughout most of the region (see chart 13). This downward trend is particularly pronounced in loans to households, which is ­especially welcome as the sector is usually unhedged against exchange rate changes. At end-2019, the foreign currency share in loans to households was virtually zero in ­Slovakia, Slovenia, the Czech Republic, Hungary and Russia (in Turkey, foreign currency loans to households have been legally banned). Several countries (most prominently Hungary and Croatia) introduced conversion schemes for foreign ­currency loans to households into local currency loans that fueled this downward trend. It needs to be noted, however, that the foreign currency share in corporate loans remains notably higher. Such loans have received more policy attention ­recently. The IMF has intensified its warnings on high levels of corporate debt in emerging markets, and the sharp depreciation of the nominal effective exchange rate of the Turkish lira in 2018 as well as the recent pandemic-­induced ­depreciation of some of the region’s currencies has illustrated potential risks.

The decline in foreign currency loans has contributed to a higher quality of banking sector assets, as has the decline in nonperforming loans (NPLs). A ­comparison of the shares of nonperforming assets in total assets for 2008 and 2019 does not yield a clear-cut regional trend (see chart 14). Compared to 2013, ­however, when the global financial crisis and crisis legacies pushed NPL ratios to 15% or more in several countries, NPL ratios have come down substantially. This positive momentum can be attributed to favorable lending developments as well as to the strong general economic momentum. Furthermore, active portfolio ­cleansing measures – including writing off bad debt, selling NPL portfolios as well as restructuring and forbearance agreements and the transfer of NPLs to bad banks – also positively impacted the stock of nonperforming assets. In Russia, NPLs also decreased somewhat in the past two years after the 2014 recession and the banking sector turbulences in late 2017 (which led to the nationalization of three ­medium-sized credit institutions together accounting for about 7% to 8% of banking assets). Turkey was the only CESEE country to report higher NPLs in the past few years. The increase in Turkish NPLs reflected the financial difficulties ­associated with the 2018 financial turbulences faced by indebted companies – ­particularly those with debts in foreign currency.

Another positive development was observed in CESEE banking sectors’ ­refinancing structure. Over the past years, the refinancing structure has shifted from external liabilities to local deposits. This makes CESEE banks less vulnerable to swings in international sentiment and the possible (and in a worst case synchronized) withdrawal of foreign capital. At end-2019, seven of the ten countries under observation reported a (partly substantial) overhang of private sector deposits ­relative to domestic banking sector claims (see chart 15). Banking sectors in Slovakia and Turkey reported a moderate funding gap (on a notable declining trend in the case of Turkey), while only Russia recorded a persistently high funding gap of around 10% of GDP.

On top of that, the risk-bearing capacity of CESEE banking sectors has ­improved. Capital adequacy ratios have increased throughout the region, in most countries substantially so (see chart 16). Capitalization is not only higher than in 2008, it is also high compared to other European countries. The by far weakest capital base was reported for Russia.

Summing up, risks emanating from the banking sector – a main transmission channel in the global financial crisis – currently seems to be relatively contained. Credit growth is swift (in certain segments probably too swift) but credit is stably funded by local deposits. Asset quality has improved, and banking sector portfolios today are generally less risky than a decade ago. Among the individual countries, Russia and Turkey are exposed to the highest risks. In Russia, domestic claims are notably higher than domestic deposits, capitalization is weak and the NPL ratio is the highest in the region (accompanied by a coverage ratio of only somewhat above 50%, which is the lowest level in CESEE). At the same time, credit growth ­remains high and strongly driven by uncollateralized consumer loans. Turkey was the only country to report negative dynamics in nonperforming assets. While consumers are banned from foreign currency borrowing, the stock of foreign currency-­denominated lending in corporate credit remains stubbornly high. This exposes corporations to exchange rate risks amid large (short-term) negative foreign ­currency positions. Finally, the Turkish banking sector still reports a positive – though declining – funding gap.

Spread of coronavirus: CESEE was affected later and to date less strongly than other regions

The Global Health Security Index (developed by the Johns Hopkins Center for Health Security, the Nuclear Threat Initiative and The Economist Intelligence Unit) provides a rough estimate of the health security capabilities in CESEE. The subcomponents of the index that are related to responding to and mitigating the spread of an epidemic and to the quality, coverage, resilience and robustness of the health sector are especially relevant in the current situation. In both categories, CESEE countries do not score notably worse than Austria or Italy. Some countries even outperform Western European peers, e.g. Slovenia, Poland and Turkey. Rather low capabilities of dealing with pandemics, however, are reported for ­Romania, Slovakia and especially Bulgaria.

While the Global Health Security Index suggests that health-related risks from the coronavirus pandemic are generally in line with other European countries, two factors should be noted: (1) The median age of the population in CESEE is rather high and the age distribution of the population skewed toward older age cohorts (this does not apply to Turkey, though). This implies a rather large high-risk group for serious COVID-19 infections. (2) Furthermore, studies ⁵ suggest that case-­fatality rates are strongly positively related to a country’s share of working-age families living with their parents (multigeneration households). Such living ­arrangements increase intergenerational contacts and enable the virus to spread more quickly to high-risk older strata of the population. Multigeneration households are especially widespread in several CESEE countries (e.g. in Slovenia, ­Bulgaria, Croatia and Poland).

Government responses to the spread of coronavirus: fast, comprehensive and associated with high economic costs

Government measures will, without any doubt, lead to large economic costs. The extent of these costs, however, is still very hard to gauge as no hard data on production and sales for the period in question is available at the time of writing (for our current forecast for GDP growth in CESEE, see the Outlook for selected CESEE countries in this issue). However, some indicators suggest that public life has temporarily come to a standstill and that economic activity has declined ­notably. Chart 19 shows mobility trends published by Google, which are based on visits and the length of stays at different places. For example, mobility observed at transit stations (e.g. public transport hubs such as subway, bus and train stations) declined by more than 50% on average in the CESEE region in mid-April ­compared to early February 2020. The reduction was even more pronounced in the area of retail and recreation (e.g. at restaurants, cafés, shopping centers, theme parks, ­museums, libraries and movie theaters), where mobility declined by 70% on ­average (up to 90% in the case of Turkey). The same is true for international ­mobility. Flight departures from international capital airports had basically come to a standstill by mid-April (see chart 20).

Mirroring these figures, sentiment in CESEE has plummeted (see chart 21). In April 2020, the European Commission’s Economic Sentiment Indicator (ESI) for the CESEE EU Member States dropped by nearly 40 points and declined to its lowest level in history. A similar development was observed in the Purchasing Managers’ Index (PMI) for Russia and for Turkey. Both indexes descended into free fall in April 2020 and reported values of only 31.3 and 33.4 points, respectively, far below the 50-point threshold indicating an economic expansion.

The impact of the coronavirus crisis on financial market indicators was immediate and substantial. A deterioration was observed in all financial market segments and in all countries. The strongest impact was reported for equity prices, which declined by up to 29% against the start of the year in Hungary. Euro-­denominated eurobond spreads also increased strongly in all countries against the backdrop of increasing sovereign credit default risk and a notable depreciation of national currencies against the euro. The countries most affected by the sell-off were Turkey, Russia, Romania and Croatia, not least because of higher macro­financial vulnerabilities (as outlined above).

Governments throughout CESEE have taken extensive measures to alleviate the fallout from the ensuing greater economic crisis (for a more detailed overview see country chapters). Fiscal support measures have included, among others, ­deferring tax and social security contributions for affected enterprises, taking over part of the salary payments to employees and extending paid sick leave. Some countries have increased the pay for key sectors, including the salaries of medical professionals. Governments in many countries have also taken measures to avoid liquidity shortages in the real sector, often jointly with commercial banks and development banks; in some cases, these measures include state guarantees. Moratoria for debt repayments have also widely been recommended and implemented.

CESEE central banks have also been quite active and made use of their full tool kit. Following the onset of the pandemic, key policy rates were cut in the Czech Republic (by 125 basis points to 1%), Poland (100 basis points to 0.5%), Romania (50 basis points to 2%) and Turkey (100 basis points to 9.75%). The Croatian ­central bank also inter­vened on the foreign currency markets to contain ­depreciation pressures. Liquidity provision measures for banks including ­longer-term refinancing operations and additional foreign currency swap auctions were launched in several countries, and some central banks also started buying bonds of their respective governments (e.g. in Croatia, Hungary, Poland, ­Romania). In mid-April, the Croatian and the Bulgarian central bank announced swap lines with the ECB. Easing measures regarding the regulatory frame­work have also been announced. They include, among others, a revision of (planned) countercyclical capital buffer rates (e.g. in Bulgaria and the Czech Republic).

Western Balkans ⁷ : strongly affected by the coronavirus crisis

The coronavirus crisis strongly impacted on the Western Balkan economies since the first lockdown measures were enacted toward mid- or end-March 2020. The far-reaching shutdown of the economies to combat the spreading of the virus have shaken all areas of life in the region with tremendous economic consequences. Overall, the Western Balkan region seems to be more vulnerable to a fast spreading of the virus and its consequences owing to generally poorer health systems and less preparedness to face a pandemic than most EU countries (chart 1).

In most Western Balkan countries, economic growth already lost some momentum in the final quarter of 2019 compared to previous quarters, with strongly diverging growth patterns (chart 2). Growth declined particularly sharply in Albania, from 4.2% in the third to –0.2% in the fourth quarter of 2019, as a devastating earthquake hit the country in November 2019, and in Bosnia and Herzegovina, from 3.1% in the third to 1.6% in fourth quarter of 2019, driven by almost stagnating private consumption. By contrast, growth in Serbia accelerated to 6.2% in the fourth quarter on the back of strong gross fixed capital formation, which lifted full-year growth to above 4%.

Overall, private consumption continued to be an important growth contributor in the second half of 2019. Swift credit expansion, higher wages and remittances and, moreover, positive developments on the labor markets were supportive factors. According to labor force surveys, the Western Balkan economies managed to bring down their unemployment rates (toward the end of 2019) compared to a year earlier. In Kosovo, the unemployment rate declined by almost 4 percentage points compared to one year earlier to (still high) 25.7% at the end of 2019, and by 3.5% in North Macedonia to 17.5%. At end-2019, Serbia had the lowest unemployment rate with below 11%. It should be noted that these positive trends are also owed to a strong brain drain in the region.

The rather volatile patterns of investment are often the result of big public investments (mainly infrastructure or energy projects) throughout the Western Balkans. It is worth noting that, in Albania, investment growth decelerated sharply in the second half of 2019 due to the finalization of large infrastructure projects. Furthermore, political uncertainty and the earthquake left their marks as well. In Serbia, by contrast, investment growth accelerated by more than 20% in the second half of 2019. Here, a huge energy project was key but FDI inflows also contributed positively to investment growth.

Export growth was particularly strong in Albania (record tourist season) and Kosovo (a new ferronickel plant started operation, increasing exports of services). In North Macedonia, export growth turned negative in annual terms in the last quarter of 2019, possibly mirroring lower international demand (the country is relatively strongly integrated in global value chains).

Strong investment activity in several countries resulted in high import growth given a substantial import content of investments. This is particularly true for Kosovo, North Macedonia and Serbia, where the growth contribution of net exports was negative in the second half of 2019. In contrast, in Albania, Bosnia and Herzegovina, and Montenegro, import growth contributed positively to growth toward the end of the year, leading to a positive contribution of net exports in these countries.

Current account deficits (see table 2 in the statistical annex) narrowed in all Western Balkan countries in the second half of 2019 compared to the first half. In Montenegro, for instance, lower imports of machinery needed for infrastructure projects and higher exports related to tourism ⁸ and transport services caused the improvement. Albania reported a record tourist season for the second half of 2019. Furthermore, secondary income (largely workers’ remittances) as a share of GDP remained high and even accelerated in some countries compared to previous periods (Albania, Kosovo). In the second half of 2019, FDI as a share of GDP on average moderated somewhat compared to the first half of 2019 but still covered the lion’s share of the current account deficits.

Inflation rates declined in almost all Western Balkan countries. In Bosnia and Herzegovina, Montenegro and North Macedonia annual inflation fell below 1% in 2019 on average. The situation looked different in Kosovo, where inflation stood at almost 3% in 2019 but moderated over the year. Inflationary pressure was largely the result of 100% tariffs imposed by Kosovo on products from Bosnia and Herzegovina as well as Serbia. In early April, Kosovo lifted the tariffs. Partly as a result of the lower oil price, inflation dropped further in March 2020 in all countries. Only in Albania inflation accelerated strongly to 2.1% year on year, driven by higher food prices, but still remained below the inflation target of 3% set by the Bank of Albania. So far, the exchange rate regimes seem to be coping relatively well with the coronavirus crisis and have remained rather stable. Albania and Serbia are the only two countries among the Western Balkans with a flexible ­exchange rate regime. In Albania, the lek temporarily lost some 6% against the euro at the end of March but has largely recovered since then, while the Serbian dinar traded stably.

