ECONOMIC REVIEW 2/2018

Transcription

1 ECONOMIC REVIEW 2/2018

2 ECONOMIC REVIEW 2/2018 BULGARIAN NATIONAL BANK

3 The BNB quarterly Economic Review presents information and analysis of balance of payments dynamics, monetary and credit aggregates, their link with the development of the real economy, and their bearing on price stability. Processes and trends in the external environment are also analysed since the Bulgarian economy is directly influenced by them. This publication contains also quantitative assessments of the development in major macroeconomic indicators in the short run: inflation, economic growth, exports, imports, trade balance and BoP current account, foreign direct investment, monetary and credit aggregate dynamics. The Economic Review, issue 2/2018 was presented to the BNB Governing Council at its 26 July 2018 meeting. It employs statistical data published up to 9 July The estimates and projections published in this issue should not be regarded as advice or recommendation. Exclusively the information user is liable for any consequences thereof. The Economic Review is available at the BNB website, Research and Publications menu, periodical Publications sub-menu. Please address notes, comments and suggestions to the BNB Economic Research and Projections Directorate at 1000 Sofia, 1, Knyaz Alexander I Square. Bulgarian National Bank, , Knyaz Alexander I Square 1000 Sofia, Bulgaria Tel.: (+359 2) , , Website: This issue includes materials and data received up to 31 July The contents of the BNB Economic Review may be quoted or reproduced without further permission. Due acknowledgement is requested. ISSN (online) Economic Review 2/2018 2

4 CONTENT Summary External Environment... 9 Current Business Situation... 9 International Commodity Prices Financial Flows, Money and Credit External Financial Flows Monetary and Credit Aggregates Interest Rates Financial Flows between the General Government and Other Sectors of the Economy Economic Activity Current Economic Environment Exports and Imports of Goods and Services Behaviour of Firms and Competitiveness Household Behaviour Fiscal Policy Effects on the Economy Inflation Bulgarian National Bank Forecast of Key Macroeconomic Indicators for Forecast Forecast Revisions Risks to the Forecast Research Topics: Methods for assessment of the financial cycle in Bulgaria Bulgarian National Bank

5 Charts Global PMIs... 9 World Trade and External Trade in Selected Regions Inflation Measured through CPI Contribution to the Change in Real GDP in the Euro Area by Country Contribution to the Change in Real GDP in the Euro Area by Component GDP Change and Manufacturing and Services PMIs in the Euro Area Euro Area Unemployment Rate and Employment Growth Euro Area Inflation Rate ECB Interest Rates, EONIA and Excess Liquidity in the Euro Area Banking System EURIBOR Dynamics Contribution to US GDP Growth by Component (Quarterly) US Manufacturing and Services ISM-PMIs and GDP Growth (Quarterly) US Consumer Confidence Indices US Inflation Rate US Unemployment Rate and Number of New Employees in the US Non-farm Sector Manufacturing and Services PMIs and GDP Growth (Quarterly) in China Manufacturing PMI in China and Major Components Fixed Capital Investment (Total) and in Selected Sectors in China China s Inflation Rate House Prices in China China s Foreign Exchange Reserves Brent Crude Oil Price World Crude Oil Supply and Demand (Quarterly) Brent Crude Oil Futures Prices Price Indices of Major Raw Materials and Commodity Groups Current and Capital Account Dynamics and Contributions of Individual Components (on an Annual Basis) Financial Account Dynamics and Contributions of Individual Components (on an Annual Basis) Direct Investment Liabilities by Type of Investment (on an Annual Basis) Gross External Debt as a Share of GDP Annual Rate of Change in М3 and Contribution of the Aggregate by Component Annual Growth of Non-government Sector s Deposits and Contribution by Sector Currency Structure of Deposits of Non-financial Corporations and Households Reserve Money Bank Deposits with the BNB Currency in Circulation Foreign Currency Purchases and Sales between the BNB and Banks (on a Monthly Basis) Annual Growth of the Credit to Non-financial Corporations and Contributions of Individual Types of Loans New Loans to Non-financial Corporations (Monthly Volumes) Annual Growth of Household Credit and Contributions of Individual Types of Loans New Loans to Households (Monthly Volumes) Changes in Demand and in Credit Standards Interbank Money Market Rates (Average Monthly Value) SOFIBOR Yield Curve Interest Rates on New Time Deposits Distribution of Interest Rates on New Household Time Lev Deposits Interest Rates on New Loans to Non-financial Corporations by Currency Interest Rates and Annual Percentage Rate of Charges on New Household Loans Distribution of Interest Rates on New Housing Loans in Euro Ten-Year-and-Six-Month Government Bond Rates on the Primary and Secondary Markets Reference Government Securities Yield Curve in Bulgaria Consolidated Budget Effect on Other Sectors Liquidity (Quarterly) Contribution to GDP Growth by Final Use Component Business Climate and Consumer Confidence Factors Limiting Economic Activities of Corporations Business Climate Indicator Dynamics Dynamics of Exports and Imports of Goods Dynamics of Exports and Imports of Services Dynamics of Exports to EU and non-eu Countries Exports of Goods: Geographical Breakdown Exports of Machines, Vehicles, Appliances, Instruments and Weapons Exports of Mineral Products and Fuels Exports of Base Metals and Related Products Exports of Animal and Plant Products, Food, Drinks and Tobacco Exports of Chemical Products, Plastics and Rubber Annual Change of Services Exports and Contribution by Sub-components Imports of Energy Resources Imports of Raw Materials Imports of Consumer Goods Imports of Investment Goods Annual Change of Services Imports and Contribution by Sub-components Value Added Growth and Contribution by Sector Industrial Turnover Dynamics Construction Production Dynamics and New Building Permits Issued Dynamics of the Nominal Retail Trade Turnover Growth Rate of Employment in the Total Economy and Contribution to Changes in the Number of Employed by Economic Sector Economic Review 2/2018 4

6 Labour Productivity Developments (Value Added per Employee) Compensation per Employee at Current Prices Job Vacancies/Unemployed Ratio and Employee Compensation Unit Labour Costs Gross Operating Surplus at Current Prices Financing Sources Contribution of Changes in Production Factors to GDP Growth Economic Activity and Share of Discouraged Persons Unemployment Rate Employment and Nominal Wage Bill Private Consumption and Consumer Confidence Retail Trade Turnover Household Propensity to Save and Expectations Contributions Major Groups of Revenue to Growth in Total Revenue and Grants, Cumulatively (on an Annual Basis) Contributions Major Tax Groups to Tax Revenue Growth, Cumulatively (on an Annual Basis) Contribution of Major Groups of Expenditure to Total Expenditures Growth, Cumulatively (on an Annual Basis) Contribution of Government Consumption Components to Real GDP Growth Inflation and Contribution of Major Commodity and Services Groups to It Rate of Change in Brent Crude Oil and A95 Petrol Prices Rate of Change in PPI in Industry and HICP Rate of Change in PPI on the Domestic Market and Contribution by Major Sub-sectors Rate of Change in Producer Prices on the Domestic Market by Major Industrial Groupings Rate of Change of Food Price Index Contribution of Major Sub-groups of Unprocessed Food to Overall inflation Core Inflation Contribution of Services and Major Services Sub-groups to Overall Inflation Contribution of Major Non-food Goods (Excluding Energy Products) Sub-groups to Overall Inflation Annual Growth Rate of Nominal Retail Trade in Non-food Goods Excluding Trade in Motor Vehicles and Lubricants Contribution of Major Sub-groups of Goods with Administratively Controlled and Tobacco Prices to Overall Inflation Diffusion Index of Major Goods and Services Groups Expectations of Selling Prices in Industry, Retail Trade and Services in the Next Three Months Share of Firms by Sectors Pointing to Insufficient Demand as a Factor Limiting Their Activity Share of Firms by Sectors Pointing to Sector Competition as a Factor Limiting Their Activity Rate of Change of House Price Index Rate of Change in the House Price Index in Cities of More Than 120,000 Inhabitants over the First Quarter of Construction Production Dynamics and New Residential Building Permits Issued Expected Annual Rate of Change of Real GDP Expected Annual Rate of Change in Inflation at the End of the Period Chart 1. Cyclical Components of the Indicators Used to Obtain an Estimate of the Financial Cycle Chart 2. Indicators Used to Obtain a Measure of the Financial Cycle in Bulgaria Chart 3. Business Cycle, Financial Cycle Potential Measure in Bulgaria FC6 and Alternatives Chart 4. Estimate of the Cyclical Component of Real GDP in Bulgaria Chart 5. Estimate of the Cyclical Component of Real Credit in Bulgaria Chart 6. Estimate of the Cyclical Component of Real House Prices in Bulgaria Tables Projections of the Annual Rate of Change of Euro Area Real GDP Projections of Euro Area Annual Inflation Rate Flows on Major Balance of Payments Accounts Gross External Debt in April Real GDP Growth by Final Use Component Net Exports of Commodity Groups by Use, January March Imports of Commodity Groups by Use, January March Gross Value Added Growth Employment and Income Dynamics Retail Trade Turnover Key Budget Indicators Performance for Key Indicators of the Housing Market Developments in Bulgaria Rates of Change in Major Goods and Services Groups Prices and Contributions of These Groups to Inflation GDP and Inflation Forecast Revisions (21 June 2018 vis-ў-vis 28 December 2017) Forecast of Key Macroeconomic Indicators for Table 1. Indicators Used to Obtain a Measure of the Financial Cycle in Bulgaria Table 2. Groups of Indicators Used to Obtain an Aggregate Measure of the Financial Cycle Table 3. Results of the Principal Component Analysis Table 4. Volatility, Maximum (Peak) and Minimum (Trough) of the Financial Cycle in the Different Indicators of the Financial Cycle in Bulgaria Table 5. Concordance Index Table 6. Concordance Index between the Potential Measures of the Financial Cycle and the Business Cycle in Bulgaria Bulgarian National Bank

7 Abbreviations ABSPP Asset-Backed Securities Purchase Programme APP Asset Purchase Programme APRC Annual percentage rate of charge BIR Base interest rate BOP Balance of Payments BTC Bulgarian Telecommunications Company b.p. basis points CBPP3 Covered Bond Purchase Programme CEECs Central and East European countries CEFTA Central European Free Trade Association CFP Consolidated Fiscal Programme CIF Cost, insurance, freight CNY Chinese Yuan CPI Consumer Price Index DXY an index measuring the exchange rate of the US dollar against the basket of six major currencies EA Employment Agency EC European Commission ECB European Central Bank EIB European Investment Bank EMBI Emerging Markets Bond Index EONIA Euro OverNight Index Average EU European Union EURIBOR Euro Interbank Offered Rate EWRC Energy and Water Regulatory Commission FDI Foreign Direct Investment FOB Free on Board FRS Federal Reserve System GDP Gross Domestic Product GFMS Gold Fields Mineral Services HICP Harmonized Index of Consumer Prices HRW Hard Red Wheat HUF Hungarian forint IEA International Energy Agency IMF International Monetary Fund ISM Institute for Supply Management LEONIA LEv OverNight Index Average LIBOR London Interbank Offered Rate М1 narrow money М2 М1 and quasi-money М3 broad money MF Ministry of Finance MFIs Monetary Financial Institutions mt metric tons NPISHs Non-profit institutions serving households NSI National Statistical Institute OPEC Organization of Petroleum Exporting Countries OTC over-the-counter PBoC People s Bank of China PMI Purchasing Managers Index p.p. percentage points PPP Purchasing Power Parity PSPP Public Sector Purchase Programme RON Romanian new leu SITC Standard International Trade Classification WTI West Texas Intermediate Economic Review 2/2018 6

8 SUMMARY Global PMI dynamics suggests that global economic activity in the January May 2018 period remained at high levels, with year-on-year growth reported in both industrial production and services sectors. Annual growth of world trade remained high despite its temporarily moderating rate in March From the beginning of the year global inflation has varied close to the end-2017 levels. At its monetary policy meeting of 14 June 2018, the ECB Governing Council announced the intention to extend the term of implementing the extended Asset Purchase Programme (APP) from October to December 2018 at a monthly volume of EUR 15 billion. The ECB Governing Council also changed the forward guidance on euro area interest rates, which are expected to remain at their current levels at least through the summer of Based on the reported and expected US labour market and inflation indicators, in June the US Federal Open Market Committee decided to raise by 25 basis points the federal funds rate corridor to per cent. In line with favourable economic developments external demand for Bulgarian goods and services are expected to further increase in the third and fourth quarters of Protectionist measures launched in world trade since early June 2018 and a possible further escalation of international trade conflicts have not been reflected in the baseline scenario of the forecast, thus posing risks to lower than expected global economic growth and weaker external demand for Bulgarian goods and services. Between January and April 2018 the overall current and capital account balance was negative, indicating a higher deficit on the same period of 2017 due mainly to an increase in the trade balance deficit as a result of the worsened terms of trade. In April 2018 the current and capital account balance was positive on an annual basis, the surplus posting a slight decrease on end In the third and fourth quarters of 2018 the current and capital account surplus is expected to further decline on an annual basis driven by the continuously increasing trade deficit and the projected higher deficit on the net primary income item. In the first five months of 2018 the inflow of funds attracted from residents in the banking system remained strong contributing mostly to gradually accelerating annual growth rate of broad money. Bank loans to the non-government sector also tended to accelerate, more strongly pronounced in loans to households. Amid high banking system liquidity and continuous inflow of attracted funds deposit rates remained at relatively low levels and lending rates continued smoothly decreasing. In the third and fourth quarters of 2018 non-government sector deposits are expected to increase at rates similar to those of the first five months of the year. Growth rate of credit to corporations and households is anticipated to continue accelerating moderately driven by the favourable macroeconomic environment and comparatively low lending rates. In the first quarter of 2018 quarterly growth rate of real GDP accelerated and reached 0.9 per cent. Net exports, gross fixed capital formation and private consumption had a positive contribution to growth, while the government consumption contribution was slightly negative. In the first quarter of 2018 the number of employed increased in all economic sectors with the exception of the agricultural one (according to seasonally adjusted data). Due to lower number of employed in the agricultural sector, the number of employees in overall economy remained at a level close to that in the previous quarter. The unemployment rate continued to decline. The nominal compensation per employee increased on a quarterly basis at a rate close to the previous quarter s rate, with all economic sectors contributing to this growth. Short-term economic indicators in the first half of 2018 continued to improve giving mostly positive signals about the economic development. Employment growth, increasing external and domestic demand 7 Summary