In response to the pandemic, all countries implemented swift and harsh measures to contain the spread of coronavirus, closing shops, businesses and borders and imposing curfews. All central banks have been very active, too, in addressing the crisis; several interest rate cuts were implemented in March and early April 2020. The National Bank of Serbia cut its key policy rate in two steps by a total of 0.75 percentage points to 1.50% and adopted several measures to supply the domestic sector with additional dinar and foreign currency liquidity. The National Bank of the Republic of North Macedonia cut the policy rate by 0.25 percentage points to 1.75% and the Bank of Albania by 0.5 percentage points to 0.5%. Furthermore, most central banks have decided to support households and businesses with moratoria on debt payments in case these are facing difficulties in repayment; they have also implemented measures to support the granting of loans.

The fiscal positions are rather heterogenous across the region (see table 4 in the statistical annex) but overall the fiscal leeway to deal with the coronavirus-induced shock is relatively limited. In 2019, Montenegro reported the biggest fiscal shortfall with –2.6% of GDP (the same as in 2018); in North Macedonia, the deficit widened, reaching –2.5% (2018: –1.8%). Bosnia and Herzegovina as well as Kosovo achieved fiscal surpluses in 2019. In Albania, the destructive earthquake in November 2019, which resulted in additional expenditure pressure, left the deficit at –1.7%. In 2019, Montenegro was the country with the highest debt-to-GDP level (77.8%), followed by Albania with 66.6%.

The coronavirus crisis is now derailing fiscal plans, causing significant fiscal pressure in 2020. A large fiscal expansion is needed on account of higher health, social and economic support spending as well as revenue shortfalls due to an expected massive economic ­contraction. So far, Serbia has released the most substantial measures to support the economy in the Western Balkans.

On a positive note, at the end of March 2020, the EU finally gave green light for opening accession negotiations with Albania and North Macedonia. Albania has been an EU candidate since 2014, North Macedonia already since 2005. Together with Montenegro (where ­accession negotiations started in 2012) and Serbia (since 2013), four countries are currently negotiating with the EU to become Member States. Bosnia and Herzegovina and Kosovo are classified as potential EU candidates. Regarding relations with the IMF, the Extended Fund Facility (EFF) of the IMF approved for Bosnia and Herzegovina in 2016 has been off track, i.a. because the country failed to form a new government after general elections in autumn 2018. In December 2019, a new government was put in place, which provided the basis for a resumption of the IMF arrangement. In December 2019, the IMF completed the third review under the Policy Coordination Instrument (PCI) with Serbia. Accordingly, the economic reform program is on track. In light of the coronavirus crisis, international organizations are playing an important role in supporting the region. The IMF has activated emergency support under its Rapid ­Financing Instrument (RFI) for Albania, Bosnia and Herzegovina, Kosovo and North ­Macedonia. Also, the EU, with a macrofinancial assistance (MFA) package and other measures, as well as other institutions or countries have stepped in to support the region.

2 Slovakia: coronavirus crisis has been a baptism of fire for the new government

Buoyed by still favorable economic conditions, the labor market remained ­robust, ­despite significant regional disparities, with unemployment figures ­improving marginally in the second half of 2019. At the same time, slowing ­employment and wage growth gradually relieved labor market tensions. As rising labor and global commodity prices had been passed through to food and energy prices, annual inflation edged up to average 3.1% in the review period, although annual price increases moderated to 2.4% in March 2020. Due to the slowdown in economic growth, the general government deficit did not continue its downward trend and remained broadly unchanged at 1.3% of GDP in 2019 compared to the year before. Nonetheless, due to the increase in the denominator, public debt relative to GDP came down by more than 1 percentage point to about 48% of GDP in 2019.

After a good start to the year, with strengthening industrial production, sales, construction and consumer confidence, the first COVID-19 patient was confirmed in Slovakia on March 6, 2020. Amid a global pandemic, Slovakia experienced a historical change in leadership. A new four-party coalition government led by Prime Minister Igor Matovič, who was elected largely due to his promise to crack down on corruption, was sworn in on March 21, 2020. With the words “Let’s go to battle,” Prime Minister Matovič and his government took over from the former administration led by the Social Democrats. Strict containment measures have since been introduced, ranging from a shutdown of nonessential stores and almost all service sector activity to the closure of borders and schools and a ban on the free movement of people (barring some exceptions). These unparalleled restrictions – compounded by similar measures taken by Slovakia’s trading partners – with their disruptive consequences for global supply chains and international trade will have a massive impact on Slovakia’s economy. Moreover, they provide a genuine stress test for the young ruling coalition. The heavy economic toll of the coronavirus crisis will be aggravated by the fact that all carmakers in Slovakia suspended production for a ­couple of weeks not only to contain the spread of coronavirus but also to adjust to lower demand and obstructions in related supply chains. This represents a major setback for an economy that has the largest car production rates per capita worldwide and whose automotive industry accounts for about 12% of GDP and 44% of industrial production. While the change of government delayed the introduction of economic anti-crisis packages, the government has meanwhile approved ­measures that include, inter alia, benefits for self-employed individuals and wage subsidies for affected employees. Moreover, the measures allow for the deferred payment of employer levies as well as bank guarantees and moratoria on the ­repayment of loans and mortgages.

3 Slovenia: no grace period for the new government

In late January 2020, the five-party minority government resigned. The new four-party coalition government had initially intended to go ahead with reforms to Slovenia’s pension, healthcare and long-term care systems as well as its national defense system, to cut red tape and to improve the country’s infrastructure. However, these plans were disrupted by the coronavirus pandemic.

Alongside restrictive measures to slow the spread of coronavirus, various ­support measures have been introduced to mitigate the economic impact of the pandemic. As a case in point, the state is providing wage compensation for short-time working schemes and temporary layoffs and has taken over, on a temporary basis, ­pension and health insurance payments as well as sick pay. Furthermore, ­financial support is being provided to parents who have to take care of their ­children while schools are closed, while pensioners with the lowest pensions, ­university students and families with at least three children have received a one-off allowance. Household electricity prices have been temporarily cut by 20%. Public sector employees working in sectors that are key to overcoming the pandemic will be rewarded with a bonus; in addition, the government has invited private companies to follow suit. Self-employed individuals who had come under pressure were given support in the form of emergency assistance, exemptions from pension and healthcare contributions and postponements of income tax prepayments. The government has also introduced tax deferrals for up to two years and the option of paying taxes in 24 installments for troubled businesses, while additional state guarantees and credit lines have been extended.

With the general government budget posting a surplus in 2019 (0.5% of GDP), the government now has some room for maneuver. The expected economic contraction, however, is likely to aggravate the situation. The Bank of Slovenia (BS) estimates that the drop in GDP will range between 6% and 16% in 2020, which would push state debt up to between 70% and 78% of GDP in 2020 (from 66.1% in 2019).

Borrowers who had become unable to settle their loan liabilities due to the ­negative effects of the pandemic were given the option to defer debt servicing for a maximum of 12 months at no additional costs. As Slovenia is a member of the euro area, the country’s banks will also benefit from the ECB’s additional asset ­purchases. In addition, since the BS has extended the supervisory measures taken by the ECB for banks under its direct supervision to less significant institutions under national supervision, all Slovene banks will benefit from further flexibility in the prudential treatment of loans backed by state guarantees. Also, the BS ­decided to defer certain deadlines for less important supervisory activities and granted banks capital relief.

4 Bulgaria: entering the coronavirus crisis with an already slowing economy

Bulgarian authorities declared a state of emergency on March 13, 2020, which is scheduled to last for at least two months and includes the closure of schools, ­universities, shopping centers, cinemas, restaurants as well as the suspension of all mass public events. In contrast to other countries, though, several smaller ­businesses already reopened after two weeks of closure. Construction sites, supermarkets, food markets, pharmacies, banks and gas stations had been exempted from the lockdown. To contain the spread of the coronavirus, entry bans were issued, domestic travel was restricted and curfews or quarantine measures were introduced for affected areas. The immediate economic impact has already been substantial: The number of passengers passing through Sofia Airport in March 2020 halved compared to a year before. Fuel consumption dropped by between 40% (in major cities) and 80% (in small towns) in the second half of March. Car parts manufacturing plants largely suspended their production. According to a poll by Gallup International in early April, about one-third of respondents indicated that their income had dropped, and another 25%, mostly elderly people, said that they expected this to happen. According to the Labor Minister, by early April, about 40,000 people had registered as unemployed since the start of the coronavirus crisis.

In early April, the planned general government deficit for 2020 was raised to 2.9% of GDP and the annual borrowing ceiling to BGN 10 billion (compared to an originally planned balanced budget with annual borrowing capped at BGN 2.2 billion). The extra spending will cover not only increased unemployment spending, support for pensioners and increased salaries for medical staff but also state aid for affected businesses by taking over 60% of gross salaries of employees facing layoffs and obliging employers to retain their staff and pay the remaining 40%. The Bulgarian National Bank (BNB) implemented a package of measures worth BGN 9.3 billion (nearly 8% of 2019 GDP figures) to strengthen banks’ capital and liquidity. The package requires all banks to retain their year-end 2019 profits and bans the redistribution of profits accumulated during previous years. The planned increases in the countercyclical capital buffer, by contrast, were cancelled. Moreover, the BNB paved the way for a temporary moratorium on debt repayments for crisis-affected borrowers for up to six months until the end of the year. As regards the ERM II, the BNB stated that the crisis will most likely delay Bulgaria’s ERM II entry until 2021, while the government plans to apply for ERM II membership by the end of April this year. Besides the BNB, the state-owned Bulgarian Development Bank also implemented several liquidity-supporting measures. Its capital was raised by the government by BGN 700 million (about 0.6% of 2019 GDP levels) and will be used to issue portfolio guarantees to commercial banks extending loans to crisis-­affected SMEs and to issue interest-free consumer loans to self-employed individuals and employees on unpaid leave.

5 Croatia: entering the downturn with high sovereign debt and strong reliance on tourism

Average inflation in 2019 was low (0.8% year on year), with energy and food price movements (VAT changes) contributing to some volatility in headline inflation. The Croatian National Bank (HNB) continued its accommodative stance and further built up international reserves (EUR 18.5 billion at end-2019, roughly 9 months of imports) through its foreign currency market interventions to counter appreciation pressures on the kuna. The Croatian banking sector is one of the most profitable and best capitalized in the region, with a return on assets of 1.4% and a tier 1 capital ratio of 22.4% at end-2019 – both slightly higher than a year before. The NPL ratio declined markedly to 5.5%.

One of the main imbalances remains the high sovereign debt level, which came to 71.2% of GDP at the end of 2019, 3.5 percentage points lower than a year ago. The Croatian government reported a very mild budget surplus in 2019. Croatia’s external debt declined to the still high level of 76% of GDP at end-2019, down 7 percentage points from the previous year and driven to a large extent by the govern­ment sector. Corporate indebtedness is elevated compared to peer countries (roughly 68% of GDP). Croatia’s economy relies heavily on tourism, which indirectly contributes some 20% to 25% to employment and gross value added. Over the past years, this has supported growth and improved the current account surplus, which increased to 2.9% of GDP in 2019 on the back of a strong tourist season. However, tourism is likely to be among the sectors most affected by the current crisis, potentially exposing the Croatian economy to a sharp downturn.

The Croatian government and the HNB have implemented strong measures to support the country’s economy and exchange rate stability. The Croatian government has approved two economic support packages worth an estimated 9% to 10% of GDP (roughly half of which relates to debt moratoria or loan guarantees). The ­financing has not yet been secured for all announced support measures. On April 3, 2020, the government reported having received 65,000 requests for tax ­deferrals from businesses and 70,000 applications from employers for the government’s net minimum wage support program aimed at saving 420,000 jobs (approximately 25% of Croatia’s labor force). During March 2020, the HNB intervened several times in foreign currency markets, injecting a total of EUR 2.2 billion into the markets. In addition, it launched a high-volume five-year longer-term refinancing operation, restarted regular weekly repo auctions with full allotment, purchased government bonds on the secondary market and eased regulations for banks.

So far, Croatia has not announced any deviations from its timeline for adopting the euro, according to which Croatia will join the ERM II in the second half of 2020.

6 Czech Republic: economy is teetering on the edge of the coronavirus abyss

Owing to a lower surplus of the trade and services balance and to a strong outflow of dividends in the primary income balance, the current account turned negative in the second half of 2019. The general government budget surplus declined noticeably in 2019 as a result of the economic slowdown, rising wages in the public sector and higher social transfers. While the labor market remained tight, the sustained rise in employment and decline in unemployment petered out in late 2019. Nonetheless, inflation was fueled by buoyant wage growth and consumer demand ­coupled with a weaker koruna and a surprisingly sharp rise in administered and food prices. Inflation thus averaged 2.8% in the second half of 2019 and gradually ­increased further to 3.7% in February 2020. After having missed its inflation target repeatedly (2% ± 1 percentage point), the Czech National Bank (CNB) raised the key policy rate by 25 basis points to 2.25% in early February 2020.

The Czech government reacted forcefully and quickly to the outbreak of ­coronavirus, declaring a state of emergency on March 12, 2020. By mid-March, borders, restaurants and most shops had been closed and a nation-wide curfew had been issued. Hence, the coronavirus-induced damages were already reflected in the most recent business and consumer sentiment indicators. Google data suggest that demand in the retail and recreation sectors dropped by more than 60% in the second half of March. Worse still, an agonizing 90% of the entire automotive ­industry – the backbone of the Czech economy, which accounts for more than 8% of GDP and for one-quarter of industrial production and exports – stopped ­production for at least a month.

Both fiscal and monetary policy have reacted vigorously to the unfolding crisis. Given the country’s rather favorable fiscal position, the government has, to date, earmarked some 18% of GDP to deal with the impact of the coronavirus ­pandemic, the third-largest amount in Europe according to the IMF. The measures adopted by the government include income support for affected employees, lump-sum ­payments to self-employed individuals as well as credit lines and guarantees for businesses. In addition, some taxes have been temporarily suspended, penalties have been waived and a six-month moratorium on some bank loans has been announced. The CNB lowered the policy rate twice by 50 and 75 basis points, respectively, to 1% in March 2020. It also increased the frequency of repo operations from one to three times a week. On the regulatory side, the CNB reduced the countercyclical capital buffer rate from 1.75% to 1%, and relaxed regulatory limits for new mortgages. In preparation for quantitative easing, the Act on the CNB has been amended to ease existing restrictions on open market transactions.

7 Hungary: coronavirus pandemic brings about cyclical turnaround

Since the outbreak of coronavirus in Hungary, the government has ­announced several direct measures to fight the pandemic (e.g. shop closures and increased funding for the healthcare sector) and has presented recovery packages to ease the negative impact on the economy. These included, in a first step, a temporary suspension of tax payments for small businesses, a substantial reduction of social security payments for the most heavily hit sectors, a temporary ban on ­evictions, property seizures and the enforcement of tax debt collection as well as an extension of expiring entitlements to childcare benefits until the end of the state of ­emergency. In a second step, the 2020 budget deficit was revised up from 1% to 2.7% of GDP following the creation of three funds (partly financed by budget ­restructuring and partly by additional taxes on the retail and banking sectors). Furthermore, the ­recovery packages included government cofinancing of wages of employees on short-time working, investment support, state guarantees and interest subsidies for corporate loans, the gradual re-introduction of an extra 13th month pension payment as well as bonusses and wage hikes in the healthcare sector.

The Hungarian parliament imposed a debt service moratorium and an interest rate cap for consumption loans, both until end-2020. To ease the adverse effects of these measures on banks, support the restart of the economy and counter increased depreciation pressure on the forint, the National Bank of Hungary (MNB) has taken comprehensive action. To mitigate liquidity tensions and support lending, it expanded its lending facilities, granted banks a moratorium on servicing loans ­under its F4G schemes and exempted banks from reserve requirements. It also initiated a new “Funding for Growth Go!” (F4G Go!) scheme with lighter ­eligibility criteria and expanded its “Bond funding for Growth Scheme” (BGS), while paying a preferential 4% interest rate on its deposit facility to sterilize the additional ­liquidity. Moreover, the MNB launched a government bond purchase program and restarted its mortgage bond purchase program. In response to mounting ­depreciation pressure on the forint in the second half of March 2020, the MNB introduced one-week deposit auctions at the base rate to withdraw short-term ­liquidity. At the beginning of April, it hiked its overnight and one-week deposit rates to 1.85% and allowed the one-week deposit rate to be set in a flexible way between –0.05% and 1.85% at future auctions. To ease the administrative and regulatory burden on the financial sector, the MNB eased capital requirements, suspended some fines and deferred certain deadlines. At the same time, it took measures to bolster banking sector stability, inter alia by tightening some ­prudential ratios, and instructed banks not to decide on and pay out dividends until end-­September 2020.

8 Poland: entering the coronavirus crisis in a comparatively robust economic position

On March 4, 2020, the MPC, pursuing an inflation target of 2.5% ±1 percentage point (CPI), considered inflation above the upper bound due to supply-side and regulatory factors beyond its direct influence and concluded that inflation would remain close to the target over the monetary policy transmission horizon. On March 17, 2020, in view of the coronavirus impact, the MPC agreed, through ­intersessional decision, to cut the rate from 1.5% to 1%, to decrease the required reserve ratio from 3.5% to 0.5% and to increase the remuneration of the required reserves from 0.5% to 1%. Moreover, it decided to offer bill discount credit aimed at refinancing new loans granted by banks and to start outright government bond purchases on the secondary market to strengthen the monetary transmission mechanism. Having bought a volume of about 0.8% of GDP by April 8, 2020, the MPC cut the policy rate to 0.5%. In parallel, commercial banks offered moratoria of up to six months to households and SMEs. Regarding fiscal policy, in fall 2019, the European Commission had expected the headline deficit to narrow to 0.2% of GDP in 2020 after widening to 0.7% of GDP in 2019, and the structural deficit to narrow slightly to 1.9% after widening to 2.2% in 2019. General government gross debt declined to 47.4% of GDP in 2019. On March 21, 2020, in view of the implications of the coronavirus crisis, the government exempted self-employed individuals and micro-companies from social security payments (for three months if revenues had dropped by more than 50% compared to February 2020) and provided support for farmers. On March 31, 2020, an economic support package worth EUR 46 billion or 9% of GDP was adopted, providing, inter alia, support for companies, the financial and health sectors as well as public investment.

9 Romania: entering the coronavirus crisis with twin deficits

Procyclical fiscal and income policies have fueled economic growth in recent years. As a consequence, the general government budget deficit rose to 4.3% of GDP in 2019 and thus exceeded the 3% limit set out in the EU’s stability and growth pact. On April 3, 2020, the Council of the European Union concluded that an excessive deficit existed in Romania. A few days later, the European ­Commission emphasized in a letter to the Romanian Finance Minister that it would fully take into account the economic and fiscal impact of the coronavirus outbreak in its ­assessment under the excessive deficit procedure. As the current account deficit widened in 2019, Romania entered the international coronavirus crisis with twin deficits.

To contain the spread of coronavirus, the Romanian authorities took various measures in early March that have since been tightened further and complemented. In addition to a series of initial measures (e.g. cancellations of events and flights to Italy and Spain and closures of schools, universities and restaurants), restrictions on movement were put in place on March 25, 2020. Since then, people have not been allowed to leave their homes with a few exceptions (e.g. to go to work or buy food).

To mitigate the impact of the coronavirus pandemic and related containment measures, the government decided to cover the wages of employees working for companies strongly affected by the pandemic (up to a limit of 75% of individual gross wages and capped at 75% of the average gross wage in the economy). In ­addition, to avoid liquidity shortages in the real sector, the state will provide ­guarantees for loans to companies (especially SMEs) and permit borrowers to ­postpone their loan repayments by nine months. Yet, the scope of the moratorium remains unclear, as the latest law on this subject has been challenged before the constitutional court. Finally, individuals and companies will be allowed to defer tax payments.

The National Bank of Romania (NBR) reacted with a policy package as well. It cut its key policy rate by 50 basis points to 2% and decided to provide liquidity to credit institutions via repo transactions and to purchase leu-denominated ­government securities on the secondary market. The NBR’s decisions came after four failed treasury bond auctions and market reports of foreign exchange ­interventions by the NBR to support the leu. Furthermore, the NBR stressed that it stands ready to cut the minimum reserve requirement ratios on leu- and foreign currency-­denominated liabilities of credit institutions, depending on how the ­situation evolves.

10 Turkey: entering the coronavirus crisis in a very weak economic position

On March 17, 2020, the TCMB reduced the policy rate to 9.75% and increased liquidity provision to banks by introducing unlimited access through intraday and overnight standing facilities, longer repo auctions and targeted additional liquidity facilities at 100-150 basis points below policy rate linked to credit provision to the corporate sector. Moreover, the TCMB lowered foreign exchange reserve requirements by 500 basis points for banks that meet certain credit growth conditions and thus released EUR 4.5 billion (0.7% of GDP). It also granted export companies a three-month moratorium on the repayment of rediscount credits until end-June 2020, thus providing about EUR 7 billion. On March 31, 2020, the TCMB announced that it will increase outright purchases of domestic government debt securities and extend Turkish lira-denominated rediscount credits to export companies at 150 basis points below policy rate by TRY 60 billion (1.4% of GDP). Moreover, the government asked all financial institutions not to cancel existing loans and to ease the requirements for new loans. Furthermore, it doubled the state credit guarantee fund. Regarding fiscal policy, in fall 2019, the European Commission had expected the headline deficit to widen to 3.5% in 2020. In view of the coronavirus impact, a first economic package worth TRY 100 billion (2.3% of GDP) was set up, which includes deferrals of taxes and social insurance payments by six months for all companies operating in particularly affected industries, several benefits for elderly people and new funds for poor families.

11 Russia: continued sluggish growth in 2019, coronavirus crisis triggers plunge in oil price, but fiscal and external buffers available

Given the decline in the oil price (–9% compared to the annual average price of Urals grade crude in 2018) and continued foreign exchange purchases of the Russian central bank (CBR) under the fiscal rule, the exchange rate of the ruble against the U.S. dollar weakened slightly (–3.5% from the annual average). The CBR’s tight monetary stance (elevated key rate), overall still sluggish demand and declining inflationary expectations helped cushion the temporary rise in inflation linked to housing and communal tariff adjustments as well as the VAT increase in January 2019. Accordingly, inflation decreased to 2.5% in March 2020 year on year (1.5 percentage points below the CBR’s target of 4%). Meanwhile, the corona­virus pandemic has swiftly changed the trajectory of the economy, triggering an oil price plunge in March-April 2020 and bringing about nation-wide lockdowns ­stifling business from late March onward. The combined effect has pushed the Russian economy into recession. The CBR has reacted by switching to an accommodative monetary stance, cutting its key rate by half a percentage point to 5.5% in April and announcing its readiness for further cuts.

The VAT increase, improved tax administration, the pension reform (adjustment of the retirement age) and restraint in public spending secured a budget surplus of 1.9% of GDP in 2019. The declining oil price in 2020, however, contributed to eliminating the surplus. Meanwhile, thanks to the transfer of oil and gas earnings of 2019, Russia’s budgetary stabilization fund, the National Wealth Fund, reached a record level of EUR 150 billion at end-March 2020 (most of the fund is part of the country’s international reserves). In April 2020, fiscal anti-crisis measures amounting to about 2.8% of GDP (mostly tax holidays and loan guarantees) were decided for 2020. The lower oil price contributed to the decline in the current ­account surplus to 3.8% of GDP in 2019. Private capital outflows decreased to 1.6% of GDP in 2019 (from 3.8% in 2018). Largely on account of nonresidents purchasing Russian obligations, the country’s foreign debt expanded to EUR 430 billion at end-2019 (+8% year on year). However, Russia’s foreign debt has been clearly outgrown by its international reserves (including gold, which the ­authorities have stocked up substantially in recent months), which stood at EUR 516 billion in mid-April 2020 (+19% over the last 12 months or 34% of GDP).

Notwithstanding Russia’s sluggish economic growth and its NPL ratio ­stagnating at a relatively high level (17.1% at end-2019), retail lending (as opposed to corporate lending) continued to expand swiftly (+15% in late 2019 in real terms and exchange rate adjusted). That said, the expansion of retail lending – notably that of unsecured consumer credit – has passed its peak thanks to the CBR’s prudential tightening measures and, in part, related easing of credit demand on the back of unstable household income dynamics. Moreover, the coronavirus crisis is expected to strongly dampen any further growth of banking activities.