9 for goods and services and still low interest rates are expected to continue favouring growth in both private consumption and investment activity of corporations. Given these factors, quarterly real GDP growth is anticipated to vary between 0.7 and 0.8 per cent in the second half of Year-on-year economic growth is expected to moderately accelerate supported mainly by domestic demand. In the first five months of 2018 domestic annual inflation followed an upward trend and reached 2.3 per cent in May, reflecting both the increasing effect of internal factors on consumer price dynamics and the 19.6 per cent rise in international oil prices since early The gradual upward trend in services inflation which started at the close of 2017 was sustained, this group contributing most significantly to the overall inflation in May Goods and services with administratively controlled prices (including tobacco products) were another group with a comparatively high contribution to the annual inflation since the beginning of Energy product inflation accelerated from the end of 2017 due to the strong hike in euro oil prices. In the second half of 2018 annual inflation is expected to increase further driven mainly by the positive contribution of services and energy product prices. Goods and services with administratively controlled prices are also expected to make a strong positive contribution to the overall inflation in the following months. This issue of Economic Review includes the BNB forecast of key macroeconomic indicators for the period. It is based on information published as of 15 June 2018 and employs European Central Bank (ECB), European Commission (EC) and International Monetary Fund (IMF) assumptions of global economic developments and international commodity price dynamics as of 22 May Based on the ECB, EC and IMF forecasts external demand for Bulgarian goods and services is expected to increase tending to slow down its growth rate. The baseline scenario of external demand does not take into account the effects of the US foreign trade measures launched in the beginning of June 2018 and does not include a possible escalation of international trade conflicts during the forecast period. International market prices are expected to increase significantly in In the period non-energy product prices are expected to continue increasing, while petroleum prices are anticipated to slightly decrease due to projected global supply growth of petroleum products. As regards the assumptions about external environment developments, there are risks of lower growth in external demand due to the uncertainty surrounding the assessments of the effects of the US trade policy on global trade developments and the uncertainty about the economic growth in some important for Bulgaria trading partners. Regarding the forecast of crude oil prices there are risks of a hike in the period due to a possible decline in global oil inventories, lower than expected US oil production and worsening of the global geopolitical situation. The improved global economic environment contributed to Bulgaria s stable economic growth. According to the macroeconomic forecast between 2018 and 2020 the economic growth rate will remain stable with a positive contribution of domestic demand (private consumption, private investment and a significant increase in public investment). Private consumption growth is supported by the improved labour market: the unemployment rate is decreasing against the background of an increasing economic activity rate, with relatively high growth of wages being an essential motivating factor for inclusion in the labour force. Increasing domestic demand pushed up the demand for imported goods and services. As a result net exports will have a negative contribution to real GDP growth despite the expected steady growth in exports. These developments will change the current account balance from a surplus to a deficit in the period, with inflows of capital transfers from the EU and foreign direct investments exceeding significantly the deficit. Inflation, influenced largely by international market price dynamics, is expected to accelerate in 2018 and then to slow down and stabilise at 2 per cent annually. Risks to the macroeconomic outlook are related mainly to the uncertainty of assumptions about external environment. An internal factor creating uncertainty to the outlook is the public investment dynamics which is determined by the EU fund absorption rate. Economic Review 2/2018 8

10 1. EXTERNAL ENVIRONMENT Global PMI dynamics suggests that global economic activity in the January May 2018 period remained at high levels, with year-on-year PMI growth reported in both industrial production and services sectors. Annual growth of world trade remained high despite its temporarily moderating rate in March From the beginning of the year global inflation has varied close to the end-2017 levels. At its monetary policy meeting of 14 June 2018, the ECB Governing Council announced the intention to extend the term of implementing the extended Asset Purchase Programme (APP) from October to December 2018 at a monthly volume of EUR 15 billion. The ECB Governing Council also changed the forward guidance on euro area interest rates, which are expected to remain at their current levels at least through the summer of Based on the reported and expected US labour market and inflation indicators, in June the US Federal Open Market Committee decided to raise by 25 basis points the federal funds rate corridor to per cent. In line with favourable economic developments external demand for Bulgarian goods and services are expected to further increase in the third and fourth quarters of Protectionist measures launched in world trade since early June 2018 and a possible further escalation of international trade conflicts have not been reflected in the baseline scenario of the forecast, thus posing risks to lower than expected global economic growth and weaker external demand for Bulgarian goods and services. Current Business Situation In the first five months of 2018 the global PMI remained significantly above the neutral limit of 50, signalling a further increase in the global economic activity. Its average value between January and May 2018 rose from the corresponding period of 2017, reflecting the simultaneous increase in both industrial production and services indices. Global trade growth 1 accelerated between January and April 2018 vis-ў-vis the same period of 2017 despite its temporary moderation in March It was most pronounced in trade of emerging market economies in Asia 2 and the USA, followed by euro area countries, Bulgaria s major trading partners. From the beginning of 2018 global inflation remained close to the end-2017 level, reaching in May 2.1 per cent on an annual basis. 3 Inflation increased from December 2017 in developed countries, while slowing down slightly in developing economies. Global PMIs Source: JP Morgan. 1 CPB Netherlands Bureau for Economic Policy Analysis data as of 23 May CPB Netherlands Bureau for Economic Policy Analysis Classification by country groups. 3 Based on World Bank data as of 2 July External Environment

11 In line with favourable economic developments, external demand for Bulgarian goods and services is expected to further increase in the third and fourth quarters of Protectionist measures launched in world trade since the beginning of June 2018 and a possible further escalation of international trade conflicts have not been reflected in the baseline forecast scenario, posing risks to lower than expected global economic growth and weaker external demand for Bulgarian goods and services. International prices of energy and non-energy commodities are anticipated to rise further on an annual basis in dollars 4 and in euro over the third and fourth quarters of Given the projected higher growth of energy than of nonenergy product prices and the larger share of petroleum products in goods imports of Bulgaria than in goods exports, developments in international commodity group prices are expected to retain the unfavourable terms of trade in Bulgaria over the projection horizon. World Trade and External Trade in Selected Regions (annual rate of change in volumes, per cent) Source: CPB Netherlands Bureau for Economic Policy Analysis. Inflation Measured through CPI (per cent, annual rate of change, seasonally adjusted data) Euro Area In the first quarter of 2018 euro area real GDP growth decelerated to 0.4 per cent on a quarterly basis, from 0.7 per cent in the fourth quarter of Economic growth was observed in all euro area countries, except for Estonia, but GDP growth in Germany, France and Italy moderated to 0.3, 0.2 and 0.3 per cent (from 0.6, 0.7 and 0.4 per cent in the previous quarter), while remaining unchanged in Spain (0.7 per cent). Private consumption, inventories and investments were the GDP components with positive contributions to euro area GDP growth on a quarterly basis. Net exports made a negative contribution and government consumption had a neutral effect on real GDP growth. Note: The World Bank measures the change of CPI in individual groups as a weighted average of CPI changes in the countries of the group. Real GDP based on purchasing power parity is used to calculate country weights. Groups include only World Bank Member States classified by the World Bank as developing and developed countries. Source: the World Bank. Contribution to the Change in Real GDP in the Euro Area by Country (per cent; percentage points; quarter-on-quarter) Leading economic indicators, including PMIs and EC indices, decreased in the second quarter of 2018, though remaining at comparatively high levels signalling retention of a relatively high economic activity in the euro area. Labour market indicators continued to improve slowly, with the unemployment rate in the euro area reaching 8.4 per cent in April and May against 8.6 per cent on average in the first 4 Hereinafter referred to as the US dollar. Sources: Eurostat and BNB calculations. Economic Review 2/

12 quarter. The expected level of unemployment over the next 12 months included in the EC consumer confidence index recorded historical lows, signalling further improving consumer expectations of labour market developments. Contribution to the Change in Real GDP in the Euro Area by Component (per cent; percentage points; quarter-on-quarter) Over the second quarter of 2018 euro area inflation accelerated, with the HICP energy component contributing most strongly. In May the annual HICP change was 1.9 per cent, from 1.3 per cent in April, with Eurostat's provisional assessment showing annual inflation of 2.0 per cent in June. In June the ECB revised downwards its euro area real GDP growth projections for 2018 (by -0.3 percentage points to 2.1 per cent) leaving its growth expectations for 2019 and 2020 at 1.9 and 1.7 per cent, respectively. Risks to euro area economic growth prospects are assessed as broadly balanced by the ECB. Risks of achieving lower than expected economic growth stem primarily from global factors such as increased protectionist pressures and financial market fluctuations. Source: Eurostat. GDP Change and Manufacturing and Services PMIs in the Euro Area (per cent) The ECB increased its euro area inflation projections for 2018 and 2019 by 0.3 percentage points to 1.7 per cent for the two years, leaving unchanged at 1.7 per cent its projection for This revision reflects mainly the increased oil prices. The ECB expects wage growth acceleration in the medium term to be an essential factor for the increase in consumer prices. At its monetary policy meeting of 14 June 2018 the ECB Governing Council left unchanged Sources: Eurostat, Markit. Projections of the Annual Rate of Change of Euro Area Real GDP (per cent) Euro Area Unemployment Rate and Employment Growth (per cent) (per cent) Date of release latest previous latest previous latest previous ECB VI EC V Sources: the ECB, the EC. Projections of Euro Area Annual Inflation Rate (per cent) Institution Institution Date of release latest previous latest previous latest previous ECB VI EC V Sources: the ECB, the EC. Source: Eurostat. 11 External Environment

13 interest rate levels, while announcing its intention to extend the Asset Purchase Programme (APP) duration from October to December 2018 and reduce the monthly pace of purchases to EUR 15 billion (half the present volume of EUR 30 billion) until the end of December 2018 and then end net asset purchases. It was considered prudent to leave the implementation of net asset purchases still conditional on incoming data, with the extended APP remaining part of monetary policy instruments. At this meeting, the ECB Governing Council changed also the forward guidance on euro area interest rates, which are expected to remain at their current levels at least through the summer of The ECB reconfirmed its forward guidance on reinvesting principal payments from maturing securities of the Eurosystem portfolio, indicating, however, that this issue would be discussed at one of the Council's forthcoming meetings. As of 29 June 2018 the cumulative amount of APP purchases was EUR billion, of which EUR billion under the Public Sector Purchase Programme (PSPP), EUR billion under the Covered Bond Purchase Programme 3 (CBPP3), EUR 27.4 billion under the Asset-backed Securities Purchase Programme (ABSPP) and EUR billion under the Corporate Sector Purchase Programme (CSPP). On 29 June 2018 excess liquidity in the euro area banking system decreased to EUR billion, from EUR billion by end-march, with the Eurosystem balance sheet figure reaching EUR trillion. In the second quarter of 2018 average EONIA (the effective overnight reference rate) remained at per cent. Overnight interbank deposit trading in the euro area decreased, with its daily average coming to EUR 4.1 billion between 2 April and 29 June 2018 (from EUR 5.1 billion in the first quarter of 2018). EURIBOR unsecured deposit rates remained unchanged or increased slightly across maturity sectors. As of 29 June 2018 one-month (-0.37 per cent) and six-month deposit rates (-0.27 per cent) remained at their end-march levels, whereas twelve-month rates rose to per cent (+1 basis point on the end of March). Euro Area Inflation Rate (per cent, on an annual basis) Source: Eurostat. ECB Interest Rates, EONIA and Excess Liquidity in the Euro Area Banking System (per cent) Note: Average EONIA data for the month. Source: the ECB. EURIBOR Dynamics (basis points) Source: the ECB. (EUR billion) Economic Review 2/

14 The United States In the first quarter of 2018 US real GDP growth decelerated to 0.5 per cent on a quarterly basis against 0.7 per cent in the fourth quarter of 2017, most largely due to the weaker growth in household consumption, consumption and public sector investments, as well as decreased residential investment. At the same time, business investment growth accelerated over the period, while net exports contribution to real GDP growth was neutral. In the second quarter of 2018 US leading economic indicators showed hesitant but mostly positive movements. In May ISM-PMI manufacturing and services indices increased, though remaining below historical highs achieved in early 2018 in the context of the tax reform approved by the US Congress at the end of Consumer confidence indicators showed similar changes in the second quarter, remaining also close to the high levels of early Following the moderation in the first quarter, monthly indicators measuring US household consumption and retail sales suggest a substantial acceleration in their growth over the April to May period. Latest construction data show an increase in economic activity compared to the beginning of the year. Overall, end-june data signal accelerating growth in US real GDP over the second quarter. In May the annual rate of change in the price index of personal consumption expenditure (PCE) rose to 2.3 per cent, from 2.0 per cent in the previous two months. The annual growth rate of the core index (excluding food and fuel) rose to 2.0 per cent in May against 1.8 per cent in April and March. May 2018 CPI data point to an accelerated annual rate of both overall inflation (to 2.8 per cent, from 2.5 per cent in April) and core inflation (to 2.2 per cent, from 2.1 per cent in April). June data on expectations of consumer price changes one to five years ahead, which are part of the consumer confidence index of the University of Michigan, showed an increase in both long-term and short-term inflation expectations. US labour market indicators underwent no essential changes in the second quarter of The average quarterly number of new employees in the US non-farm sector declined Contribution to US GDP Growth by Component (Quarterly) (per cent; percentage points) Source: Bureau of Economic Analysis. US Manufacturing and Services ISM-PMIs and GDP Growth (Quarterly) (per cent) Source: Institute for Supply Management, Bureau of Economic Analysis. US Consumer Confidence Indices (2000 = 100) Source: The Conference Board. 13 External Environment

15 insignificantly from the first quarter s level. After declining in May to 3.8 per cent (the lowest level since April 2000), the US unemployment rate increased to 4.0 per cent in June. Based on reported and expected US labour market and inflation indicators, at its 12 and 13 June meeting, the Federal Open Market Committee (FOMC) took a decision to raise by 25 basis points the federal funds rate corridor to per cent. Simultaneously, the Committee made some changes to its communications by removing the forward-guidance language stating that the federal funds rate is likely to remain, for some time, below levels that are expected to prevail in the longer run. After the meeting, changes were also announced in the monetary policy framework involving a technical adjustment to the IOER rate that would place it at a level 5 basis points below the top of the FOMC s target range for the federal funds rate. Positive expectations of FOMC members regarding developments in the economy and labour market in the medium term led them to revise upwards their forecast of the expected reference rates in 2018 and 2019 incorporating four instead of three hikes in 2018 and three instead of two hikes in At the same time, Committee members assessments of the target rate at the end of 2020 and its equilibrium level remained unchanged at 3.38 and 2.88 per cent. At the conference after the FOMC meeting, Jerome H. Powell, the Chairman of the Board of Governors of the Federal Reserve System, pointed to the lack of any indications so far that changes in the US trade policy since early 2018 has imported US macroeconomic data, though he designated them as a potential risk factor. US Inflation Rate (per cent; on an annual basis) Note: Inflation is measured by the personal consumption expenditure deflator. Source: Bureau of Economic Analysis. US Unemployment Rate and Number of New Employees in the US Non-farm Sector (per cent) Source: Bureau of Labour Statistics. (new employees, thousand) Manufacturing and Services PMIs and GDP Growth (Quarterly) in China (per cent) China In the second quarter of 2018 services and manufacturing PMIs in China did not change significantly, signalling retention of economic growth at a level close to the first quarter s level. In June the new export orders index, which is a sub-component of manufacturing PMI, declined to 49.8 points, from 51.2 points in May. The decrease below the neutral limit of 50 points is likely to reflect the uncertainty surrounding unfavourable effects of US trade policy measures (involving duties on certain imports Sources: National Bureau of Statistics of China and China Federation of Logistics and Purchasing. Economic Review 2/

16 Manufacturing PMI in China and Major Components Fixed Capital Investment (Total) and in Selected Sectors in China (per cent; on an annual basis) Source: China Federation of Logistics and Purchasing. Source: National Bureau of Statistics of China. China s Inflation Rate (per cent, on an annual basis) House Prices in China (per cent; on an annual basis) Source: National Bureau of Statistics of China. from China) on country s economic activity since early Note: Group 1 includes the four largest cities: Beijing, Shenzhen, Guangzhou, and Shanghai. Group 2 includes the capitals of the rest provinces. Source: National Bureau of Statistics of China. In April and May the annual growth rate of industrial output in China remained at relatively stable levels, while total investment in the economy decreased its growth as a result mainly of the weaker increase in infrastructure investment. Over the second quarter Chinese CPI dynamics was largely driven by the slowing growth in food prices in April and May following their temporary rise in February when the Chinese New Year was celebrated. In May annual growth in producer prices accelerated to 4.1 per cent, from 3.4 per cent in April and 3.1 per cent in March, reflecting primarily commodity price increases over the recent months. In the second quarter annual growth in house prices in China slowed down due to their declines in large cit- China s Foreign Exchange Reserves (USD billion) Source: People s Bank of China. (USD billion) 15 External Environment

17 ies (group 1), remaining almost unchanged from last two quarters in group 2 smaller cities. At the end of the second quarter, the People s Bank of China announced its intention to cut the minimum reserve rate by 50 basis points for the five largest and 12 smaller banks to 15.5 per cent. This measure aims to increase the lending capacity of these institutions for small enterprises by some CNY 700 billion (USD 108 billion). In addition, banks are being encouraged by the authorities to use funds in debt-for-equity swap transactions with large mostly state-owned enterprises. Brent Crude Oil Price (USD per barrel) (EUR per barrel) Over the second quarter China s international reserves decreased by USD 32 billion to USD 3.11 trillion. According to the State Administration of Foreign Exchange (SAFE), this decline was almost entirely due to financial asset revaluation as a result of the US dollar appreciation, with no capital outflows observed over the period. Source: the World Bank. World Crude Oil Supply and Demand (Quarterly) (million barrels per day) International Commodity Prices Crude Oil In the first five months of 2018 international crude oil prices soared strongly on an annual basis, reaching on average USD 69.8 per barrel (Brent). The year-on-year increase in demand for petroleum products since early 2018 led to a decline in global inventories and a rise in oil prices. Production cuts agreed between the OPEC and other major oil producers and temporary production disruptions in the USA and Venezuela supported additionally the increase in oil prices. At the same time, the risk of impeding Iran s and Venezuela s exports as a result of the US foreign policy also boosted oil prices. *Including IAE latest available data on crude oil demand in March Source: the IAE. Brent Crude Oil Futures Prices (average monthly price of contract, USD per barrel) In the first five months of 2018 International Energy Agency (IEA) data showed an increase in global crude oil demand compared to the corresponding period of 2017, reflecting global economic growth and unfavourable weather conditions in both Europe and the USA in early In June market oil price expectations for the third and fourth quarters of 2018, embedded in futures prices, increased vis-à-vis March expectations, despite the announced measures Source: JP Morgan. Economic Review 2/

18 to raise oil production after the OPEC meeting on 22 June This reflected market expectations of enhanced demand for oil and oil products exceeding supply due to the high economic activity rate and increasing global industrial output. Market expectations of crude oil prices point to a level around USD 74 per barrel in the third and fourth quarters of Price Indices of Major Raw Materials and Commodity Groups (2013 = 100) Metals Commodity and Food Prices In the first five months of 2018 metal prices continued to increase on an annual basis, following the 2017 tendency. Broken by subcomponent, nickel prices showed the highest annual growth resulting from the increase in stainless steel production, supply disruptions and decreased global inventories. Strong annual growth was also recorded in aluminium prices, driven by China s measures to reduce air pollution and the April sanctions imposed by the United States to the world's second largest aluminium producer located in Russia. Despite the annual growth in dollar metal prices, in the first five months euro prices declined following the appreciation of the euro against the US dollar. The price of copper, which is of key importance to Bulgarian goods exports, continued to increase on an annual basis in both dollars and euro. This growth was driven by stronger demand for copper ore worldwide and enhanced global industrial production. The risk of temporary production disruptions in the Chilean largest copper mine in renewing labour contracts also contributed to the increase in copper prices. Copper Food Wheat Metal prices are expected to increase in the third and fourth quarters of 2018 in both dollar and euro. 5 Possible escalation of international trade conflicts that would increase the risk of lower than targeted global economic growth may consequently result in a weaker rise or even decline in metal prices. 5 The euro price forecast is based on the technical assumption about the euro/dollar exchange rate. It is fixed at the average value for the last ten days on 15 June 2018, the date when the assumptions were made. In line with the technical assumption, the euro/dollar exchange rate will increase in the third and fourth quarters of 2018 compared with the corresponding periods of the previous year. Sources: the ECB and BNB calculations. 17 External Environment

19 For the period January to May 2018 on average, the food price index in dollars remained broadly unchanged from the corresponding period of 2017, with divergent developments observed across components. Prices of certain products, such as sugar, coffee and tea, declined on an annual basis. Prices of wheat, which is essential for the Bulgarian exports of goods, increased in US dollars on an annual basis. This was attributable to the expected lower yields at the global level, with market expectations in Russia, the EU, India and Mexico being lowered due to unfavourable weather conditions. Though in the first five months the dollar food price stayed close to the level of the same period of 2017, it declined in euro due to the euro appreciation against the dollar. Food price projections in US dollar terms point to a year-on-year increase in the third and fourth quarters of 2018, albeit an expected slight decline in euro terms, as a result of the euro appreciation against the dollar. 6 If market expectations of international fuel, commodity and food price dynamics materialise, the terms of trade in Bulgaria are expected to remain unfavourable in the third and fourth quarters of Please refer to footnote 5. Economic Review 2/

20 2. FINANCIAL FLOWS, MONEY AND CREDIT Between January and April 2018 the overall current and capital account balance was negative, indicating a higher deficit on the same period of 2017 due mainly to an increase in the trade balance deficit, as a result of the worsened terms of trade. In April 2018 the current and capital account balance was positive, the surplus posting a slight decrease on end In the third and fourth quarters of 2018 the current and capital account surplus is expected to further decline driven by the continuously increasing trade deficit and the projected higher deficit on the net primary income item. In the first five months of 2018 the inflow of funds attracted from residents in the banking system remained strong contributing mostly to gradually accelerating annual growth rate of broad money. Bank loans to the non-government sector also tended to accelerate, more strongly pronounced in loans to households. Amid high banking system liquidity and continuous inflow of attracted funds deposit rates remained at relatively low levels, while lending rates continued to follow a decreasing trend. In the third and fourth quarters of 2018 non-government sector deposits are expected to increase at rates similar to those of the first five months of the year. Growth in credit to corporations and households is anticipated to continue accelerating moderately, driven by the favourable macroeconomic environment and comparatively low lending rates. External Financial Flows In the January April 2018 period the balance on the current and capital account was negative at EUR million, with the deficit posting an increase compared to the corresponding period of The main factor behind this was the higher trade balance deficit 8, which posted a rise due to worsened terms of trade for Bulgaria over the review period. Lower surpluses on the net secondary income item and the capital account, reflecting the lower amount of transfers under EU programmes and Bulgaria s larger contribution to the EU, contributed to a smaller extent to the increase in the overall negative current and capital account balance. The higher surplus on the net services trade item and lower deficit on the net primary deficit income item partly limited the increase in the current and capital account deficit. The higher surplus on the net services trade item in the January April 2018 period compared to the corresponding period of 2017 reflected the good performance of exported tourist and transport services, while imports decreased on an annual basis. Between Flows on Major Balance of Payments Accounts (EUR million) On an annual basis as of April 2018 Current account Trade balance Services, net Primary income, net Secondary income, net Capital account Financial account Change in reserves (per cent of GDP) On an annual basis as of April 2018 Current account Trade balance Services, net Primary income, net Secondary income, net Capital account Financial account Change in reserves The analysis of balance of payments flows employs information from its analytical reporting. Preliminary data. 8 For a more detailed analysis of foreign trade flows, see The Exports and Imports of Goods and Services Section in Chapter Financial Flows, Money and Credit

21 January and April 2018 the deficit under the net primary income decreased compared to the corresponding period of 2017, due to lower paid dividends and distributed profits, and reinvested earnings. 9 Current and Capital Account Dynamics and Contributions of Individual Components (on an Annual Basis) (EUR million) Movements of flows under individual current and capital account items in the first four months of 2018 almost matched those in As a result in April 2018 the total surplus accumulated on the current and capital account for the last 12 months posted a slight decrease on December 2017, which was mostly attributable to the larger trade deficit. In the third and fourth quarters of 2018 the current and capital account surplus year-on-year is expected to retain its downward path 10 driven by further growing trade deficit and a projected increase in the net primary income deficit due to growth in gross operating surplus in the economy in the last quarters and a possible increase in reinvested earnings. At the same time, the improved absorption of funds under EU-funded operational programmes for the programming period will be reflected in increased EU fund inflows and growing annual capital account surplus as a share of GDP compared to the end of 2017 and early Between January and April 2018 the financial account was positive, reflecting mainly the increase in foreign assets by EUR million and to a lesser degree by the decrease in foreign liabilities of Bulgarian residents by EUR 94.7 million. Banks transactions contributed most significantly to the increased assets on the financial account. 11 In addition, other sectors activities 12, increasing mainly their foreign portfolio investments, also contributed to Bulgarian residents foreign assets growth in the January April 2018 period. Between January and April 2018 the general government and other sectors reported a decrease in foreign liabilities on the financial account, while banks increased their foreign liabilities. As a result of described flow 9 Preliminary data subject to revision. Revisions usually show an increase in outflows to non-residents. 10 It should read the balance of the account for the last four quarters. 11 More detailed analysis of bank operations is provided in the Monetary and Credit Aggregates Section in Chapter Sectors other than central banks, other monetary financial institutions and general government. Note: For each quarter, the chart shows item balances accumulated in the last four quarters. February 2018 data on the chart are accumulated for the last 12 months. Financial Account Dynamics and Contributions of Individual Components (on an Annual Basis) (EUR million) Notes: The Other item includes Other Investments, Net, and Financial Derivatives (Other than Reserves) and Employee Stock Options, Net. For each quarter, the chart shows item balances accumulated in the last four quarters. February 2018 data on the chart are accumulated for the last 12 months. Economic Review 2/