Outlook for selected CESEE countries

Economic activity in the CESEE-6 region will take a deep dive in 2020 and then recover hesitantly, Russian economy set to contract in 2020 ^{9 , 10}

The coronavirus pandemic and falling oil prices have seriously eroded prospects for the Russian economy. ¹² Thus, we expect Russia’s growth to turn negative this year, reaching –3%. Growth is expected to revive next year as the outlook for the global economy recovers and Russia’s public sector spending rises further. The risk of a weaker than expected outcome is significant due to the uncertainty in commodity markets and effects from the coronavirus pandemic.

1 CESEE-6: coronavirus pandemic has fueled uncertainty and will lead to a broad-based decline in economic activity

Even before the outbreak of the pandemic, a slowdown in economic momentum for the CESEE region was evident, given the U.S.-China trade conflict, shrinking world trade amid weaker global growth in general, and a sluggish performance in Germany. In the fourth quarter of 2019, average quarter-on-quarter growth in the CESEE-6 countries stood at 0.7%, the lowest level in three years. Nevertheless, at 3.9%, average growth for the whole year 2019 came in only moderately lower than in 2018 (4.4%).

Since the beginning of the year, and even more so in recent weeks and days, a completely new situation has been rapidly unfolding. At the cutoff date for this forecast, far-reaching economic restrictions have been imposed across CESEE and the duration, coverage and extent of these restrictions remains unclear. For several sectors of the economy, activity has been brought close to a standstill and a general economic recession is becoming inevitable. The question how deep the recession is going to turn out, however, is surrounded by an extraordinarily high degree of uncertainty.

First, there is still a lack of hard data for February and March that could provide facts on the impact of the coronavirus pandemic on CESEE economies and that could inform projections for the following quarters. One of the few available ­figures is the latest release of the European Commission’s Economic Sentiment Indicator (ESI). The March reading of the indicator for the CESEE-6 countries came in at an average of 97.3 points, 3.6 points lower than in the previous months. While this represents the lowest value since early 2014, the drop in the euro area’s ESI reading was even more pronounced (–8.9 points to 94.5 points). The comparatively low decline in the CESEE-6 countries could be related to the timing of the fieldwork for the survey (between February 26 and March 23); it cannot be ruled out that most survey responses were collected before strict containment measures were enacted.

Second, the duration of shutdown measures (and consequently the depth and duration of recession) is wide open and depends on how well people will comply with hygiene and social distancing requirements and on the subsequent development of coronavirus infections.

Third, national authorities have entered partly uncharted territory with the policy measures taken to contain the economic fallout from the crisis. The current crisis is unique in that it stems from a politically imposed, almost complete reduction in economic activity for public health reasons, resulting in a drastic shortage of supply and a subsequent loss of demand. Against this backdrop, previous crises can only partly serve as a blueprint for finding an adequate policy mix and gauging the current situation’s implications for near- and medium-term economic developments.

Fourth, individual sectors of the economy will be affected to a very different degree. While sectors like tourism, arts, cultural, recreational and sports activities, as well as large parts of wholesale and retail trade are literally lying idle, other ­sectors are running at full speed (e.g. health care, trade of daily consumer goods). The composition of the slump in economic activity will therefore differ markedly from previous crises. We have reason to expect that private consumption will ­collapse to a much greater extent than in previous recessions and – on the supply side – the service sector will not be able to act as a stabilizing element.

Fifth, given the very nature of this crisis, a rather swift rebound in economic activity after the lifting of containment measures cannot be ruled out. As most CESEE economies entered the current downturn in generally solid economic shape without any major macroeconomic imbalances, impediments to kick-starting ­economic activity should, in principle, be low. However, a swift rebound crucially depends on a number of factors: how long the shutdown measures will last (the longer they last, the more likely they will have an impact on the growth potential), whether disruptions of the global value chains are of a temporary or permanent nature and how the international environment will evolve. The major regions of the world economy are sequentially entering the shutdown phase and the question arises whether the lifting of related measures will occur simultaneously or in a more staggered approach.

Given the lack of data available to capture the impact of the coronavirus ­pandemic, we assume a moderately negative shock to economic activity in the first quarter of 2020, a strongly negative shock in the second quarter of 2020 and, again, a ­moderately negative shock in the third quarter of 2020. Given positive base effects and catch-up effects buttressed by massive demand-side fiscal and ­monetary ­support, we have penciled in a moderately positive shock for 2021.

Against this backdrop, we project real GDP growth to decline to an average
of –4.5% in 2020. This is a hard brake on economic activity compared to 2019 (+3.9%). The projected recession will be even deeper than in the year 2009 at the height of the global financial crisis (–1.8%), mostly because the Polish economy will no longer act as a stabilizing anchor for the region. On the country level, we expect output to decline, with rates ranging from –3.5% in Hungary to – 7.6% in Croatia. The response to the shock will strongly depend on idiosyncratic factors such as the availability of fiscal space and the specific structure of the economy.

In general, we are currently witnessing an unprecedented decline in growth projections for the CESEE region as is shown by a comparison of different forecast vintages. This is illustrated in chart 1, which shows projections for the year 2009 at different points in time compared to projections for the current year.

For the year 2021, we see a muted recovery to an average growth rate of 3.0%, mainly on the back of positive base effects and fiscal and monetary stimuli for the region. Furthermore, we penciled in a somewhat stronger rebound of investment spending as the financing periods of the EU multiannual financial frameworks for 2014–2020 and 2021–2027 overlap. However, the level of economic activity will remain notably below the baseline of our previous forecast, not least given ­second-round effects of the recession (e.g. higher unemployment, weaker wage growth and a deterioration in consumer confidence). Real GDP growth will return to positive levels in all CESEE-6 countries in 2021 and range from 2.7% in Romania to 4.3% in Bulgaria.

Economic dynamics in 2022 will be very much like those expected in 2021, with an average real GDP growth rate of 3.3% and a broadly similar range of ­outcomes at the country level as in the previous year.

The downturn in 2020 will be broadly based and encompass all components of GDP except for public consumption. The largest negative growth contributions will come from private consumption, but gross fixed capital formation will have a largely equal negative impact on growth in Romania and Poland. In line with weaker projections for euro area import growth, CESEE-6 export growth will also substantially lose steam in 2020. Given the high import content of exports and fading domestic demand, however, we also project a strong decline in import growth. This translates into a roughly neutral growth contribution of the external sector.

Comparing these trends with the experience of the 2009 recession, we expect a broadly similar decline of investments but a much larger contraction of private consumption in the current year than back then. Furthermore, the external sector will not act as a stabilizing element as it did in 2009, as the collapse of imports will not be more pronounced than the collapse of exports according to our projections. (Import growth in the global financial crisis had come down from very high and partly unsustainable levels.)

The recovery in 2021 and 2022 will be just as broadly based as the downturn in 2020, with all components of domestic demand bolstering GDP growth. The growth contribution of net exports will remain about neutral, as both, export and import growth are set to accelerate. Compared to 2009, we project a stronger ­rebound of domestic demand on the back of comprehensive monetary and fiscal measures and a larger base effect.

As outlined above, the current situation is surrounded by an unusually high degree of uncertainty. Therefore, the risks to our CESEE-6 forecast are high and crucially depend on the accuracy of the assumed shocks our projections are based on.

The most important risk relates to the duration of the shutdown measures. Far-reaching restrictions to economic activity beyond the second quarter of 2020 represent the most important downside risk to our forecast. A quicker lifting of restrictions than assumed in our scenario represents an upside risk to our forecast. Furthermore, the possibility of repeated waves of infections (and shutdown ­measures) until the development of a vaccine also is a downward risk that could alter the outlook for 2021 and 2022.

Certainly, economic growth of the world economy, in general, or the euro area, in particular, could turn out higher (or lower) than assumed in our baseline scenario and would thus translate into higher (or lower) growth prospects for the CESEE-6 countries. A better than expected coordination of the international ­policy response to the approaching recession represents an upward risk to our ­projections. In our overall risk assessment, however, we conclude that the risks to external demand are tilted to the downside. The U.S.A. – the largest economy of the world and an important component of final demand for CESEE products – are currently developing into a hot spot of the coronavirus pandemic and infection is spreading at alarming speed. The U.S.A. have so far been rather reluctant to fight the spread of the virus, but the introduction of more full-fledged (European style) measures is becoming more likely. At the European level, there are still many ­uncertainties regarding the future relationship between the U.K. and the EU.

The coronavirus crisis could potentially accelerate the process of de-globalization that we have been witnessing in recent years and thus impact negatively on supply as well. The closing of European (and international) borders, the ensuing inter­ruption of the free flow of goods and people and the consequent scarcity of supplies that are vital to our economies and health-care systems (the most prominent ­example being protective masks) could lead to a reconsideration of the current ­international division of labor. A shift or a complete dismantling of global production chains would be especially harmful for CESEE, as the region is tightly interwoven in such structures. However, in the medium term, the region could also benefit from a reallocation of production capacities from more remote destinations to ­Europe and closer to European consumer markets given the still existing wage ­advantage vis-à-vis Western Europe.

The CESEE-6 countries’ ability to manage a quick economic recovery after the acute phase of the crisis might be impaired by a permanent loss of productive ­capacity, following waves of bankruptcies and surging unemployment. A shift in demand patterns and/or a substantial sectoral restructuring would render previous investment decisions and qualifications obsolete and thus cause notable hysteresis effects. While policy measures are being introduced to avoid such a scenario, it still represents a downside risk especially for the years 2021 and 2022.

Governments throughout CESEE are trying to shield companies, workers and households from the most severe consequences of the coronavirus-induced ­economic slump by lending fiscal support, which will lead to higher public debt. A drastic tightening of global financial conditions and/or the deterioration of ­sentiment could impinge on (re)financing this debt. The same is true for other ­ ­European countries, especially those with already high debt levels. Should the ­solvency of a core member of the euro area be at risk, a general loss of confidence in the single currency could not only have detrimental effects on the euro area but also on the CESEE region. This represents a negative tail risk.

2 Projections for Bulgaria, Croatia, the Czech Republic, Hungary, Poland and Romania

The projected V-shaped impact of this once-in-a-lifetime shock is obviously subject to the country’s resilience and crisis mitigation capacities. The largest short-run risks are related to the poor quality of the health-care system due to ­underinvestment and emigration of medical personnel in the past years. According to the OeNB’s Euro Survey (2018 wave), 80% of respondents in Bulgaria are ­dissatisfied or very dissatisfied with the public delivery of health-care services – one of the highest rates in the region. As regards the scope for countercyclical fiscal policy measures, Bulgaria’s government is in a comfortable position, having ­produced budgetary surpluses in recent years that are partly stored in the Fiscal Reserve Account, which can now be tapped. Bulgaria is also by far less indebted than other EU countries, with a gross public debt-to-GDP ratio of only slightly above 20% in 2019. However, an extensive use of discretionary fiscal stimulus measures is de facto limited by the currency board arrangement in place. The ­latter remains comfortably backed by a comparatively large share of gross official reserves (excluding gold) – nearly 40% of GDP at the end of 2019. As regards the scope for countercyclical financial sector policies, the Bulgarian banking sector has overall been well capitalized and profitable, and additional leeway has been created by the Bulgarian National Bank (BNB) by requiring full capitalization of the ­banking system’s profit and by canceling the planned increases of the countercyclical buffer in 2020 and 2021. Moreover, several liquidity-enhancing measures have been implemented by both the BNB and the state-owned Bulgarian Development Bank, most notably a temporary moratorium that suspends or postpones payments on bank loans, portfolio guarantees on bank loans and interest-free consumer loans for citizens on unpaid leave.

Our first point estimate for Croatian GDP growth foresees a sharp contraction of 7.6% in 2020, largely based on the assumption of severe tourism-related revenue losses this year. Tourist revenues in the second quarter of 2020 will likely be ­completely lost due to coronavirus containment measures. It seems plausible to assume that border restrictions will be among the measures that will be lifted comparatively late and that the uncertainty regarding a renewal of measures, ­combined with lost holiday leave entitlements and income across Europe, will also have severe negative effects on the main tourist season in the third quarter. The tourism sector has an indirect share in total employment and in gross value-added of around 20% to 25%.