22 movements in financial account components the balance accumulated for the last 12 months was positive, posting an increase from December 2017 due to the stronger rise in foreign assets against foreign liabilities. Between January and April 2018 foreign direct investment liabilities decreased compared to the corresponding period of 2017 to EUR million 13 (against EUR million in the January April 2017 period). The highest foreign direct investment inflows were from the Netherlands, Germany and Russia. As a result of lower inflows in early 2018, in April 2018 inflows on direct investment liabilities accumulated for the last 12 months declined from December Reflecting net current, capital and financial account flows, Bulgaria s gross international reserves 14 decreased by EUR million between January and April In April 2018 the international reserve coverage of the average nominal imports of goods and non-factor services remained high (8.1 months against 8.7 months in December 2017). In April 2018 Bulgaria s gross external debt declined from the end of 2017 for all economic sectors except for intercompany lending. The share of long-term debt in Bulgaria s total gross external debt accounted for 76.2 per cent in April 2018 and remained almost unchanged from December The new external debt taken up by Bulgarian residents in the January April 2018 period decreased compared with the corresponding period of 2017 due almost entirely to the lower new external debt of the banking sector. Conditions for servicing foreign obligations of Bulgarian residents remained favourable. Between January and April 2018 period interest payments on external debt also posted a decline on the corresponding period of Preliminary data subject to revision. Revisions usually show an increase in foreign direct investment liabilities. 14 Valuation adjustments and price revaluation excluded. Direct Investment Liabilities by Type of Investment (on an Annual Basis) (EUR million) Note: For each quarter, the chart shows item balances accumulated in the last four quarters. April 2018 data on the chart are accumulated for the last 12 months. Gross External Debt as a Share of GDP (per cent of GDP) Gross External Debt in April 2018 (EUR million) Amount Since December 2017 Change For the last 12 months General government Central bank Banks Other sectors FDI intercompany loans Total Note: More information about individual institutional sectors is available on the BNB website, Statistics, External Sector/Gross External Debt. 21 Financial Flows, Money and Credit

23 Monetary and Credit Aggregates In the first five months of 2018 the inflow of attracted funds in the banking system remained high and contributed to the moderate acceleration of annual growth in the broad monetary aggregate M3 which amounted to 8.8 per cent in May (7.7 per cent at the end of 2017). Broad money dynamics continued to be driven by overnight deposits and, to a lesser extent, by currency in circulation outside MFIs, and quasimoney made a slight positive contribution. Major factors affecting the M3 component dynamics again included the relatively high savings rate in the economy and corporate and household preferences for easier access to savings amid low deposit rates. Non-government sector s deposits continued to account for a major part of attracted funds in the banking system, their share reaching 87.3 per cent on average for the first five months of the year. In May the annual growth of nongovernment sector s deposits 15 in the banking system amounted to 6.1 per cent (6.2 per cent at the end of 2017), with deposits of households and non-financial corporations having major positive contribution and deposits from financial corporations contributing negatively. Household deposits accelerated moderately from the end of 2017 to 6.5 per cent on an annual basis amounting to BGN 48.6 billion. Deposits of nonfinancial corporations posted a 12.5 per cent annual increase to BGN 21.9 billion. Financial corporations deposits, which usually fluctuate significantly, tended to further decrease. In May the decline accounted for 30.5 per cent on an annual basis. In terms of the currency structure, lev-denominated deposits retained their high share. At the end of May they accounted for 61.7 per cent of all non-government sector deposits. Annual Rate of Change in М3 and Contribution of the Aggregate by Component (per cent, percentage points) Note: The marketable instruments component is not shown on the chart due to its insignificant contribution to broad money growth. Annual Growth of Non-government Sector s Deposits and Contribution by Sector (per cent, percentage points) Currency Structure of Deposits of Non-financial Corporations and Households (BGN million) The effective implicit rate of minimum required reserves 16 remained at levels similar to the previous year s average levels, reflecting the 15 Non-government sector s deposits include deposits of households, non-financial corporations and financial corporations. Deposits of households and non-financial corporations comprise the largest share of all non-government sector s deposits (96.0 per cent on average for the last 12 months in May 2018) and therefore the analysis is focused on these two sectors. 16 According to Article 3 of BNB Ordinance No 21 on the Minimum Required Reserves Maintained with the Bulgarian National Bank by Banks, the minimum required reserve rate on funds attracted from residents is 10 per cent of the deposit base, from non-residents 5 per cent and from the state and local government budgets 0 per cent. Economic Review 2/

24 retained high inflow of attracted funds from residents in the banking system in the first half of In June it was 9.37 per cent (9.41 per cent on average for 2017) with bank deposits with the BNB comprising 8.36 percentage points in the fulfilment of minimum required reserves, and the remaining 1.01 percentage point forming recognised cash balances. Following a prevailingly downward trend at end-2017 and early 2018 reserve money posted positive annual growth at 4.1 per cent at the end of the second quarter of This dynamics reflected mainly the slower decline of bank reserves (-3.1 per cent in June against per cent on average in the first quarter of 2018) and, to a lesser extent, moderately accelerating annual growth in currency in circulation (10.9 per cent in June against 10.4 per cent on average in the first quarter of 2018). The decline in bank deposits with the BNB on an annual basis pertained to the decrease in banks excess reserves (by 22.4 per cent on an annual basis at the end of June). Concurrently, the annual growth of minimum required reserves accounted for 6.3 per cent. The dynamics of banks excess reserves on an annual basis continued to reflect the effect of the changed methodology for setting interest rates on accounts with the BNB, effective from October In the second quarter of 2018 banks excess reserves remained at levels close to the average ones for the first quarter amounting to BGN 2.9 billion. In June 2018 the excess funds on banks minimum reserve accounts with the BNB over the required minimum of reserve assets under Ordinance No 21 was 39.3 per cent on an average daily basis. The year-onyear decrease in bank deposits with the BNB following excess reserve declines was partially offset by the increased bank funds in TARGET2. In the first six months of 2018 banknotes and coins in circulation continued to increase at comparatively high rates. This dynamics was mostly driven by growth in private consumption amid improving labour market conditions, low deposit interest rates and savers preferences to hold cash. Reserve Money (BGN million) Bank Deposits with the BNB (BGN million) Currency in Circulation (BGN million) (per cent) (per cent) 17 The change became effective following a decision of the BNB Governing Council of 4 October Changes in the methodology refer to the interest rate on excess reserves of banks with the BNB which was reduced by 20 basis points below the ECB deposit facility rate. 23 Financial Flows, Money and Credit

25 Under currency board arrangements, foreign currency trade with the BNB is the main tool to manage bank liquid resources. In June 2018 the BNB net sales of euro to banks amounted to EUR 463 billion on an annual basis. Foreign Currency Purchases and Sales between the BNB and Banks (on a Monthly Basis) (EUR million) (EUR million) In May banks assets grew 5.6 per cent on an annual basis, with claims on the non-government sector and their foreign assets having the largest positive contribution. Concurrently, banks excess reserves with the BNB and claims on the general government sector reported a decline. Lower claims on the general government sector was due to the negative issue of government securities over the review period. Annual growth in credit to households and nonfinancial corporations 18 continued to accelerate in the first five months of 2018, accounting for 5.1 per cent in May (3.3 per cent by end-2017). Favourable macroeconomic environment and comparatively low lending rates weigh on these developments. Consumer loans contributed most significantly to growth due to the inclusion of a new reporting unit in the scope of monetary statistics from April Notes: Net means currency purchased minus currency sold by the BNB. Data refer to all bank transactions in foreign currency including liquidity management operations related to the transfer of own funds from lev accounts with the BNB to own accounts with the BNB in euro and vice versa. In May 2018 loans to non-financial corporations rose by 2.9 per cent on an annual basis (1.6 per cent in December 2017). Overdrafts continued to contribute most strongly to growth of loans to non-financial corporations, and loans, excluding overdraft, reported low positive annual growth. In May 2018 new 20 corporate loans reached higher levels than those reported at the end of The share of newly extended corporate 18 Loans represent a major part of bank claims on the nongovernment sector with a share of 98 per cent on average for the last 12 months as of May 2018, and the analysis was therefore focused on them. In addition to loans, claims include also repurchase agreements, securities other than shares, and shares and other equity instruments. Nongovernment sector s deposits, in turn, include loans to households, loans to non-financial corporations and loans to financial corporations. The share of loans to households and non-financial corporations in total loans to the non-government sector accounted for 96 per cent on average in the last 12 months as of May 2018 and therefore developments in these two sectors are addressed. 19 As of April 2018 other monetary financial institutions sector includes BNP Paribas Personal Finance S.A. Bulgaria branch with balances reclassified from other financial intermediaries sector. The reclassification results from the transformation through merger of BNP Paribas Personal Finance EAD specialised in lending and reported to-date in other financial intermediaries sector for the purposes of monetary statistics. 20 The terms new and newly extended hereinafter referred to as the statistical category new business. 21 Based on 12-month moving average. Annual Growth of the Credit to Non-financial Corporations and Contributions of Individual Types of Loans (per cent, percentage points) Economic Review 2/

26 loans in levs decreased by 2.1 percentage points on average for the last 12 months to 52.4 per cent compared with December 2017, while new corporate loans in euro increased by 2.3 percentage points and their share reached 45.9 per cent. New corporate loans in US dollars comprised 1.7 per cent of all new corporate loans (1.9 per cent at the close of 2017). In the first five months of 2018 the growth of household loans accelerated and in May accounted for 8.8 per cent (6.0 per cent at the end of 2017). Consumer loans contributed most substantially (15.0 per cent annual growth), with the accelerated growth rate strongly affected by the new reporting unit included in the scope of monetary statistics from April Housing loans (9.8 per cent annual growth in May) also contributed significantly to higher household loans, while overdraft to households continued to decline on an annual basis. Other loans to households fell by 21.6 per cent on an annual basis. Loans extended under the National Programme for Multi-Occupant Residential Building Energy Efficiency (energy efficiency programme) reported an annual decline due to the high base in the same period of the previous year and loan repayments under this programme launched by the government in the middle of Newly extended consumer and housing loans to households retained their upward pattern over the first five months of Summarised weighted results of the bank lending survey 23 in the first quarter of 2018 suggest tightening of bank standards 24 in approving credit applications of corporations and easing the standards for crediting households. At the same time, banks continued to ease lending conditions 25 for corporations in terms of interest rates, interest rate spreads, and fees and commissions. Strong competition in the banking sector was the main reason for easier New Loans to Non-financial Corporations (Monthly Volumes) (BGN million) Note: For data characterised by a significant volatility, additionally moving averages computed for suitably selected periods are provided, in order to smooth the fluctuations in corresponding time series and present the trends in their development. Annual Growth of Household Credit and Contributions of Individual Types of Loans (per cent, percentage points) New Loans to Households (Monthly Volumes) (BGN million) 22 See footnote Summarised results of the bank lending survey are presented through weighting bank responses by their market share in the relevant credit segment. 24 Credit standards are understood as internal bank guidelines or criteria for loan approvals established prior to negotiating the terms of extended loans. Credit standards determine the type of the loan and collateral considered admissible by banks, taking into account specific priorities by sector, etc. Credit standards specify also all relevant conditions to be met by a borrower. 25 Credit conditions typically involve the reference interest rate surcharge, the loan amount, conditions for its utilisation and other conditions, fees and commissions, collateral or guarantees to be provided by a borrower. 25 Financial Flows, Money and Credit

27 lending (lending standards and conditions) to corporations and households over the quarter. Sustained ample liquidity of the banking system, continuously increasing volume and low cost of attracted funds along with lowery yield of alternative investment prompted easing of credit standards and conditions for corporations. In addition to bank competition, lowery risk assessment as a result of overall macroeconomic environment improvement, favourable prospects in the housing market, increased borrowers solvency, and lower collateral risk also contributed to easier lending standards and conditions for households. Banks expectations in the April June 2018 period point to some tightening of standards for crediting large corporations, sustaining of the standards for small and medium-sized enterprises and insignificant easing of the standards for households. As regards demand, banks continued to report stronger demand for corporate and consumer loans, while demand for housing loans slightly decreased from the fourth quarter of Enhanced demand for corporate loans in the first quarter of 2018 reflected the demand for investment funds and low interest rates. Major factors encouraging household demand for consumer loans included low interest rates and funds required for purchasing current consumption and durable goods. Bank expectations for the April June 2018 period show that demand for funds by corporations and households will increase, with the strongest growth anticipated in consumer loans. In the context of continuously improving economic activity and comparatively high savings rate in the economy, in the third and fourth quarters of 2018 deposits of the nongovernment sector are expected to increase at rates similar to those in the first half of the year. Growth in credit to corporations and households is anticipated to continue accelerating moderately driven by the favourable macroeconomic environment and comparatively low lending rates. Over the projection horizon the contribution of the loans extended under the National Programme for Energy Efficiency to the overall growth of household loans may be expected to decrease further, assuming continued government repayments on these loans and decreased volumes of new loans under this programme. Factors likely to limit additionally credit growth in the third and fourth quarters are maintenance of bank policies to write-off non-performing Changes in Demand and in Credit Standards (balance of opinions; percentage points) а) corporate loans b) consumer loans c) housing loans Notes: As regards credit standards, charts present banks balance of opinions defined as a difference in percentage points between the percentage of banks responding tightened' ( considerably and somewhat ), and the percentage of banks responding eased ( considerably and somewhat ). As regards credit demand, the balance of opinions is defined in percentage points as a difference between the percentage of banks responding increased' ( considerably and somewhat ) and the percentage of banks responding reduced ( considerably and somewhat ). All opinions are weighted by the bank s market share in the relevant credit segment. Economic Review 2/