The current forecast projects a drop in private consumption of 12%. We assume that, in the second quarter of 2020, domestic consumption will experience a much sharper drop due to the closure of shops, restaurants, etc., but that these domestic containment measures will be slowly lifted in the course of the same quarter. The Croatian government has passed an economic package worth 7.5% of GDP to ­support businesses and households. Whether this package will prove sufficient to shield the private sector from major permanent income and job losses that would weaken consumption even further will depend on how soon domestic businesses and trade relations can return to some normality. The weak tourist season will, however, dampen the recovery of consumption in the second half of 2020. Public consumption is expected to grow by 6% year on year in 2020, largely on the back of higher health-care expenditure, and then shrink due to fiscal consolidation needs.

Gross fixed capital formation is expected to shrink by 12% in 2020. The ­uncertainty triggered by the coronavirus crisis, revenue losses and higher debt ­levels for businesses will weigh on private investments. However, a much larger share of investments in Croatia is related to EU funds, for which we assume only short-lived disruptions during the second quarter of 2020; over the remainder
of the forecast horizon, fiscal consolidation needs could negatively impact the co-financing of EU-funded projects.

Our assumptions for the tourism sector imply a sharp drop in service exports and thus overall exports (–23%). Moreover, Italy is Croatia’s largest trading partner, which worsens supply chain disruptions for Croatian firms and the ­outlook for goods exports. However, given the high import content of private consumption, investments and tourism, we project a similarly sharp drop in imports (–24%) for 2020.

We currently see mostly downside risks to our GDP forecast for 2020. On the back of strong tourist seasons in 2021 and 2022, Croatian GDP growth will start to recover. The rebound will likely be moderate, however. While purchases of goods may just have been postponed during the crisis and may help a strong ­rebound later, revenues from services, such as tourism, are likely permanently lost. Moreover, income losses, lower savings and higher debt in the private sector as well as fiscal consolidation needs as a result of the downturn in 2020 will likely dampen the recovery. We estimate GDP growth to reach 4.2% and 4.5% in 2021 and 2022, respectively. Croatia will therefore need some years to recover the GDP losses of the current crisis. Notwithstanding the current economic downturn, Croatia’s plans to enter ERM II and to adopt the euro remain unchanged.

Economic growth in the Czech Republic has been gradually losing steam for some time, coming down from 4.4% in 2017 to 2.4% in 2019. Even before the coronavirus shock, a further slight moderation of GDP growth was expected for 2020. The coronavirus pandemic and the worldwide containment measures seen since early March have turned the entire economic outlook on its head within a couple of weeks. Not only has the forecast growth performance changed dramatically but also its structure has made a U-turn. From the current perspective – which is ­obviously subject to an enormous amount of uncertainty and significant downside risks – the Czech economy seems to be headed for a deep recession in 2020, which will be at least as pronounced as seen in 2009. Only a gradual recovery will follow in the medium term. With respect to demand-side components, the slide will be driven predominantly by private consumption and investment as a ­result of the shutdown measures.

Concerning actual health risks, the Czech health-care system appears better suited to cope with the pandemic than the health-care systems of many other ­countries in and beyond the CESEE region as various health-care indicators and rankings suggest. ¹³ More importantly, the Czech government introduced rather strict containment measures very early into the pandemic. Only 11 days after the first confirmed coronavirus patient on March 1, the government declared a state ­of emergency and by mid-March the country had largely closed its borders, most shops, restaurants as well as cultural and sports venues, had forbidden the entry of ­foreigners and issued a nationwide curfew. In addition, virtually the entire automobile industry – the lung of the Czech economy, making up nearly 25% of the country’s industrial production and exports as well as roughly 7.5% of GDP and all FDI – will come to a total standstill for at least four weeks. Moreover, the very high level of internationalization of the Czech economy and its deep integration in global value chains are likely to amplify the extent of the shock.

Overall, such massive domestic impediments magnified by an unprecedented drop in external demand will have a devastating impact on private consumption, investment as well as trade in goods and services, probably well beyond the period of the acute coronavirus threat. The damage to the economy is likely to rise over time. Hence, we expect the Czech economy to remain in a strong recession with a pronounced anti-inflationary effect for the rest of the year and only a gradual ­recovery in the medium term. Domestic demand will contribute to the decline in economic activity most strongly. The unprecedented monetary and fiscal stimulus measures as well as a weaker koruna, which has depreciated significantly as a result of the dramatic economic outlook and financial market panic, may partly offset the economic damage but only in the medium term.

Hungary’s GDP growth in 2019 surpassed our expectations (4.9% against 4.4%) and thus decelerated only slightly compared to 2018. However, we now ­expect a contraction of GDP by around 3.5% in 2020, to be followed by a gradual recovery in 2021 and 2022.

The pandemic will knock down private consumption, which will be directly hit by the suspension of entertainment activities and the closure of shops, restaurants and cafés. Second-round effects will likely include an increase in unemployment, potential wage losses due to temporary layoffs or shorter working hours, weaker wage growth, a deterioration in consumer confidence and probably weaker household credit growth (owing to greater caution on part of both households and banks). The moratorium on debt servicing until end-2020, the introduction of a temporary interest rate cap on consumption loans, the temporary suspension of tax payments and reduction in social security contributions in sectors most ­affected by the pandemic and the fact that households’ net financial wealth is at record highs are expected to provide a cushion. With pandemic effects easing off toward ­end-2020, we expect a recovery of private consumption in 2021 and 2022.

We expect economic policy to continue counteracting the impact of the ­coronavirus pandemic. The Hungarian central bank (MNB) has substantially ­expanded its liquidity provision to banks via stepped-up foreign currency swaps and the introduction of new collateralized loan facilities. It has also taken several measures to ease the administrative and regulatory burden of banks, while at the same time strengthening banking sector stability. The government and parliament have already enacted various temporary measures. The government will also make substantial changes to the 2020–2021 budget to restructure expenditure and ­possibly increase the 2021 deficit target to up to 3% of GDP. We expect some ­fiscal correction in 2021 (if, as assumed, the pandemic recedes by end-2020), to be followed by a temporary acceleration in 2022, in line with the electoral cycle ­(parliamentary elections in spring 2022). Hungary’s emergency law that enables the prime minister to rule by decree without time limits has been met with international criticism.

Investment activity was set to decelerate in 2020 and 2021 already in our ­baseline scenario, mostly on account of weakening public sector investments and housing construction. As a result of the coronavirus crisis, we expect deteriorating household income and consumer sentiment and a potentially weaker uptake
of subsidized loans to additionally slow housing construction in 2020. Capacity utilization in industry has already fallen back to its long-time average and is now expected to drop further in the coming months. Companies will likely cut back investments in 2020 in response to the damage to global value chains, weaker ­near-term demand prospects (affecting both domestic and export demand) and likely deteriorating economic sentiment. Favorable and abundant financing ­
facilities (not least due to the central bank’s various funding schemes) should partly mitigate the decline in investment activity. Investments are expected to rebound in 2021 and continue to be supported by EU funds.

We expect exports to contract sharply in 2020 as the pandemic is causing ­notable damage to global value chains, foreign supplies are being interrupted, ­foreign demand is nose-diving, and short-term global growth prospects are deteriorating. Given the high import content of exports and the decline in domestic ­demand, we expect imports to contract somewhat more than exports, leading to a small positive contribution of net real exports to overall GDP growth in 2020. For 2021 and 2022, we expect a recovery of exports, as foreign demand recovers and additional impulses come from new production capacities going online. ­Imports are expected to rise on the back of recovering domestic demand, leading to a minor negative contribution of net real exports to the GDP growth rate.

In Poland, GDP is forecast to decline by –4.3% year on year in 2020, after growth of 4.0% in 2019. In the last quarter before the coronavirus outbreak, ­export growth had declined further and imports even slightly contracted, while domestic demand growth had remained stable at almost 3%. The economic fallout from the coronavirus crisis in 2020 will arise from both gross exports and domestic ­demand, with the former having a weight of about 35% in total final demand. Given our assumptions about the developments in the euro area, Poland’s main trading ­partner, exports of goods and services are forecast to decline by about 7%, with nominal depreciation of the currency helping preserve Polish market shares in the euro area. Domestic demand in Poland, like in all other countries, will be severely hit by the health policy measures, implying a contraction by about 4% to 5% – ­indeed, there is not much scope for repeating the exceptional relative ­performance of 2008–2009.

In 2021, the economy is forecast to expand by almost 3.0%, 2.8% to be ­specific. Despite starting from a significantly lower level and being supported by a beneficial base effect, growth is forecast to be smaller in absolute terms than during the ­previous decline, on account of economic hysteresis effects and a protracted route toward normalization in terms of health policy measures. While exports will start expanding again, driven by the recovery of demand in the euro area, some ­export-oriented SMEs will likely vanish, and others struggle to regain their previous strength. Together with a recovery of sentiment in the financial markets, these losses are forecast to dampen the speed of export growth.

As a result, imports are forecast to decline even more strongly than exports in 2020, reflecting the price effect of currency depreciation and the significant ­contraction of domestic demand on top of the impact of the export decline. Thus, there will be a positive contribution of net exports to GDP growth of almost 1 percentage point. By contrast, in 2021, import growth is forecast to outpace ­export growth, as pent-up domestic demand will compound the effect of restarting foreign demand growth. This will cause the contribution of net exports to GDP growth to enter negative territory, reaching almost –1 percentage point.

Looking at the components of domestic demand, in the last quarter before the coronavirus outbreak and at the beginning of the year, consumption had already received several stimuli, ranging from fiscal transfer measures, including the ­widening of family benefits to include higher income segments and the one-off thirteenth-month pension payment, and an increase in public sector wages, to changes in the personal income tax system. At the same time, enterprises had been in a favorable financing situation with respect to both own funds (profitability, ­accumulated deposits) and external funds (low interest rates). These factors are expected to provide some cushion against the immediate impact of the corona­virus-related health policy measures. In addition, monetary policy easing (rate cut, reserve requirements, repo operations and outright government bond purchases), commercial banks’ moratorium offers to households and SMEs of up to six months, the exemption of self-employed people and micro-companies from social security payments and the economic/fiscal support package of EUR 46 billion (about 10% of GDP) will help contain a domestic demand contraction in 2020 and support a re-start in 2021. Fixed investment will be hit particularly hard in 2020, as the coronavirus effects exacerbate the already ongoing slowdown due to the impact of the fading-out EU funding cycle on local governments and (publicly owned) companies.

After a period of strong growth with signs of overheating, Romania will experience a deep recession in 2020. Both domestic and external demand will be ­substantially hit by the economic impact of the spread of the coronavirus and ­related containment measures. While the first quarter of 2020 will already be partly affected, a deeply negative growth rate is projected for the second quarter. The recovery will likely be gradual, as uncertainties will depress economic sentiment and hysteresis effects stemming from rising unemployment and insolvent companies will prevent a more forceful recovery.

The temporary closure of the two major car producers Dacia and Ford and other companies in the manufacturing sector starting from mid-March, will have an immediate impact on exports. We do not expect a fast return of production to full capacities, though closures were initially announced to last only for some weeks. In addition to supply-side factors, exports will also be negatively affected by plummeting external demand in 2020. After a marked decline of exports in 2020, base effects and recovering external demand will enable exports to grow again in 2021. Due to the high import content of exports and shrinking domestic demand, imports will also contract considerably in 2020. This might induce a very small, positive contribution of net exports in 2020, which will turn negative in 2021.

High wage growth will come to an end due to rising unemployment and wage losses triggered by temporary layoffs. This will add to the negative ­impact of ­containment measures (such as restrictions on movement) and confidence ­effects on private consumption. Hence, we expect negative private consumption growth in 2020, followed by a hesitant recovery in 2021 and 2022. Households’ deteriorating income position and loss of confidence will also have an impact on capital ­formation, in the form of less residential building construction. While debtors are allowed to postpone their loan repayments, demand for new housing loans might decline during the economic downturn. Residential building investment is likely to ­recover only slowly but investments supported by EU funds will enable some recovery of capital formation (provided the Romanian government manages to free up ­budgetary resources for cofinancing). At the same time, shrinking earnings and confidence effects will negatively affect corporate investment – definitely in 2020 and possibly also beyond. However, state guarantees for investment loans to SMEs will provide some cushion. In the medium term, Romania could benefit from a reallocation of production capacities, as multinational European companies might try to increase capacities in Europe to make production chains less vulnerable to global disruptions.