28 loans from balance sheets and possible sales of credit portfolios. Interest Rates Interbank Money Market Rates (Average Monthly Value) а) overnight deposits (per cent, percentage points) LEONIA Plus index 26 remained relatively stable in the first half of 2018 and stood at per cent (down 2 basis points on December 2017) in June. LEONIA Plus/EONIA spread also remained unchanged and stood at -14 basis points in June. In the first half of the year volumes traded in the interbank lev market, dominated mainly by overnight transactions, remained relatively high. In the first six months of 2018 interbank money market rates 27 based on SOFIBOR 28 posted a decrease in all maturities moving downwards the yield curve. In June 2018 unsecured three-month lev deposit rates declined compared to December 2017 by 10 basis points to per cent, six-month rates fell by 8 basis points to 0.08 per cent and unsecured 12-month lev deposit rates dropped by 14 basis points to 0.49 per cent. Note: With effect from 1 July 2017, LEONIA Plus index replaces LEONIA. LEONIA Plus monthly values are calculated as an arithmetic average for those days when unsecured lev overnight lending transactions are concluded in the interbank market. Sources: the BNB, the ECB. b) three-month deposit interest rates (per cent, percentage points) The stronger decline in the interbank money market quotations in Bulgaria compared to that in the euro area resulted in a fall in the spreads between money market quotations in Bulgaria vis-ў-vis the euro area. As compared to end spreads on three-month and 12-month unsecured deposit rates were lower by 10 basis points and 15 basis point reaching 28 and 67 basis points respectively. Interest rates on time deposits remained at low levels in the first five months of The strong inflow of attracted funds and high liquidity in the banking system further favoured the retention of low deposit rates. In May the average weighted interest rate on new time deposits of non-financial corporations and households was 0.3 per cent, remaining broadly unchanged on previous year average. Interest rates on non-financial corporations deposits posted an annual growth of 17 basis points to 0.4 per cent since year-start, which was offset by a fall Note: Average monthly SOFIBOR for three-month deposits. Sources: the BNB, the ECB. SOFIBOR Yield Curve (per cent) 26 LEONIA Plus (LEv OverNight Interest Average Plus) is a reference rate of unsecured overnight deposit transactions in Bulgarian levs on the interbank market. 27 Interbank money market rates based on SOFIBOR are indicative in nature with no real transactions concluded. 28 According to the BNB Governing Council decision of 16 March 2017 the BNB discontinued the activities in relation to the calculation and publication of the SOFIBID and SOFIBOR reference rates as from 1 July For further details, see the BNB press release of 16 March 2017 ( PR_ _EN Note: Average monthly SOFIBOR for the relevant maturity. 27 Financial Flows, Money and Credit

29 in household deposit rates (down 4 basis points to 0.2 per cent). Interest rates on new time deposits in levs and euro declined, and those on deposits in US dollars raised on December In May 2018 the breakdown by bank of average weighted interest rates on new time lev deposits of households showed an increase on March in the number of banks with average interest rates within ranges from 0.25 to 0.50 per cent and from 0.50 to 0.75 per cent at the expense of banks with interest rates ranging up to 0.25 per cent and from 0.75 and 1.00 per cent. Variation of interest rates offered by banks on new time lev deposits of households remained low, continuing to decline slightly. The smooth downward trend in interest rates on new loans was sustained in the first five months of Results from Bank Lending Survey show that major factors favouring these developments were enhanced competition and high liquidity in the banking sector, increasing volume and low cost of attracted resources, and limited alternatives for high-yield investment. In May 2018 interest rates on new loans to nonfinancial corporations decreased on December 2017 by 2 basis points for lev-denominated loans, 46 basis points for euro-denominated loans and 99 basis points for dollar-denominated loans. In May interest rates on new loans to non-financial corporations in the three major currencies were 3.8 per cent in levs, 3.4 per cent in euro and 3.8 per cent in US dollars 29. The annual percentage rate of charge (APRC) on housing loans to households retained the downward tendency, while it posted a slight increase on consumer loans. In May 2018 the APRC on housing loans fell by 25 basis points to 4.2 per cent compared to December 2017, while that on consumer loans went up by 7 basis points to 10.2 per cent 30. The APRC decline in housing loans was mainly driven by the downward dynamics of the interest rate component of costs while the APRC increase in consumer loans was due to the increase in the implicit rate of non-interest service charges. Interest Rates on New Time Deposits a) by sector (per cent) Notes: The average interest rate is calculated for all sectors, maturities and currencies weighted by the relevant volumes of new deposits. Average deposit rates for non-financial corporations and households are based on interest rates for all maturities and currencies weighted by relevant volumes of new business. b) by currency (per cent) Note: The average interest rate by currency is based on rates for all sectors and maturities weighted by relevant volumes of new business. Distribution of Interest Rates on New Household Time Lev Deposits (frequency number of banks) The bank breakdown of average weighted interest rates on new housing euro loans showed 29 The above values of interest rates on loans to non-financial corporations are weighted averages on a 12-month basis. 30 Values indicating the APRC and interest rates are weighted averages of the interest rates on household consumer and housing loans on a 12-month basis. Economic Review 2/

30 ranges of up to 4.0 per cent in most banks. In May 2018 the standard deviation decreased by 4 basis points to 0.80 per cent compared with March Over the same period the average interest rate on new housing euro loans decreased by 14 basis points to 3.72 per cent. In the second half of 2018 lending rates on new time deposits and loans are expected to remain broadly unchanged compared to the levels in the first half of the year. Factors that will support the retention of low interest rates do not change and include the high inflow of attracted funds in the banking system, its high liquidity and the competition in the sector. An external factor that also helps keeping interest rates on loans and deposits close to the current level relates to the market expectations of retaining the ECB key interest rates unchanged over the forecast horizon. In the second quarter of 2018 the Ministry of Finance held no government bond auctions in the primary market. The lack of supply of newly issued securities and the high banking system liquidity contributed to keeping the secondary market yield at levels similar to those of the first quarter end. In June the long-term interest rate used for assessing the degree of convergence posted a slight decline from the March level reaching 0.99 per cent. The yield of Bulgarian Eurobonds traded in international capital markets also remained broadly unchanged from the end of the first quarter despite the higher yield of some EU countries government securities in May and June, reflecting the increased political uncertainty in Italy. Euro area debt market fluctuations were behind lowering market participants risk appetite and higher demand for lowrisk assets as German government securities, which led to increased spreads between yields of Bulgarian and German government securities. The increase was most strongly pronounced in the long-term end of the maturity curve. In the second half of 2018 Bulgarian government bond yields are expected to be further driven by external and internal factors. In case purchase volumes decrease as planned from October 2018 until the final completion of ECB s Asset Purchase Programme (APP) in December 31, external factors could contribute to higher yields of the Bulgarian government securities. At the same time, the effect of the internal factors 31 For details, see the External Environment Section in Chapter 1. Interest Rates on New Loans to Non-financial Corporations by Currency (per cent) Interest Rates and Annual Percentage Rate of Charges on New Household Loans а) consumer loans (per cent) b) housing loans (per cent) (percentage points) (percentage points) Notes: Interest rates in all maturities and currencies are weighted by the relevant volumes of new loans for a 12-month period. Implicit rate is the difference between the APRC and the relevant interest rates and reflects the approximate proportion of all non-interest service charges on loans (including fees and commissions). 29 Financial Flows, Money and Credit

31 Distribution of Interest Rates on New Housing Loans in Euro (frequency number of banks) Ten-Year-and-Six-Month Government Bond Rates on the Primary and Secondary Markets (per cent) supporting the retention of Bulgarian securities yields at levels close to the current ones is expected to continue. Reference Government Securities Yield Curve in Bulgaria (per cent) Financial Flows between the General Government and Other Sectors of the Economy Government revenue and expenditure policies and budget financing operations affect the allocation of liquidity among economic sectors. In the first quarter of 2018 the financial sector (excluding the BNB) was beneficiary of a great amount of liquidity from the budget (5.9 per cent of GDP for the quarter) as a result mainly of the negative net issue of government securities for the period (BGN million). In the first quarter of the year interests paid to residents and non-residents in the amount of BGN 249 million and BGN 130 million 32 also contributed to redirecting liquid funds from the budget to the financial sector and the external sector. Outflows to the external sector resulting from the EU budget contribution costs (BGN million) were lower than the inflows to the budget, reflecting mainly the funds reimbursed by the EU under the EU Common Agricultural Policy (BGN million) and grants under EU programmes (BGN million). Consequently, net financial flows from the external sector to the budget were positive accounting for 2.0 per cent of GDP in the quarter. The fiscal policy had a negative effect on non-government non-bank sector s liquidity in the first quarter accounting for 2.6 per cent of the Notes: The reference yield curve of Bulgarian government securities is based on own calculations under the extended version of Nelson Siegel Svensson model (1994). The yield change refers to the previous quarter. The chart employs daily yield data on Bulgarian government securities issued and traded in international capital markets, published in the MF Central Government Debt and Guarantees Monthly Bulletin. Sources: The MF, BNB calculations. 32 The assessment is based on the MF Central Government Debt and Guarantees Monthly Bulletin as of 31 March 2018 concerning Eurobonds held by residents. Economic Review 2/

32 GDP mainly due to the primary surplus reported in this period (BGN million). Positive net financial flows from the non-government non-bank sector and the external sector failed to offset the liquid funds redirected from the budget to the financial sector, resulting in a decrease of government deposit with the BNB by BGN million (1.2 per cent of GDP in the quarter) compared to the end of December Based on preliminary monthly data as of mid- July 2018 an assessment can be made for the second quarter of 2018 that financial flows from the external sector, the financial sector and the non-government non-bank sector to the budget will be positive, though small in size. Liquid funds redirected to the budget are reflected in the growth in government deposit with the BNB by BGN 313 million compared to the end of the first quarter. Consolidated Budget Effect on Other Sectors Liquidity (Quarterly) (as a percentage of GDP, per cent) Sources: the MF, the BNB. Developments in financial flows generated in the process of external and internal economic factor interaction contributed to the decrease in gross international foreign exchange reserves. By the end of June 2018 the market value of international reserves (including valuation adjustments and price revaluations), an asset on the BNB Issue Department balance sheet, was EUR 23.5 billion (BGN 46.0 billion), posting a decrease by EUR 0.2 billion on end According to the currency board principles, the decline in international reserves corresponds to the drop in the balance sheet value of the Issue Department liabilities. Decreased liabilityes by the end of the first quarter of 2018 compared to December 2017 mainly resulted from lower bank excess reserves with the BNB which was partly offset by an increase in the deposit of other depositors as a result of the annual contributions in the total amount of BGN million made in May in the Deposit Insurance Fund and Bank Resolution Fund. In the third and fourth quarters of 2018 bank reserves with the BNB are expected to remain broadly unchanged compared to the end of June. Government deposit dynamics will be determined mainly by the consolidated fiscal programme performance and the schedule for reimbursement of funds by the EU under the EU Common Agricultural Policy. In the second half of the year the annual growth in currency in circulation is expected to remain at a relatively high level, reflecting the favourable developments in private consumption and retention of interest rates on deposits at current low levels. 31 Financial Flows, Money and Credit

33 3. ЕCONOMIC ACTIVITY In the first quarter of 2018 quarterly growth of real GDP accelerated on the previous quarter and accounted for 0.9 per cent. Net exports, gross fixed capital formation and private consumption had a positive contribution to growth, while government consumption contribution was slightly negative. In the first quarter of 2018 the number of employed increased in all economic sectors with the exception of the agricultural one (according to seasonally adjusted data). Due to lower number of employed in the agricultural sector, the number of employees in overall economy remained at a level close to that in the previous quarter. The unemployment rate continued to decline. The nominal compensation per employee increased on a quarterly basis at a rate close to the previous quarter rate, with all economic sectors contributing to this growth. Short-term economic indicators in the first half of 2018 continued to improve giving mostly positive signals about the economic development. Employment growth, increasing external and internal demand for goods and services and still low interest rates are expected to continue favouring growth in both private consumption and investment activity of corporations. Given these trends, quarterly GDP growth is anticipated to vary between 0.7 and 0.8 per cents in the second half of Year-on-year economic growth is expected to moderately accelerate supported mainly by domestic demand. Current Economic Environment In the first quarter of 2018 quarterly growth of real GDP accelerated from the previous quarter to 0.9 per cent (seasonally adjusted NSI national account data). Net exports contributed most to this growth, reflecting the quarterly decline in goods imports, while goods and services exports grew in line with the strong external demand. Lower goods imports over the first quarter reflected both slowing private consumption growth and significant declines in supply for domestic refineries in March and April following the planned renovation of production capacities 33. The previous year s trend to a strong positive contribution of domestic demand components to the quarterly growth in real GDP was sustained in the beginning of Over the review quarter private consumption continued rising, albeit at a slower rate compared to the end of Consumption growth was underpinned by the persistent long-term trend toward improving consumer confidence, increasing employment and households labour income as well as of Contribution to GDP Growth by Final Use Component (per cent, percentage points; quarter-on-quarter) Note: Non-additive data due to direct chain-linked and seasonal adjustment of GDP and its components; the contribution of the change in inventories has not been included. Sources: the NSI, BNB calculations. 33 NSI data on foreign trade and production and supply of energy products. For details, see Exports and Imports of Goods and Services Section in Chapter 3. Economic Review 2/

34 low interest rates which continue to stimulate the demand for consumer loans. Moderation in the quarterly growth of final household consumption expenditure reflected partially the weaker wage bill growth in real terms 34 compared to the end of 2017, as well as allocation of a higher proportion of income towards assets, such as real estate and deposits. 35 In the first quarter of 2018 gross fixed capital formation increased significantly from the previous quarter, contributing positively to the domestic demand growth. BNB estimates 36 indicate that investment growth is driven by the increase in both public and private investment. Public investment growth reflects mainly the increase in investment under EU funding programmes. Business Climate and Consumer Confidence (per cent) Sources: the NSI, BNB calculations. (per cent) In the second quarter the business climate indicator tracking business sentiment went up on average vis-ў-vis the first quarter of 2018, reflecting largely more optimistic assessments of corporations about the current and expected business situation 37. Across economic sectors an improvement in the business climate indicator was recorded in construction, industry and services, while retail trade registered a slight drop. In the second quarter of 2018 again an increasingly smaller proportion of firms identified the uncertain economic environment and intra industry competition as major factors for limiting their activities, while the share of firms pointing to the labour shortage as a factor hampering their activity continued to increase. Business optimism about the future economic situation gives grounds to expect that economic activity will increase further in the coming quarters. The composite business climate indicator exceeding its historical average value, coupled with the broad-based sound developments in domestic demand components and labour market improvements suggest that the economy is in an expansion phase. Factors Limiting Economic Activities of Corporations (relative share of all corporations) Note: Average for the period calculated as sector-weighted (industry, construction, trade and services). Sources: the NSI, BNB calculations. 34 Wage bill has been converted from nominal into real terms using the private consumption deflator. 35 The analysis employs Eurostat data on residential buildings investment and BNB data on household deposits in the banking system. 36 The analysis employs data from quarterly non-financial account of the general government sector and reports on the implementation of the Consolidated Fiscal Programme, published by NSI and MF respectively. 37 Seasonally adjusted data. 33 Economic Activity

35 Consumer confidence increased in the second quarter, driven by the positive consumer expectations about the overall economic situation in the next 12 months and improved consumer expectations about unemployment and their financial position. Retention of optimistic consumer assessments is also confirmed by the continuous strong growth of retail sales in food and non-food segments 38. In the third and fourth quarters of 2018 growing labour demand by firms, rising nominal and real household labour income and optimistic consumer sentiment are expected to contribute to the increase in final consumption expenditure by households. Over the second half of 2018 real GDP growth is projected to be largely driven by domestic demand following the accelerated growth in private and government consumption and persistently high growth rates in fixed capital investment since early Strong increase in public investment under the influence of accelerated growth in EU co-financed investment and a recovery of national investments after their low performance in 2017 are anticipated. It could be expected that firms will continue to expand their investment activities in the third and fourth quarters encouraged by the improving economic indicators, high capacity utilisation in the first two quarters, the expected growth in domestic and external demand, retained favourable financing conditions and improved corporate profitability. Exports of goods and services are projected to increase further in the third and fourth quarters in line with strengthening external demand, though at slower rates than imports, with the result that net exports will have a negative contribution to the growth. Given the final use component dynamics, quarterly real GDP growth is expected to vary between 0.7 and 0.8 per cent in the second half of At the same time annual real GDP growth is projected to follow a gradual upward trend. The uncertainty surrounding assumptions on external factor and external environment effects gives rise to risks of lower than projected growth in real GDP. Another factor for achieving Business Climate Indicator Dynamics Notes: The chart compares the business climate level (the vertical axis) and its quarterly change (the horizontal axis). The business climate series has been transformed using the HP filter (with a parameter l = 100) in order to eliminate short-term fluctuations after which it has been standardised. The four chart grids allows to distinguish the four phases of the business cycle. 38 For details on developments in nominal retail trade volumes, see the Household Behaviour Section in Chapter 3. Economic Review 2/

36 Real GDP Growth by Final Use Component (per cent, quarter-on-quarter; real rate, seasonally adjusted data) I II III IV I II III IV I II III IV I Consumption incl. Household consumption Government final consumption expenditure Collective consumption Gross fixed capital formation Exports of goods and non-factor services Imports of goods and non-factor services GDP Source: the NSI. weaker than expected economic growth is the possibility of lower than projected growth in government investment. Exports and Imports of Goods and Services In the first quarter of 2018 nominal exports of goods increased by 2.4 per cent on an annual basis and nominal imports of goods by 5.7 per cent. 39 According to non-seasonally adjusted GDP data, in the first quarter of 2018 real growth in both exports and imports of goods was 4.1 per cent on an annual basis, though nominal values show faster growth in imports compared to exports. In the first quarter of the year the terms of trade in Bulgaria worsened which can be traced through export and import deflator dynamics (national account data). In the context of equal real terms growth in goods exports and imports, the worsened terms of trade over the first quarter were the major factor for the increased trade deficit. 40 Dynamics of Exports and Imports of Goods (EUR million) 39 Foreign trade data. 40 According to foreign trade data under the Standard International Trade Classification (SITC) in the first quarter of the year the export deflator is positive and import deflator is negative which suggests an improvement in terms of trade. This conclusion conflicts the assessment about developments in the terms of trade based on the national accounts data, which show that the change in the export deflator on an annual basis is negative and that of import deflator positive. Insofar the dynamics of export and import deflators from the national accounts match to a large extent the dynamics of international prices, this part of the analysis is based on national accounts data. 35 Economic Activity

37 Balance of payments data show that in April 2018 nominal goods exports and imports continued to rise on an annual basis, with growth in exports outpacing that in imports. In the third and fourth quarters of 2018 real goods exports and imports are expected to continue increasing year on year, with growth in imports outpacing that in exports. Real exports of goods will reflect the expected increase in external demand. Concurrently, real imports dynamics will be determined by both projected acceleration in annual investment growth and continued increases in private consumption in Bulgaria. If the projected growth in real volumes and prices of goods exports and imports materialises, nominal foreign trade flows are expected to increase on an annual basis, more pronounced in imports. Balance of payments data show that between January and April services exports rose year on year, while services imports declined from the same period of There is great uncertainty on whether services exports and imports dynamics was due mainly to changes in real volumes rather than price components since a significant divergence in nominal values exists between balance of payments and national accounts statistics. 41 This divergence is more visible in imports of services which increased over the first quarter by 16.6 per cent in nominal terms according to non-seasonally adjusted national account GDP data, while balance of payments data show a decline of 5.5 per cent on January March National account data for the first quarter show that nominal exports and imports of services went up on an annual basis, with both real trade volume growth and price dynamics contributing positively to this. In the third and fourth quarters of 2018 annual price increases in services exports and imports and the projected rise in real trading volumes amid growing external and domestic demand are expected to lead to year-on-year growth in services exports and imports. Dynamics of Exports and Imports of Services (EUR million) Dynamics of Exports to EU and non-eu Countries (EUR million) Exports of Goods: Geographical Breakdown In the first quarter of 2018 goods exports to EU Member States increased by 13.1 per cent on an annual basis, reflecting the improved economic activity within the EU. Exports to 41 Preliminary data subject to revision. Economic Review 2/

38 Germany, Italy and Spain contributed most significantly to this growth. Exports of animal and plant products followed by machines 42 and metals 43 contributed most significantly to the export dynamics in EU Member States by commodity group. Exports to non-eu countries declined by 17.5 per cent on an annual basis, driven mainly by the weaker exports of energy products 44 to Turkey. In the first quarter of 2018 the share of exported goods to EU Member States in total exports of goods increased year-on-year to 71.9 per cent. Exports of metals had the largest contribution to the nominal annual growth of goods exports under the Combined Nomenclature in the first quarter. A year-on-year decline in nominal terms was recorded only in exports of energy products. Nominal exports of goods continued to follow the observed trend towards growth in the last years, exceeding the world trade growth rate. Consequently, Bulgaria continued to increase its market share in world trade in 2017, reflecting largely the improving competitiveness of the Bulgarian economy. Against the backdrop of rising unit labour costs over the recent years, it can be concluded that the improving competitiveness was largely attributable to non-price factors. Exports of machines rose substantially in the first quarter of 2018, remaining one of the majors factors behind the overall goods export growth a trend also observed in previous years. The increase was due to the exports to EU countries, while exports to non-eu countries declined on the January compared to March 2017 period. Developments in machines exports over the first quarter of 2018 was driven by higher exported volumes, while prices declined 45. Reflecting the growing external demand, the upward trend in machines exports Net Exports of Commodity Groups by Use, January March 2018 Exports of Machines, Vehicles, Appliances, Instruments and Weapons (EUR million) Exports of Mineral Products and Fuels (EUR million) Balance (EUR million) Change* (EUR million) Growth of exports** (per cent) Growth of imports** (per cent) Consumer goods Raw materials Investment goods Energy resources Other exports Total * Balance change on the same period of previous year. ** Exports and imports growth for the period on an annual basis. 42 In this chapter it should read the machines, vehicles, appliances, instruments and weapons group under the Combined Nomenclature. 43 In this chapter, it should read the base metals and their products group under the Combined Nomenclature. 44 In this chapter, it should read mineral products and fuels group under the Combined Nomenclature. 45 The quarterly export deflators for the machines, equipment and vehicles group under SITC were used to estimate the price effect in the first quarter of Economic Activity

39 is expected to continue in the third and fourth quarters of In the first quarter nominal exports of energy products declined substantially as a result almost entirely of lower physical volumes. 46 The major reason was the strong decline in production of oil products due to the planned renovation of production capacities. 47 A decline was reported in exports to both EU and non-eu countries. With completion of the renovation work and given the annual increase in petroleum product prices in dollars and euro, declines recorded in nominal exports of mineral products are expected to be reversed in the third and fourth quarters of Exports of base metals made the largest contribution to the overall goods export growth in the first quarter. By geographical breakdown, metal exports to EU Member States rose significantly year on year, while those to non-eu countries decreased vis-ў-vis January to March The nominal year-on-year change over the first quarter was primarily due to the higher real export volumes rather than metal price rises. 48 In the third and fourth quarters of 2018 exports of metals are expected to continue growing in nominal terms on an annual basis, reflecting the increasing real export volumes and projected annual growth in international metal prices in euro and US dollars. Nominal exports of animal and plant products increased on an annual basis over the first quarter of 2018, with cereals exports contributing most significantly to this. Increased exports of animal and plant products was attributable to the higher real export volumes, whereas export prices declined on an annual basis. 49 Exports to EU Member States recorded growth, while those to non-eu countries declined compared with the first quarter of This group s exports are expected to drop on an annual basis in the Exports of Base Metals and Related Products (EUR million) Exports of Animal and Plant Products, Food, Drinks and Tobacco (EUR million) Exports of Chemical Products, Plastics and Rubber (EUR million) 46 The quarterly export deflators for the mineral fuels, oils and related products group under the SITC were used to estimate the price effect in the first quarter of NSI data on production and supplies of oil and oil products of 27 June The quarterly export deflators for the precious and other non-ferrous metals under SITC were used to estimate the price effect in the first quarter of The quarterly export deflators for the food and live animals group under SITC were used to estimate the price effect in the first quarter of Economic Review 2/

40 third and fourth quarters of 2018 as a result of lower cereals yields in Bulgaria and other EU countries. 50 Annual Change of Services Exports and Contribution by Sub-component (per cent, percentage points) In the first quarter of 2018 exports of chemical products 51 rose nominally on an annual basis, with plastic and inorganic products contributing most strongly to this growth. An increase in exports was reported in exports to both EU and non-eu countries. Nominal growth of exports in this group was due to both higher prices and exported larger physical volumes. 52 In the third and fourth quarters of 2018 stronger external demand is expected to further contribute to the increase in exports of chemical products on an annual basis. According to balance of payments data between January and April 2018 exports of services increased by 8.3 per cent annually with exports of services from the travel sub-item contributing most to this growth, followed by transport subitem. Between January and April 2018 year-onyear earnings from visits of foreign nationals to Bulgaria increased by 10.4 per cent. According to NSI data in the January April 2018 period visits of foreign tourists to Bulgaria increased by 8.8 per cent compared to the corresponding period of 2017, mainly due to visits of Romanian tourists. In the first quarter of 2018 nominal imports of goods increased on an annual basis. By use, growth was reported in all sub-groups with the exception of energy resources. Raw materials, followed by investment goods contributed most substantially to growth of nominal imports. In the first quarter of 2018 imports of goods from EU Member States rose by 8.5 per cent on the first quarter of Imports from Germany and Italy contributed most strongly to growth. Imports from non-eu countries rose by 2.7 per cent on an annual basis. Imports from China and Turkey made the largest positive contribution to this dynamics, while imports of goods from Russia reported the strongest decline. In the first quarter of 2018 the share of imports 50 For further details see Wheat Outlook of USDA, June 2018, p The chemical products, plastics and rubber group under the Combined Nomenclature. 52 The quarterly export deflators for the chemicals and chemical products group under SITC were used to estimate the price effect in the first quarter of Imports of Energy Resources (EUR million) Imports of Commodity Groups by Use, January March 2018 EUR million Change* (EUR million) Growth** (per cent) Contribution** (percentage points) Consumer goods Raw materials Investment goods Energy resources Other imports Total imports (cif) * Change on the corresponding period of previous year. ** Growth/contribution to total import growth over the period on an annual basis. Economic Activity

41 from the EU slightly declined to 52.4 per cent (against 51.0 per cent in 2017). Between January and March 2018 nominal imports of energy resources decreased significantly compared to the first quarter of 2017 due to lower real imported volumes 53. The deceleration of real imports in the group was most strongly pronounced in March, reflecting a significant fall in production and supply of oil products, due to a planned renovation of production capacities in Bulgaria 54. At the same time in the January March 2018 period import prices of energy resources increased on an annual basis, as a result of the price increase in petroleum products in international markets. Based on expectations of a continuous annual increase in international petroleum product prices and rising domestic demand in Bulgaria in the second half of 2018, nominal imports of energy products are projected to boost on an annual basis. The raw materials group had the largest contribution to annual growth of total imports in the first quarter of This reflected both higher imported physical volumes and the increase in import prices. 55 Higher imports of the raw materials group was supported by the improved year-on-year economic activity in industry 56. Imports of ores followed by imports of iron and steel had the largest positive contribution in nominal terms. In the third and fourth quarters of 2018 nominal imports of raw materials are expected to continue increasing due to the anticipated rise in international prices of these goods. Imports of Raw Materials (EUR million) Imports of Consumer Goods (EUR million) 53 The quarterly import deflators for the mineral fuels, oils and related products group under the SITC were used to estimate the price effect in the first and second quarters of See footnote In the first quarter nominal exports of energy products declined substantially as a result almost entirely of lower physical volumes. The major reason was the strong decline in production of oil products due to the planned renovation of production capacities. A decline was reported in exports to both EU and non-eu countries. With completion of the renovation work and given the annual increase in petroleum product prices in dollars and euro, declines recorded in nominal exports of mineral products are expected to be reversed in the third and fourth quarters of The quarterly import deflators of the groups of food and live animals, inedible (crude) materials (excluding fuels); chemicals and chemical products and manufactured goods classified chiefly by materials under the SITC were used to estimate the price effect in the first quarter of More detailed information is provided in the Behaviour of Firms and Competitiveness Section in Chapter 3. Imports of Investment Goods (EUR million) Economic Review 2/