As growth in recent years was fueled by procyclical fiscal policy leading to a high budget deficit, fiscal space to counteract recessionary developments appears limited. Fiscal consolidation is inevitable in the near future and will slow down the recovery in 2021 and 2022. By purchasing government bonds on the secondary markets, the central bank has supported the financing of public spending. Yet, monetary policy is subject to certain constraints, too. In this respect, the current account deficit and the remaining stock of foreign currency loans, in particular
in the household sector, should be noted. External imbalances might reinforce ­depreciation pressures on the exchange rate in times of international financial ­market turbulences. Hence, under tense market conditions, the central bank needs to strike a balance between managing the exchange rate (to contain balance sheet effects) and injecting more liquidity. It should be noted that cutting minimum ­reserve requirements for leu and foreign currency liabilities still is available as a policy option for the central bank.

Negative spillover effects from Italy cloud the outlook for the Romanian economy further due to a comparatively high degree of trade integration, the presence of Italian investors and the declining income of labor migrants. On a positive note, the impact of the crisis through tourism will be quite limited in Romania, as tourism plays a much smaller role in Romania than e.g. in Croatia and is also markedly less relevant than in neighboring Bulgaria.

3 Russian economy set to contract in 2020

If the weakening of the global economy remains temporary, Russia is expected to stick to its current spending plans. Under the approved 2020–2022 budget framework, public sector spending will increase by 6% to 7% a year in nominal terms. With the current oil price outlook, however, the net deficit will turn ­negative this year. Russia’s ability to deal with external shocks is supported by the Russian ­government’s comfortable financial position. The Russian government carries little debt (14% of GDP), but U.S. sanctions restrict its ability to borrow inter­nationally. Russia has also built up its sovereign wealth fund again. The ­National Welfare Fund now holds about USD 150 billion in liquid assets (9% of GDP). So far, public sector support measures related to the coronavirus situation have been relatively limited in terms of spending and have been realized within the current budget framework, but ­additional measures may be decided.

Consumption is expected to contract this year hampered by slower economic activity and measures to contain the coronavirus pandemic. Recent ruble depreciation is likely to increase inflation and thus reduce purchasing power. In addition, consumption growth will be restrained by lower growth in consumer credit due to tighter regulation. In coming years, consumption growth is expected to pick up, supported by a gradual rise in incomes, including planned increases in ­social spending. Fixed investment recovered slightly last year, growing by slightly more than 1%. Fixed investment is expected to decline substantially this year due to the deteriorating economic outlook over the near term; notably, private sector capital formation is expected to decline sharply, whereas public sector investment may slightly ­increase, with the government attempting to push forward national projects. Fixed investment should then gradually pick up in the coming years as spending on national projects gets into full swing.

Russian exports contracted by 2% in 2019 for the first time in decades. Much of the contraction was due to a decline in metal and grain exports. Exports are expected to contract even more this year due to the weakness in the global economy. Exports should recover gradually next year as demand returns. Russian imports grew moderately last year supported by ruble appreciation. Imports are expected to fall substantially this year, constrained particularly by ruble depreciation from lower oil prices. Imports should also be dampened by investment demand’s public sector focus, which favors domestic suppliers. In addition, imports of ­tourism services will fall substantially due to the coronavirus pandemic.

In the current situation, forecasting risks are exceptionally high. Russia’s economic development may turn out considerably weaker if the coronavirus ­pandemic is prolonged and oil prices fall further for an extended period. Russia, however, has the capacity to soften the impact of negative shocks through government support. The planned referendum on changes to the Russian constitution and the State Duma elections in 2021 could also create pressure to increase public ­sector spending. In addition, a stronger than expected fall in imports could ­mitigate the GDP contraction.

Studies

Macroprudential policies in CESEE – an intensity-adjusted approach

Markus Eller, Reiner Martin, Helene Schuberth, Lukas Vashold ¹⁴

We assess the overall intensity with which macroprudential policies were used in eleven Central, Eastern and Southeastern European (CESEE) countries from 1997 until end-2018. To this end, we construct an intensity-adjusted macroprudential policy index, which also allows us to gauge the impact macroprudential measures had on credit growth and housing prices. Our new index reveals that some of the eleven CESEE countries had already intensively implemented macroprudential policy tools before the global financial crisis (GFC), while others became more active in this respect only in its aftermath. The considerable macroprudential tightening evident since 2010 mainly reflects the introduction of borrower-based measures, like loan-to-value (LTV) and debt service-to-income (DSTI) limits, and the implementation of capital buffers. In the empirical assessment, we find that macroprudential measures are associated with lower private sector credit growth, in particular for households. Moreover, borrower-based macroprudential measures tend to have a larger and more robust impact on credit growth than other macroprudential instruments that also include capital- and liquidity-based measures. These findings also hold for the impact of macroprudential measures on house price growth.

JEL classification: E58, E61, G18, G28

Keywords: macroprudential policies, intensity adjustment, composite indicator, CESEE, credit growth, house price growth, financial stability

Macroprudential policy encompasses a large number of tools. Some resemble traditional microprudential tools, e.g. capital and liquidity requirements; others target borrowers’ behavior, e.g. loan-to-value (LTV) and debt service-to-income (DSTI) limits. Macroprudential tools may be gradual, such as a change in the LTV limit from 90% to 80%, binary, e.g. foreign currency (FX) loans are permitted or not, and some can change from gradual to binary or vice versa. Accurately quantifying MPP activity by capturing not only the occurrence of such policies but also their strength is thus highly challenging ¹⁵ .

In fact, most of the literature on macroprudential policy captures only the ­extensity, i.e. the occurrence of macroprudential policies, using very simple indices. Some authors use a basic binary indicator signaling whether a certain instrument was in place at a given time (e.g. Reinhardt and Sowerbutts, 2015; Cerutti et al., 2017a). Most studies use an index where a tightening measure is coded with +1, a loosening measure with –1 and ambiguous measures are not taken into account. By cumulatively summing up the values over time, a simple time series of overall ­macroprudential extensity can be compiled (see, for instance, Shim et al., 2013; Cerutti et al., 2017b; Ahnert et al., 2018). The intensity of the measures is, however, not taken into account. To take an example from Vandenbussche et al. (2015), these indices would record a lowering of the maximum LTV ratio from 100% to 90% in the same way as a reduction from 100% to 60%. Obviously, however, the effect of these two measures is unlikely to be the same.

It is no easy task to construct an index that genuinely reflects the intensity of macroprudential policy given the difficulty of comparing and quantifying the ­effects that different measures have on the overall macroprudential stance. One could even argue that constructing a “perfect” index is practically impossible. However, in order to help make informed policy decisions and to explore and understand the effects of macroprudential policy on different economic developments, having an indicator that could be used for such modeling exercises would prove very useful. ¹⁶

There are only few examples of intensity-adjusted macroprudential indices. Most notably, Vandenbussche et al. (2015) construct an intensity-adjusted macroprudential policy index for 16 CESEE countries to investigate the effects of macroprudential measures on housing prices. Dumičić (2018) also accounts for possible differences in the intensity of measures. Richter et al. (2019) and Alam et al. (2019) both focus on LTV limits and provide detailed information on the intensity of usage of this specific instrument. However, the authors of both papers stopped short of compiling an overall MPP index.

In this paper, we build and expand on previous intensity-adjusted approaches and construct a novel, overall index for macroprudential policy in CESEE. Our macroprudential policy index (MPPI) accounts not only for “classic” macroprudential instruments but also for other requirements motivated by macroprudential objectives. The index is based on information derived from four databases of macroprudential measures and applies a set of weighting rules partly leaning on those developed by ­Vandenbussche et al. (2015). The MPPI covers eleven CESEE EU Member States from 1997 to 2018 on a quarterly basis. The novelty of our index compared to the existing literature lies – next to the aforementioned intensity adjustment – in ­covering a comparatively long time span (of more than 20 years) and a large variety of MPP instruments as well as in differentiating between the announcement and implementation of macroprudential measures.

This paper is structured as follows: section 1 discusses a taxonomy of macroprudential policy measures and introduces the MPPI, with section 2 providing an overview of its country-specific development over time. In section 3, we use the MPPI to assess the impact of macroprudential policy measures on key macrofinancial variables, such as credit growth and house prices. Section 4 concludes.

1 Constructing an intensity-adjusted macroprudential policy index

Chart 1 provides a hierarchical overview of the various subindices and other components we draw on to construct the MPPI. Classic (“narrow”) macroprudential instruments are grouped into three subindices for capital-based, borrower-based and liquidity-based measures, respectively. First, the capital-based macroprudential policy subindex (CB-MPPI) comprises the countercyclical capital buffer (CCyB), the systemic risk buffer (SyRB), the capital conservation buffer (CCoB), institution-specific buffers for other systemically important institutions (O-SII buffer) and risk weights (RWs) ¹⁷ . Second, the borrower-based macroprudential policy subindex (BB-MPPI) encompasses limits in the LTV ratio for new collateralized house purchase loans, the DSTI ratio and outright bans on foreign currency lending. Other borrower-based instruments like debt-to-income (DTI) or loan-to-income (LTI) limits, having only played a negligible role in the CESEE countries, have not been included so far. ¹⁸ Third, the liquidity-based macroprudential policy subindex (LB-MPPI) consists of liquidity requirements (like short-run liquidity ratios) and a variety of other prudential measures, such as limits on large exposures and specific sectors. It also captures intragroup exposure limits and rules on foreign currency mismatches. The three “classic” macroprudential subindices taken together constitute what we call the narrow macroprudential policy index (N-MPPI).

To obtain the extended version of the macroprudential policy index, or MPPI, we include the additional measures mentioned at the beginning of this section. First, we account for system-wide (Pillar 1) minimum capital requirements, comprising both the applicable capital adequacy ratio and the tier 1 capital ratio. Second, we include minimum reserve requirements. Many macroprudential measures, in particular in the CESEE countries, differentiate between local and foreign currencies, with foreign currency requirements being usually stricter than national currency require­ments to account for exchange rate risks. Foreign currency requirements were not included in the MPPI as a separate subindex. Instead, they are taken into account within the various subindices.

The second step in the construction of the MPPI is weighting and aggregating the individual measures, which is a difficult and, to a certain extent, inevitably ­arbitrary task, as already alluded to in the introduction. To go beyond simple ­binary indicators, it is necessary to define specific weights for the various macroprudential instruments included in the MPPI. Increasing the CCyB by 50 basis points might, for example, constitute a more, or less, severe change in the intensity of overall macroprudential policy than, say, reducing the upper limit on the LTV ratio by 5 percentage points. Depending on the nature of the macroprudential instruments, we apply three different weighting rules to the individual instrument feeding into the MPPI. Many of these rules are fully or partly based on Vandenbussche et al. (2015), which, to our knowledge, reflects the most comprehensive previous attempt to calculate an intensity-weighted ­macroprudential index. However, the database used in Vandenbussche et al. (2015) ended nearly ten years ago and does not include newer instruments, such as capital buffers. Moreover, the database provides only the implementation dates of measures and not the announcement dates. Our three weighting rules are as follows:

First, we use face value aggregation for instruments that are captured by a single number and where a change clearly indicates a tightening or loosening of macroprudential policy intensity. For example, a 1-percentage-point increase of the CCyB leads to a 1-index-point increase in both the extended and the narrow MPPI as well as in the CB-MPPI.

Second, for more complex measures or measures where a decrease represents a tightening of macroprudential intensity, a formula-based aggregation is used. This usually involves a default limit from which the actual limit is subtracted and a weighting scheme to ensure that the impact of changes in the measure concerned is appropriately reflected. For instance, a 5-percentage-point reduction in the LTV limit is equivalent to a 1-index-point increase in the MPPI.

Third, for macroprudential instruments that are too specific or too complex to be properly captured through a formula, such as changes in definitions for the calculation of minimum capital requirements, we use an augmented dummy approach. It works similarly to the approach used in many other studies and assigns fixed positive (negative) values for tightening (loosening) incidents. However, even for such cases we introduce some nuances in the scoring. For example, a change in the type of capital required for overall reserve requirements has a stronger impact on the MPPI than a change in the type of capital applicable only to certain exposures. This approach allows us to include a range of macroprudential measures that are hard or impossible to quantify otherwise. In total, around one-third of the measures included in the overall index were coded in this way.

Another important aspect when aggregating the MPPI is that many instruments included in the MPPI may either be recommendations or legally binding acts. Given that the former are not as stringent as the latter, the weight attributed to recommendations in the index is lower than the weight attributed to equivalent measures that are legally binding.

As we focus in our analysis on changes in the intensity of the various types of macroprudential measures, the initial level of the indices is of secondary importance. In this study, we subtract the value of the index in the first period from the whole series thereafter. This way, the index and each subindex start at zero. Positive (negative) values in the subsequent periods then signal a net tightening (loosening) of aggregate macroprudential policy. Moreover, given that most CESEE countries in our sample had implemented only few if any macroprudential policy measures before the mid-1990s, it is realistic to assume that cross-country differences in the macroprudential policy stance were not that large in 1997, when our index starts.

For compiling the MPPI, we use and cross-check information derived from four MPP databases: the ECB’s Macroprudential Policies Evaluation Database (MaPPED) documented in Budnik and Kleibl (2018), the CESEE-related database of Vandenbussche et al. (2015), the IMF iMaPP database described in Alam et al. (2019), and the database of the European Systemic Risk Board (ESRB) described in Kochanska (2017). ¹⁹ All these databases provide implementation dates, i.e. the dates when certain measures take effect. MaPPED and the ESRB database also provide announcement or decision dates. In the econometric analysis, we generally use the announcement dates for tightening and the implementation dates for loosening ­incidents. The rationale behind this is that a credit institution is likely to react to the announcement of a tightening measure, e.g. an increase in minimum capital requirements instantaneously by building up capital reserves if needed. In contrast, the announcement of a loosening policy action for the same instrument is unlikely to trigger an immediate reaction given that the old regulations stay in place until the date of implementation.

The online supplement to this article, available on the OeNB’s website , provides more details on underlying data sources, the categorization of the different macroprudential policy measures and the applied weighting rules. In it, we moreover show the evolution of each subindex by country over time.

2 Country-specific analysis of macroprudential policies

The tightening in the late 1990s was mainly due to increased minimum capital requirements (shown in dark green in chart 2) and, to a lesser extent, to liquidity-­based measures (marked brown). Minimum reserve requirements (shown in light green) were eased countercyclically in many countries in the wake of the GFC and then largely remained at these levels. The tightening evident since 2010 mainly reflects the implementation of borrower-based measures (marked orange) that gained importance after the GFC, but their use stagnated somewhat in recent years. Finally, the accelerated tightening since 2014 was mainly due to the implementation of capital buffers (marked dark blue) with a view to complying with Basel III rules and the respective EU directives. These regulatory changes affected all countries in our sample rather uniformly (and are therefore not discussed further in the subsequent country-specific paragraphs).

Minimum capital adequacy ratios and tier 1 capital ratios were harmonized in 2014–2015, which implied an easing for a few countries (Bulgaria, Croatia, Estonia and Latvia) and a tightening for several others (the Czech Republic, Hungary, Lithuania, ­Poland, Slovakia and Slovenia). At the same time, all countries activated various capital buffers. Starting in 2014 (or later), positive rates were set (1) for the CCoB in all countries (reaching 2.5% throughout in 2019), (2) for the institution-specific O-SII buffer in all countries but the Czech Republic and (3) for the SyRB in all countries but Latvia, Lithuania and Slovenia. The SyRB was initially often used as a substitute for the O-SII buffer, which became available in 2016; and later to ­complement the maximum O-SII buffer rate of 2%. ²² Some of the countries also used the SyRB to compensate for a decrease in Pillar 1 capital requirements, which resulted from the above-mentioned adjustments to EU legislation. Regarding the CCyB, there is more country-specific discretion: until end-2018, positive CCyB rates had been set only in the Czech Republic and in Slovakia (effective since 2017), in Lithuania (effective since end-2018) and Bulgaria (effective since 2019).

The aggregate picture (see the CESEE-11 panel in chart 2) conceals pronounced heterogeneity across countries. ²³ Let us focus, first, on countries that tightened macroprudential policy already in the late 1990s or early 2000s and reinforced the use of MPP in the years before the GFC, mostly motivated by extraordinarily strong credit growth at the time, i.e. Bulgaria, Croatia and Romania.

2.1 Bulgaria, Croatia and Romania

On the back of strong credit growth, mostly denominated in foreign currency, and rising financial and macroeconomic imbalances, Croatia started using macroprudential policies in the late 1990s. The minimum capital adequacy ratio was ­increased from 8% to 10% in 1998, foreign exchange liquidity requirements were introduced in 2001 and maximum limits on banks’ open foreign currency claims in 2003. Single client exposure limits were implemented in 1998 and intragroup exposure limits in 2002. With the introduction of marginal and special reserve requirements, the overall minimum reserve requirement rates were gradually ­increased between 2004 and 2006. In 2006, risk weights for FX mortgage loans to unhedged borrowers were raised from 50% to 75% and, in 2008, to 100%, those for foreign currency-indexed consumer loans from 100% to 150%. ²⁵ When the GFC hit, Croatia relaxed its macroprudential policy stance until 2014, e.g. by ­releasing the system reserves accumulated during the boom phase. Risk weights for currency-induced credit risk were also relaxed in 2010, as part of the adjustment to Basel II, but this easing was offset by an increase in the minimum capital adequacy ratio to 12%. In 2014, the minimum capital adequacy ratio was lowered again to 8% with a view to complying with Basel III but was counterbalanced by implementing several capital buffers.

Romania significantly tightened its macroprudential policy stance already before 2007, mainly by increasing the minimum capital requirements and, to some extent, also the minimum reserve requirements. In 2004, the central bank started to raise reserve requirements on foreign currency deposits while reducing those for local currency. These instruments were complemented with borrower-based measures: an LTV ceiling for housing loans of 75% and DSTI limits (30% on consumer loans, 35% on mortgage loans) were introduced in 2004 already, as well as intragroup exposure limits. However, to harmonize its minimum capital requirements with those of the EU, Romania had to loosen its macroprudential stance by reducing the minimum capital adequacy ratio from 12% to 8% in 2007, and by abandoning ­explicit ceilings on LTV and DSTI ratios. ²⁶ When the GFC hit, Romania reversed some of the tightening (especially by loosening reserve requirements) but retightened its macroprudential stance in 2011. The increase in the MPPI at the time mainly ­reflects the reintroduction of LTV limits and further restrictions for FX loans ­regarding both the LTV and DSTI ratio. A tightening of the maximum DSTI ratio to 40% was announced in 2018 and implemented in 2019.

2.2 Poland, Slovenia and Estonia

Poland escaped the GFC relatively unscathed, and the early use of MPP measures arguably played an important role in preventing banks from suffering large losses during the GFC (Sławiński, 2020). While the MPP stance was loosened between 1999 and 2005, almost exclusively reflecting a reduction of risk weights on local and FX mortgages, it was tightened thereafter. Risk weights for loans with a high LTV ratio were increased starting in 2004. In June 2006, the Polish Commission for Banking Supervision introduced “Recommendation S,” according to which LTV and DSTI limits were tightened and the risk weights for FX mortgage loans were raised from 35% to 75% with effect from 2008. The immediate impact of this regulation was not so much a curbing of the lending boom, but it directed most FX loans toward households that could withstand a sizable ­depreciation of the złoty (Zettelmeyer et al., 2011). Risk weights for FX mortgage loans were further tightened to 100% in 2012, while, in 2014, the LTV threshold above which a risk weight of 100% applies was raised from 50% to 80% for loans denominated in both foreign and domestic currency, which translated into an easing in this policy area. In another tightening step, the risk weights on FX loans were increased to 150% in 2018. Starting in 2010, borrower-based measures were significantly tightened by introducing a maximum DSTI ratio of 50% for borrowers earning less than the average national wage, and of 65% for other borrowers. Similar DSTI restrictions were applied for FX-denominated loans to unhedged borrowers. Since 2014, LTV limits have been reduced from 100% to 80%. In 2014, FX lending to households without income in the same currency was banned altogether. ²⁷

Slovenia’s MPP stance remained unchanged until 2004, when risk weights on mortgage loans were increased from 50% to 100%. They were relaxed again to 35% in 2007, but, at the same time, the LTV threshold above which a risk weight of 100% applies became more binding, having been set at 60%. Minimum capital reserves were tightened somewhat in 2007, when capital requirements for operational risk were introduced, and in 2015, when the minimum tier 1 capital ratio was increased from 4% to 6%. Capital buffers gained some importance in recent years but not as strongly as in the other CESEE countries. Since 2007, the central bank required banks to assess clients’ creditworthiness based on euro equivalent values of FX loans (Bank of Slovenia, 2008). Borrower-based measures have been used more intensively since 2016, when the central bank started issuing several tightening recommendations regarding LTV and DSTI limits.

Like Bulgaria and Croatia, Estonia tightened minimum capital and reserve require­ments considerably in the late 1990s to improve the resilience of the banking sector in view of speculative attacks against the Estonian kroon in October 1997 and potential spillovers from the Russian financial crisis in 1998. In 2006, in light of very rapid credit growth, the central bank increased the minimum capital requirement further and for the first time raised the risk weights on housing loans (from 50% to 100%). The latter were loosened again considerably (to 35%) during the GFC. Before Estonia joined the euro area, minimum reserve requirements were relaxed significantly in 2010 (from 15%) to gradually approach the rate set by the ECB. Besides the activation of several capital buffers in 2014, borrower-based measures were tightened in 2015, when the LTV cap for new housing loans was decreased to 85% and a DSTI limit of 50% was introduced.

2.3 Czech Republic, Slovakia, Hungary, Lithuania and Latvia

The Czech Republic left its MPP stance largely unchanged until it raised the minimum tier 1 capital ratio from 2% to 6% and activated various capital buffers in 2014, which it continued to tighten. Notably, the Czech Republic was among the first countries in the EU that applied a positive CCyB rate. In parallel, since 2015, liquidity requirements (i.e. floors for liquidity coverage ratios) and borrower-based measures were tightened gradually. To date, the Czech National Bank (CNB) only has a mandate to issue recommendations for borrower-based measures. It issued tightening recommendations in 2015 and 2016 for gradually reducing the LTV limit from 100% to 90%, and, in 2018, for reducing the maximum DSTI ratio. The pronounced tightening of the overall MPP stance in the Czech Republic in recent years was primarily motivated by an expansionary phase of the credit cycle since 2014 and very dynamic housing loan growth (CNB, 2019).

Slovakia followed a pattern very similar to that of the Czech Republic. Even though a formal LTV cap had already been in place since 1996, it was only applicable to a narrow definition of mortgage loans and could be bypassed by applying for other housing loans (Vandenbussche et al., 2015) ²⁸ . In response to an increase in these other housing loans, a new package of LTV limits covering all collateralized mortgages was launched. The central bank, Národná banka Slovenska (NBS), ­recommended a further tightening in 2014, which became legally binding in 2016. Moreover, in 2014, the LTV threshold above which a risk weight of 100% applies was raised, which resulted in a loosening in this policy area. In light of rising ­property prices and household indebtedness, borrower-based measures were ­tightened again in 2018. Among other things, mortgages with an LTV ratio of more than 90% were prohibited and the share of new loans with an LTV ratio ­between 80% and 90% was restricted (NBS, 2018). Together with the activation and gradual increases of applicable capital buffers, this has substantially tightened the macroprudential environment until end-2018.

Hungary considerably tightened its macroprudential policy stance only after the GFC, having since then put a strong emphasis on borrower-based measures, above all to reduce the high share of FX borrowing. In 2010, Hungary introduced ­maximum LTV ratios on household mortgage lending (and car loans). Upper limits were set for real estate borrowing, namely 75% for loans denominated in forint (raised to 80% in 2012), 60% for loans in euro, and 45% for loans in other currencies. Maximum DSTI ratios for FX loans differentiated by currency were also set in 2010, and later in that year, FX mortgages were banned altogether. The ban was discontinued in 2015, which contributed to an easing of the macroprudential stance, while the LTV and DSTI limits were tightened further that year (DSTI limit again in 2018). Moreover, more stringent liquidity requirements and the activation and tightening of various buffer rates had a considerable macroprudential tightening impact since 2014.

Lithuania used a comparatively rich macroprudential toolkit. The net easing until mid-2011 was mainly due to reduced risk weights on mortgages (from 100% to 50% in 2001) and gradually relaxed minimum reserve requirements since 1997. By contrast, liquidity-based measures had a counterbalancing tightening impact. The net overall MPP tightening since 2011 was mainly driven by borrower-based measures and the activation of various capital buffers. For mortgage loans, the LTV cap was reduced from 100% to 85% and the DSTI limit from 60% to 40% in 2011.