42 Between January and March 2017 year-on-year imports of consumer goods also grew. This growth was in line with growing private consumption according to non-seasonally adjusted GDP data. As in 2017 the food, drinks and tobacco sub-group contributed most significantly to growth of consumer good imports. Given the forecast for accelerating the annual rate of private consumption real growth in the second half of 2018, nominal imports in the group are expected to increase further on an annual basis. Between January and March 2018 year-on-year imports of investment goods increased significantly with their dynamics corresponding to nominal and real growth of investment according to non-seasonally adjusted GDP data. Imports of machines and equipment were the major contributor to growth of imported investment goods. The assessment of price changes suggests that the higher imports in the group reflected the increase in traded physical volumes, while import prices dropped on an annual basis 57. Our forecasts for the third and fourth quarters of 2018 point to a continuous upward trend in nominal exports of investment goods given the expected increase in investments in Bulgaria compared with the corresponding quarters of According to balance of payments data, in the July April 2018 period imports of services decreased by 1.9 per cent year on year. Despite the fall in total imports of services Bulgarians residents foreign travel and transport services spending rose in the first quarter of 2018 compared with the first quarter of According to NSI data, Bulgarians visits abroad in the first quarter of 2018 increased by 6.5 per cent on an annual basis, with the visits to Turkey and Romania having the most significant contribution to this growth. Annual Change of Services Imports and Contribution by Sub-component (per cent, percentage points) Value Added Growth and Contribution by Sector (per cent, percentage points, quarter-on-quarter; seasonally adjusted data) Note: Non-additive data on contributions due to direct seasonal adjustment of value added and its components. Sources: the NSI, BNB calculations. Behaviour of Firms and Competitiveness In the first quarter of 2018 real value added growth in the total economy matched the previous quarter s growth rate, reporting a 0.6 per cent increase on a quarterly basis. The services sector made the major positive contribution to 57 The quarterly import deflators for the machines, equipment and vehicles group under SITC were used to estimate the price effect in the first quarter of Economic Activity

43 value added growth, followed by manufacturing and construction, while the agricultural sector had a negative contribution. Value added in industry went up 1.2 per cent in the first quarter on a quarterly basis driven by the increased economic activity in both industry and construction. Similar developments in early 2018 were also observed in construction production index at constant prices according to NSI short-term business statistics, while the industrial turnover index reported a decline. The fall in industrial turnover by geographical breakdown (at constant prices) in the first quarter of 2018 was due mainly to international markets, with energy products playing a major role to this effect. This dynamics was in line with the reported decline in exports and imports of these goods 58. The quarter-on-quarter increase in the construction production in the first quarter of 2018 was entirely due to growth of building construction 59 and was driven by the increasing house prices from the beginning of , while civil/engineering construction 61 had a negative contribution. Services value added continued to accelerate its quarterly growth rate, increasing by 1.2 per cent in the first quarter of 2018 from 1.0 per cent in the previous quarter. Trade, transportation and storage, accommodation and food service activities 62 and real estate activities had the most sizeable positive contribution to its growth. Value added growth in trade was accompanied by the continuous upward trend in the retail trade turnover at constant prices, suggesting enhanced demand for both food and non-food goods in the first quarter of Value added growth in the other services subsectors was in line with the preliminary NSI data on turnover index in services in the first quarter, Industrial Turnover Dynamics (per cent, percentage points, quarter-on-quarter; seasonally adjusted data) Note: Data for the first quarter of 2018 are averaged for the April May period. Sources: the NSI, BNB calculations. Construction Production Dynamics and New Building Permits Issued (number of buildings; (2010=100; seasonally adjusted data) seasonally adjusted data) Note: Data for the second quarter of 2018 are averaged for the April May period. Sources: the NSI, BNB calculations. Dynamics of the Nominal Retail Trade Turnover (per cent, percentage points, quarter-on-quarter; seasonally adjusted data) 58 For a more detailed analysis of goods exports, see Chapter 3, Exports and Imports of Goods and Services Section. 59 Building construction includes construction of residential and non-residential buildings. 60 For further information on housing price dynamics, see the Inflation Section in Chapter Civil and engineering construction covers infrastructure construction of roads, bridges, pipes, power lines, gas pipelines, telecommunications and other construction works. 62 Wholesale and retail trade; repair of motor vehicles and motorcycles; transportation and storage; accommodation and food service activities by economic activity groupings (A 10). Note: Non-additive data on contributions due to direct seasonal adjustment of the total amount and its components. Data for the second quarter of 2018 are averaged for the April May period. Sources: the NSI, BNB calculations. Economic Review 2/

44 Growth Rate of Employment in the Total Economy and Contribution to Changes in the Number of Employed by Economic Sector (per cent, percentage points, quarter-on-quarter; seasonally adjusted data) Labour Productivity Developments (Value Added per Employee) (per cent, quarter-on-quarter; seasonally adjusted data) Note: Non-additive data on contributions due to direct seasonal adjustment of the total amount and its components. Sources: the NSI, BNB calculations. Sources: the NSI, BNB calculations. Compensation per Employee at Current Prices (per cent, quarter-on-quarter; seasonally adjusted data) suggesting a further increase in sales revenue generated by transportation, storage and posts, information and communication, telecommunications and other business services. The NSI short-term business statistics data on production and turnover in industry, and turnover in trade and trade in April and May 2018 gave mostly positive signs of the change in gross value added in the respective sub-sectors in the second quarter of On average, over the April May 2018 period the nominal industrial turnover increased on the previous quarter. This increase reflected domestic market real growth, which was partly offset by a fall in the real component of international markets. The price components on domestic and international markets had a slight positive contribution to growth of industrial turnover at current prices. Between April and May 2018 retail trade turnover increased further both at constant and current prices on a quarterly basis. Concurrently, the construction production index posted a decline on the previous quarter. Despite the improvement of the economic activity measured by the value added dynamics, in the first quarter of 2018 employment in the overall economy remained at a level close to that in the previous quarter according to seasonally adjusted NSI national account data. Sources: the NSI, BNB calculations. The Job Vacancies/Unemployed Ratio and Employee Compensation (per cent, job vacancies/unemployed; seasonally adjusted data) (per cent on corresponding quarter of the previous year; seasonally adjusted data) Note: The quotient between job vacancies over the month and unemployed is an indicator of the labour market conditions. The increase in the indicator is interpreted as a limited potential of labour supply to satisfy demand as a smaller number of unemployed compete for one job. Sources: NSI Labour Force Survey, Employment Agency, BNB calculations. 43 Economic Activity

45 Higher employment on the previous quarter was reported in the services sector and in industry to a lesser extent, while the agricultural sector posted a decline. Employment growth across services and industry sub-sectors was broadbased, with real estate activities alone reporting a decline on a quarterly basis. As a result of positive developments in the economic activity of construction, employment in this sub-sector of industry exhibited quarter-on-quarter growth for a third consecutive quarter. Firms optimistic expectations of future business climate in Bulgaria, anticipated higher external and domestic demand for goods and services along with gradually increasing number of job vacancies are indicative of stronger labour demand by firms. These factors are expected to retain their favourable effect on employment recording a further increase in third and fourth quarters of Labour productivity accelerated its growth rate on a quarterly basis in the first quarter of 2018 due to a faster growth rate of value added than that of employment. The improved labour productivity was favoured by private investment growth, reflecting the increased share of managers who state that the bulk of investment in 2018 will be used for expansion and rationalization of production 63. All main sectors contributed positively to labour productivity growth. Given the sustainable domestic and external demand economic activity is expected to increase further in the second half of 2018, which will stimulate employment growth. As a result the growth rate of productivity is expected to remain at a level similar to that in the first quarter of In the first quarter of 2018 compensation per employee in nominal terms continued to pick up on a quarterly basis driven by the increasing labour demand and weaker labour supply 64. Other factors contributing to this effect relate to the increase in the minimum wage, minimum social security income and pension contributions, and higher wages in the public sector 65. Compensation per employee growth Unit Labour Costs (2010 = 100; seasonally adjusted data) Sources: the NSI, BNB calculations. Gross Operating Surplus at Current Prices (BGN billion; seasonally adjusted data) Sources: the NSI, BNB calculations. Financing Sources* (million BGN, quarter-on-quarter volume change) 63 The analysis employs data derived from EC investment surveys. 64 Labour supply is measured by the labour force. 65 For details see the Fiscal Policy Effects on the Economy Section in Chapter 3. * Sources of financing other than gross operating surplus. Economic Review 2/

46 in agriculture and industry moderated on the previous quarter, while a slight acceleration in services was observed in line with higher labour productivity in the sector. The compensation per employee is expected to increase further quarter-on-quarter, owing mainly to the private sector under the influence of the above factors. Over the quarter labour productivity in the total economy increased at a slower pace compared with the growth of the nominal compensation per employee in the same period, leading to an increase in unit labour costs in nominal terms, while in real terms they remained at a level similar to that of the previous month. This development was consistent with the acceleration of the economic activity in the country and the progress in achieving nominal and real convergence. Contribution of Changes in Production Factors to GDP Growth (per cent, percentage points, quarter-on-quarter; seasonally adjusted data) Sources: the NSI, BNB calculations. Seasonally adjusted data show that the gross operating surplus remained unchanged in the first quarter of 2018 as a result of its decline in the industry, while the other major sub-sectors reported profit growth. Funding attracted by firms from sources other than the gross oper- Gross Value Added Growth (per cent, quarter-on-quarter; real rate, seasonally adjusted data) I II III IV I II III IV I II III IV I II III IV I Agriculture, forestry and fishing Mining and quarrying; manufacturing; electricity, gas, steam and air conditioning supply; water supply; sewerage, waste management and remediation activities Construction Wholesale and retail trade; repair of motor vehicles and motorcycles; transportation and storage; accommodation and food service activities Information and communication Financial and insurance activities Real estate activities Professional, scientific and technical activities; administrative and support service activities Public administration, education, human health and social work activities Culture, sport and entertainment; other activity; activities of households as employers; non-identified activities of households producing goods and services for own use; activities of extraterritorial organisations and bodies Gross value added, total for the economy Source: the NSI. 45 Economic Activity

47 ating surplus suggested an increase in the attracted funds in all sectors of the economy except for construction. This growth was mainly due to the increased foreign direct investment 66 attracted in Bulgaria and loans to non-financial corporations over the quarter. GDP growth decomposition by production factor suggests that total factor productivity was the main contributor to the growth in the first quarter. Capital retained its strong positive contribution observed throughout the previous year. Labour had an insignificant negative contribution as a result of the slight decline in the number of worked man-hours in the economy. This was due to the smaller number of worked man-hours per employee and the number of employed which remained close to that in the previous quarter (NSI seasonally adjusted data). Household Behaviour Economic Activity and Share of Discouraged Persons (per cent, seasonally adjusted data) Sources: NSI Labour Force Survey, BNB calculations. Unemployment Rate (per cent, share of the labour force) (per cent, seasonally adjusted data) Household consumption continued to increase in the first quarter of 2018 driven by high consumer confidence and further increase in household labour income and employment in the economy. The NSI Labour Force Survey shows that in the first quarter of 2018 labour supply, as measured by the labour force, declined slightly according to seasonally adjusted data, remaining, however, at a relatively high level. As a result, the labour force participation rate declined on the previous quarter. The long-term downward trend in the number of discouraged persons continued. The Labour Force Survey shows that in the first quarter unemployment continued to decrease, reaching 5.5 per cent (non-seasonally adjusted data: 5.7 per cent). The fall in the unemployment rate was mostly due to the decline in the long-term unemployed (more than one year) and, to a lesser extent, to unemployed for less than a year. The seasonally adjusted unemployment rate calculated using the Employment Agency data also continued to decrease, reaching 6.1 per cent in May 2018 (non-seasonally adjusted data: 6.3 per cent). * NSI data. ** Employment Agency data. Sources: NSI Labour Force Survey, Employment Agency, BNB calculations. Employment and Nominal Wage Bill (annual change, per cent) Reflecting labour supply decline measured by the number of unemployed and rising labour demand in the economy, labour remunerations 66 Preliminary data subject to revision. Sources: the NSI SNA, BNB calculations. Economic Review 2/

48 Private Consumption and Consumer Confidence (annual change, per cent; seasonally adjusted data) (per cent; seasonally adjusted data) Retail Trade Turnover (seasonally adjusted data at current prices, per cent, on the previous quarter) Sources: the NSI, BNB calculations. continued to increase in the first quarter of 2018 (both annually according to non-seasonally adjusted data on the wage bill and average wage per employee and quarterly according to seasonally adjusted national accounts data). Note: Non-additive data on contributions due to direct seasonal adjustment of the total amount and its components. Data for the second quarter of 2018 are averaged for the April May period. Sources: the NSI, BNB calculations. The consumer confidence indicator went up in April 2018 due to the improving expectations of households for their financial position and the overall economic situation in the country as reported in the NSI Consumer Survey. At the time of the survey the share of persons intending to increase their spending on durable goods also rose. Household consumer behaviour continues to reflect the clear trend of growing employment and income. According to national account data, household consumption increased further on both annual and quarterly basis in the first quarter of Over the same period nominal retail trade turnover increased, with May data suggesting that private consumption will continue growing on a quarterly basis in the second quarter. Real traded volumes were the main driver behind the increased turnover in April and May. Newly extended consumer loans continued to increase in May, which is likely to additionally stimulate consumer demand in the following months. Private consumption is expected to increase further in the second half of the year driven by the growing labour income and the expectations of continuous recruitment of employees in the real economy given the sustainable labour demand. Household Propensity to Save and Expectations (per cent; seasonally adjusted data) * BNB data. ** NSI data. (BGN million; seasonally adjusted data) Sources: NSI Household Budget Survey, NSI Consumer Survey, BNB. 47 Economic Activity