Latvia is a special case since its overall macroprudential stance remained broadly unchanged during the whole observation period. Tightened minimum capital ­requirements, buffer requirements and borrower-based measures were outweighed by reduced risk weights and reserve requirements. When we focus only on “narrow” MPPs, there was some net tightening since 2014, which was mainly attributable to the implementation of capital buffers.

3 The impact of macroprudential tightening on macrofinancial variables

According to several papers, tighter MPPs are associated with slower credit growth, especially if the focus is on household credit and on the impact of borrower-based measures (e.g. Akinci and Olmstead-Rumsey, 2018; Alam et al., 2019; Dumičić, 2018). Using a large panel for the period from 2001 to 2013, Cerutti et al. (2017a) find that, in general, borrower-based measures have a stronger effect on credit growth than other MPPs, in both advanced and emerging economies. Furthermore, their results suggest that while both household and corporate credit have a negative ­relationship with MPPs, the effect on corporate credit is weaker and often statistically insignificant. They argue that this is the case because MPPs are usually directed at financial institutions or households rather than corporations (while personal loans taken out by firm owners to finance their businesses could still be affected by ­borrower-based measures).

The empirical evidence regarding the impact of MPPs on house prices is less clear cut, though. Referring to a few of the related papers, Vandenbussche et al. (2015) found that tighter capital-based MPPs and nonstandard liquidity measures (mostly minimum reserve requirements) have a dampening effect on house price inflation in CESEE. Akinci and Olmstead-Rumsey (2018) concluded that almost exclusively housing-related MPPs, particularly LTV and DSTI caps, constrain housing credit growth and house price appreciation, especially in countries where private sector financing via the banking sector plays an important role. Kuttner and Shim (2016) compared the effectiveness of MPPs with that of fiscal policy measures: while housing credit growth would be mitigated by both tighter ­borrower-based MPPs and an increase in housing-related taxes, the latter is the only policy tool with a discernible impact on house price appreciation. Finally, there could also be differences across different levels of economic development: Alam et al. (2019), for instance, identified a negative effect of MPPs on house prices only in the case of advanced economies.

3.1 Econometric specification

yi,t = βMPPIt–1 + Xiγ + τi + εi,t,

(1)

To mitigate the problem of endogeneity (see Galati and Moessner, 2013), at least to some extent, we include lags of the policy measures and of the other ­covariates (except for the crisis dummy) instead of using contemporaneous data. Pre-estimation diagnostic tests also supported the inclusion of one lag of each ­regressor (based on the minimal Schwarz’s Bayesian information criterion). ­Moreover, the dependent variables seemed to be stationary based on the panel unit root test by Pesaran (2007). We conducted robustness checks, including a larger number of control variables, such as inflation, equity prices or the real effective exchange rate, the results of which are available on request.

As pointed out in section 2, the intensity with which macroprudential policies have been used in the CESEE countries under investigation is very heterogeneous, a finding also supported by a pre-estimation panel test for slope homogeneity (Pesaran and Yamagata, 2008). To account for panel heterogeneity, we estimate two other models that allow for heterogeneous slope coefficients, namely the dynamic fixed effects model (DFE) and the mean group estimator model (MG; see Pesaran and Smith, 1995). However, as a caveat, these procedures are mostly applied for panels with large N and T. As our dataset consists of eleven countries, this may lead to results that are driven by outliers, especially for house price growth, where the time series is relatively short. Nonetheless, these additional estimation models are useful complements to our baseline OLS-FE specification. A detailed description of the definitions, data sources and data availability can be found in table A1 in the annex.

3.2 Results

The decline in total credit growth following a tightening of macroprudential policies appears to be primarily driven by a decrease in household credit growth. For this variable, the estimated coefficients are larger in magnitude and statistically significant across the MPPI, N-MPPI and BB-MPPI indices (in the baseline OLS-FE specification). Corporate credit growth also shows a negative relation but is seldom statistically significant. Borrower-based MPPs again seem to have the most significant effect in dampening both household and corporate credit growth.

For house price growth, all estimated coefficients are negative and largest in the case of the borrower-based subindex, for which they are also statistically ­significant. This provides further evidence for the effectiveness of such measures to dampen house price growth. However, the aforementioned rather short time series for house price growth renders the interpretation of these estimates somewhat less reliable than those for credit growth.

The general observation that the magnitude of coefficients increases for more narrowly defined MPP indices could imply that additional measures included in the extended MPPI are in fact diluting the impact of the more effective ones, such as ­borrower-based measures. This is confirmed when we run the same estimations for other MPPI subindices, namely the capital-based (CB-MPPI) and liquidity-based (LB-MPPI) ones. For the sake of brevity, we do not show the results here in detail. The estimated coefficients often exhibited a positive sign and were not statistically significant, except for some specifications looking at house price growth.

In general, the results of this empirical exercise are broadly in line with previous empirical work and suggest that tighter MPPs are associated with lower private sector credit growth. Furthermore, we can confirm key findings of Cerutti et al. (2017a), namely that borrower-based measures seem to be more effective than other MPPs in containing credit growth and that the ­effect is more pronounced for household than for corporate credit growth. For house price growth, we find that even though broader MPPs tend to have a ­negative relationship with house prices, again only borrower-based measures seem to have a statistically significant effect (which corroborates the findings of Akinci and ­Olmstead-Rumsey, 2018).

4 Summary and conclusions

There are considerable heterogeneities across CESEE countries with respect to the composition of instruments and the timing of MPP instrument activation. We identified three clusters of countries. First, Bulgaria, Croatia and Romania implemented macroprudential policy already in the late 1990s or early 2000s and reinforced its use in the years before the GFC – mostly motivated by extraordinarily strong credit growth at the time. Second, Estonia, Poland and Slovenia also used MPP measures before the GFC but to a lesser extent than the first group of countries, and they partly applied a less differentiated instrument toolkit. Third, the Czech Republic, Hungary, Lithuania, Slovakia and partly Latvia initially made only limited use of MPP tools before considerably tightening their MPP stance just after the GFC and especially in the past few years.

Given that the recent MPP tightening in the region was driven more strongly by capital- than borrower-based measures and that it went along with widespread house price increases, the question arises whether there is room to optimize the choice of instruments. To assess the impact of macroprudential policy intensity on the key macrofinancial variables credit and house price growth, we use a set of panel regressions and find that the use of macroprudential policies is effective in lowering credit growth, both in the short and medium term. In line with ­previous research, we find that borrower-based measures, such as LTV and DSTI limits, tend to have a stronger and more significant impact than other MPP ­measures that also include capital- and liquidity-based instruments. Broadly confirming previous findings, we conclude that MPPs are more effective in containing household than corporate credit growth. In the same vein, borrower-based measures seem to be relatively more effective in dampening house price growth.

Further research is required to study more closely the role of different transmission channels to better understand the reasons for cross-country heterogeneity and to explore the effectiveness of different sets of macroprudential instruments and their possible interactions with other policy instruments (especially those of monetary policy). The index and its components presented in this paper could serve as an important contribution to the quickly evolving literature in this field of research.

References

Ahnert, T., K. Forbes, C. Friedrich and D. Reinhardt. 2018. Macroprudential FX Regulations: Shifting the Snowbanks of FX Vulnerability? NBER Working Paper No. w25083.

Akinci, O. and J. Olmstead-Rumsey. 2018. How effective are macroprudential policies? An empirical investigation. In: Journal of Financial Intermediation 33(C). 33–57.

Alam, Z., A. Alter, J. Eiseman, R. Gelos, H. Kang, M. Narita, E. Nier and N. Wang. 2019. Digging Deeper – Evidence on the Effects of Macroprudential Policies from a New ­Database. IMF Working Paper No. 19/66.

Bank of Slovenia. 2008. Financial Stability Review. May.

Budnik, K. B. and J. Kleibl. 2018. Macroprudential Regulation in the European Union in 1995–2014: Introducing a New Data Set on Policy Actions of a Macroprudential Nature. ECB Working Paper No. 2123.

CNB. 2019. Financial Stability Report 2018/2019. June.

Cerutti, E., S. Claessens and L. Laeven. 2017a. The use and effectiveness of macroprudential policies: New evidence. In: Journal of Financial Stability 28. 203–224.

Cerutti, E., R. Correa, E. Fiorentino and E. Segalla. 2017b. Changes in Prudential Policy Instruments – A New Cross-Country Database. In: International Journal of Central Banking 13(1). 477–503.

Dumičić, M. 2018. Effectiveness of Macroprudential Policies in Central and Eastern European Countries. In: Public Sector Economics 42(1). 1–19.

ESRB. 2019. Features of a macroprudential stance: initial considerations. Report of the ESRB Expert Group on Macroprudential Stance. April.

Galati, G. and R. Moessner. 2013. Macroprudential policy – a literature review. In: Journal of Economic Surveys 27(5). 846–878.

Galati, G. and R. Moessner. 2018. What do we know about the effects of macroprudential policy? In: Economica 85(340). 735–770.

Gardó, S. 2008. Croatia: Coping with Rapid Financial Deepening. In: Focus on European Economic Integration 1/08. OeNB. 61–81.

Hilbers, P., I. Otker-Robe, C. Pazarbasioglu and G. Johnsen. 2005. Assessing and Managing Rapid Credit Growth and the Role of Supervisory and Prudential Policies. IMF Working Paper No. 05/151. July.

Kochanska, U. 2017. The ESRB Macroprudential Measures Database. In: IFC Bulletins chapters 46.

Kuttner, K. and I. Shim. 2016. Can non-interest rate policies stabilize housing markets? Evidence from a panel of 57 economies. In: Journal of Financial Stability 26. 31–44.

NBS. 2018. Financial Stability Report. November.

Pesaran, M. H. 2007. A simple panel unit root test in the presence of cross-section dependence. In: Journal of Applied Econometrics 22(2). 265–312.

Pesaran, M. H. and R. Smith. 1995. Estimating long-run relationships from dynamic heterogeneous panels. In: Journal of Econometrics 68(1). 79–113.

Pesaran, M. H. and T. Yamagata. 2008. Testing slope homogeneity in large panels. In: Journal of Econometrics 142(1). 50–93.

Reinhardt, D. and R. Sowerbutts. 2015. Regulatory Arbitrage in Action: Evidence from Banking Flows and Macroprudential Policy. Bank of England Working Paper No. 546.

Richter, B., M. Schularick and I. Shim. 2019. The Costs of Macroprudential Policy. In: Journal of International Economics 118. 263–282.

Shim, I., B. Bogdanova, J. Shek and A. Subelyte. 2013. Database for Policy Actions on Housing Markets. BIS Quarterly Review Q3 2013.

Sławiński, A. 2020. Central, Eastern and Southeastern Europe’s reunion with Europe. In: Holzmann, R., D. Ritzberger-Grünwald and H. Schuberth (eds.). 30 Years of Transition in Central, Eastern and Southeastern Europe. Cheltenham/Northampton: Edward Elgar, forthcoming.

Vandenbussche, J., U. Vogel and E. Detragiache. 2015. Macroprudential Policies and Housing Prices: A New Database and Empirical Evidence for Central, Eastern, and Southeastern Europe. In: Journal of Money, Credit and Banking 47(S1). 343–377.

Vandenbussche, J., P. Kongsamut and D. Dimova. 2018. Macroprudential Policy Effectiveness: Lessons from Southeastern Europe. In: Journal of Banking and Financial Economics 1(9). 60–102.

Vujčić, B. and M. Dumičić. 2016. Managing Systemic Risks in the Croatian Economy. BIS Papers No. 86. 75–84.

Zettelmeyer, J., P. Nagy and S. Jeffrey. 2011. Addressing private sector currency mismatches in emerging Europe. In: Kawai, M. and E. Prasad (eds.). Financial Market Regulation and Reforms in Emerging Markets. Washington, D.C.: Brookings Institution Press.

Annex

CESEE-related abstracts from other OeNB publications

Assessing credit gaps in CESEE based on levels justified by fundamentals – a comparison across different estimation approaches

Mariarosaria Comunale, Markus Eller, Mathias Lahnsteiner

OeNB Working Paper 229. March 2020.

Austrian banks’ expansion to Central, Eastern and Southeastern Europe: Milestones – review and outlook

Stefan Kavan, Tina Wittenberger

To be published in Monetary Policy & the Economy Q1–Q2/20.

Mapping financial vulnerability in CESEE: understanding risk-bearing capacities of households is key in times of crisis

Nicolas Albacete, Pirmin Fessler, Maximilian Propst

To be published in Financial Stability Report 39.

Statistical annex