49 Employment and Income Dynamics I II III IV I II III IV I II III IV I II III IV I (per cent, quarter-on-quarter; seasonally adjusted data) Employed persons Nominal wage per employee* Real wage per employee** Wage bill, nominal terms Wage bill, real terms** (per cent, on corresponding quarter of previous year, non-seasonally adjusted data) Employed persons Nominal wage per employee* Real wage per employee** Wage bill, nominal terms Wage bill, real terms** * The wage is calculated according to NSI data (SNA), with social security contributions paid by the employer deducted from the compensation per employee. The difference is divided by the number of employees. ** Data deflated by HICP. Sources: NSI SNA, BNB calculations, Eurostat. Retail Trade Turnover (per cent, quarter-on-quarter; seasonally adjusted data at constant prices) IV I II III IV I II III IV I II III IV I II Retail trade, excluding motor vehicles and motorcycles incl. Food, drinks and tobacco products Textile, clothing, footwear and leather Household appliances, furniture and other household goods Computer and communication equipment Pharmaceutical and medical goods, cosmetics and toiletries Unspecialised shops with different kinds of goods Automobile fuels and lubricants * Data for the second quarter of 2018 are averaged for the April May period. Source: the NSI. Short-term statistics. Economic Review 2/

50 Despite the favourable consumer confidence, high household savings rates were sustained. The NSI Household Budget Survey data show that the share of savings in household disposable income increased in the first quarter and the share of households intending to increase their savings in the following 12 months (according to the NSI Consumer Survey of April 2018) remained at the high levels attained. Household assets measured by their net position in the banking system grew in the first quarter of Contributions of Major Groups of Revenue to Growth in Total Revenue and Grants, Cumulatively (on an Annual Basis) (per cent; percentage points) Fiscal Policy Effects on the Economy As of end-may 2018 the consolidated fiscal programme surplus came to BGN million (1.4 per cent of projected GDP) 67, with budget balance posting a decline by BGN million compared to the same period of last year (0.3 percentage points of GDP). During January - May period total budget revenue rose by 8.6 per cent on the same period a year earlier driven mainly by tax revenue (8.2 percentage points)) and, to a much lesser degree, non-tax revenue (0.7 percentage points). Concurrently, the negative effect of grant dynamics on the budget revenue was almost entirely exhausted in the last two years. 68 Sources: the MF, BNB calculations. Contributions of Major Tax Groups to Tax Revenue Growth, Cumulatively (on an Annual Basis) (per cent; percentage points) By end-may the annual growth rate of tax revenue was 9.9 per cent, underpinned mainly by the dynamics of the revenue from social and health contributions (up 14.1 per cent), indirect taxes (up 6.4 per cent) and personal income tax (15.8 per cent). The sustainable growth in revenue from social and health contributions continued to reflect employment and labour income growth along with the increase in pension contributions (by 1 percentage point) and minimum wage and minimum social security income in force as from early Favour- 67 GDP forecasts of the Ministry of Finance for 2018 based on the spring macroeconomic forecast for the period, published in April While in 2015 and in the beginning of 2016 sizeable expenditure made in the final stage of implementing projects of the programming period had been refunded, receipts under EU programmes in the May 2016-May 2017 period were lower because of the early stage of implementing projects of the programming period. The second half of 2017 and the first half of 2018 saw an accelerated rate of absorbing funds under EU programmes. 69 See the Basic Parameters of the Budgetary Framework box, Economic Review, 4/2017. Sources: the MF, BNB calculations. 49 Economic Activity

51 able labour market dynamics was the major factor behind the increase in personal income tax revenue which was further driven by the continuous positive effects from measures initiated to improve tax collection. 70 Concurrently, the observed volatility of tax revenue growth was temporary and due, on the one hand, to shifting one month forward the revenue from annual corporate tax balance payment (in April instead of March) 71, and, on the other hand, to uneven monthly VAT revenue, probably resulting from a planned renovation of production capacities of oil products in March and April After the effect of these temporary factors was depleted, corporate tax and VAT revenue growth recovered in May to 9.9 per cent and 6.9 per cent respectively, continuing to reflect gross operating surplus and private consumption positive developments. In May the annual growth in total CFP expenditure remained at 11,7 per cent, exceeding significantly the annual rates of change observed in the previous year. Social expenditure (3.4 percentage points) and capital expenditure (3.0 percentage points) contributed most substantially to total expenditure growth. The accelerated increase in social expenditure since the year start was driven by both higher pension expenditure as a result of the continuous effects of the rises effective as of the second half of 2017 and the significant increase in health insurance payments. This upward dynamics is indicative of fiscal policy continuous positive impact on economic activity through disposable income of households. In the first five months of the year the high rates of growth in the staff costs and operating expenditure remained, accounting for 11.3 per 70 Based on Ministry of Finance Monthly Bulletin on Budget Performance and Major Indicators under the Consolidated Fiscal Programme in the first five months of 2018 there are still positive effects on the collection of taxes on dividends as a result of the NRA campaign for control over business entities and non-profit institutions with sizeable cash balances reported in their 2016 financial statements. 71 Based on Ministry of Finance Monthly Bulletin the deadline for submitting annual tax declaration and making the corporate tax balance payment was 2 April 2018 (31 March being a Sunday), which shifted forward the tax revenue in April According to the MF the possibility to change the corporate tax prepayments made by the financial institutions during the calendar 2017 could impact the corporate tax revenue, which in turn will have effect on the annual corporate tax due. Contribution of Major Groups of Expenditure to Total Expenditures Growth, Cumulatively (on an Annual Basis) (per cent; percentage points) Note: The Consolidated Fiscal Programme reports of January 2016 including staff expenditure, include wage, insurance and other remuneration expenditure, while in the reports for past periods, the latter were not included in operating expenditure. To prevent data inconsistencies prior to and after January 2016 resulting from the methodological change staff and operating expenditure are presented aggregately in the chart and separately in the table Key Budget Indicators Performance for Sources: the MF, BNB calculations. Economic Review 2/

52 cent and 12.9 per cent by end-may. The key factors behind this upward dynamics include public sector wage growth in the second half of 2017 and in the beginning of 2018, which was most pronounced in the education sector, and the additional expenditure related to the Bulgarian Presidency of the EU Council in the January to June 2018 period. Despite the accelerated growth in the staff costs and operating expenditure on an annual basis the contribution of the government consumption to the quarteron-quarter real GDP growth remained almost neutral in the first quarter of the year. Government capital expenditure continued to grow at high rates reaching 61.4 per cent by end-may. The period saw stronger growth in both capital expenditure under EU programmes and national capital expenditure to 78.3 percent and 44.5 per cent respectively on an annual basis with EU co-financed investment contributing most to total capital expenditure growth. For the first five months of the year the dynamics of the capital expenditure under the consolidated fiscal programme indicates a positive contribution of government investment to GDP growth in the first two quarters of the year. Contribution of Government Consumption Components to Real GDP Growth (percentage points, quarter-on-quarter, seasonally adjusted data) Sources: the NSI, BNB calculations. In the second half of 2018 favourable developments in tax and social security revenue are expected to be sustained with a stronger growth in the receipts under EU programmes which will be driven by the current accelerated increase in expenditure on EU programmes. It is anticipated that the current measures aimed at increasing social payments and wages coupled with the average pension rise by 3.8 per cent as from 1 July 2018 will have a dominant effect on current non-interest government expenditure dynamics contributing to sustain the fiscal policy s positive impact on economic activity through higher government consumption and larger transfers to households. Current significant growth rates in expenditure on EU programmes and national investment expenditure are projected to be retained, implying a corresponding positive contribution of government investment to the economic growth in the third and fourth quarters of Economic Activity

53 Key Budget Indicators Performance for Consolidated Fiscal Programme January May 1 January May 1 BGN million per cent 2 BGN million per cent 2 Total revenue and grants Tax revenue incl. social security and health insurance contributions Non-tax revenue Grants Total expenditure (incl. the contribution to EU budget) Staff Operating expenditure Interest Social expenditure, scholarships Subsidies Capital expenditure and government reserve growth incl. capital expenditure Contribution to general budget of the European Union BGN million difference 3 (BGN million) BGN million difference 3 (BGN million) Budget balance, on a cash basis Tax revenue under the state budget January May 4 January May 4 BGN million per cent 2 BGN million per cent 2 Tax revenue Corporate tax Income tax for individuals Value added tax Excise duties Customs duties Insurance premia tax Other taxes Note: The difference between the sum of individual components and total sum is due to rounding. 1 Based on monthly reports on cash-based performance of the Consolidated Fiscal Programme. 2 Annual rate of change on the same period of previous year. 3 The change of the budget balance on the same period of previous year. 4 Based on monthly reports on cash-based budget performance. Source: the MF. Economic Review 2/

54 4. INFLATION In the first five months of 2018 domestic annual inflation followed an upward trend and reached 2.3 per cent in May, reflecting both the increasing effect of internal factors on consumer price dynamics and the 19.6 per cent rise in international oil prices since early The gradual upward trend in services inflation which started at the close of 2017 was sustained, this group contributing most significantly to the overall inflation in May Goods and services with administratively controlled prices (including tobacco products) were another group with a comparatively high contribution to the annual inflation since the beginning of the year. Energy product inflation accelerated from the end of 2017 due to the strong hike in euro oil prices. In the second half of 2018 annual inflation is expected to increase further driven mainly by the positive contribution of services and energy prices. Goods and services with administratively controlled prices are also expected to make a strong positive contribution to the overall inflation in the following months. Over the first five months of 2018 the annual inflation rate accelerated from the end of 2017 to 2.3 per cent in May (1.8 per cent in December 2017). 72 By HICP component, the groups of services, energy products and administratively controlled prices had the largest positive contributions to the overall inflation. Inflation and Contribution of Major Commodity and Services Groups to It (per cent; percentage points; on an annual basis) Internal factors to inflation between January and May were associated with tobacco excise duties raised for a third consecutive year and increased prices of accommodation and package holiday services. The upward dynamics of consumer prices since early 2018 was strongly affected by external factors, such as euro oil price increases (by 19.6 per cent since the beginning of the year and 41 per cent on an annual basis in May 2018). The increased oil prices had both direct effects on energy product inflation and indirect effects on other HICP components through production costs of corporations and producer prices. Notes: This structure corresponds to the Eurostat classification; tobacco products and goods and services with administratively controlled prices are presented separately. The index of goods and services with administratively controlled prices is calculated through the elementary aggregates level in the consumer basket. Sources: the NSI, BNB calculations. The effect of international fuel, food and commodity prices on domestic inflation may be seen through price developments in goods imported into Bulgaria and their spillover through the supply chain. According to national accounts data, the annual growth in the price deflator for imported goods was 0.8 per cent in the first quarter of 2018 (against 5.1 per cent at the 72 The analysis in this Section employs NSI data on HICP. 53 Inflation

55 end of 2017). 73 Mineral fuels 74 and unprocessed raw materials 75 had positive contributions to the change in imported goods prices, whereas the import deflator for foods 76 and some finished products 77 recorded a decline on an annual basis. Declining dynamics in imported food prices corresponded to the downword trend since the second half of 2017 in international euro food prices on an annual basis. 78 Over the first five months of 2018 the total producer price index (PPI) in the domestic market continued to grow on an annual basis, its growth rate reaching 4.8 per cent (5.1 per cent at the end of 2017). The sub-sectoral breakdown shows that the increase in producer prices was driven mainly by production and distribution of electricity and heating, 79 consistent with increased international oil prices and higher regulated gas prices since early The positive contribution of manufacturing to the increase in producer prices in May 2018 was due to the rise in industrial production prices related to the manufacture of refined petroleum products 81 and metals. 82 Rate of Change in Brent Crude Oil and A95 Petrol Prices (per cent; on an monthly basis) Sources: the ECB, the NSI, BNB calculations. Rate of Change in PPI in Industry and HICP (per cent; on an annual basis) According to the end-use classification, the PPI increase in the January to May 2018 period was mainly due to the higher energy product prices (electricity, natural gas and refined petroleum products) which form part of firms production 73 Based on Standard International Trade Classification (SITC) import data. 74 This refers to the group of mineral fuels, oils and related products under the SITC. 75 This refers to the group of inedible (crude) materials (excluding fuels) under the SITC. 76 This refers to the groups of food and live animals and animal and vegetable fats, oils and waxes under the SITC. 77 This refers to the group of miscellaneous manufactured articles under the SITC. 78 The analysis employs ECB data on international food price index. 79 This refers to the group of production and distribution of electricity, heating, and gaseous fuels. 80 Changes in natural gas prices in the first quarter of 2018 as approved by the Energy and Water Regulatory Commission (EWRC) at the end of 2017 reflect the price dynamics of oil and alternative fuels in international markets over the last nine months, including movements in the USD/BGN exchange rate. For more details, see: news/143/65/komisiyata-za-energijno-i-vodno-regulirane-utvrditsenata-na-prirodniya-gaz-za-prvoto-trimesechie-na-2018-g. html 81 This refers to the production of coke and refined petroleum products group. 82 This refers to the manufacture of basic metals and fabricated metal products, excluding machinery and equipment group. Source: the NSI. Rate of change in PPI on the Domestic Market and Contribution by Major Sub-sectors (per cent; percentage points; on an annual basis) Source: the NSI. Economic Review 2/

56 Rate of Change in Producer Prices on the Domestic Market by Major Industrial Groupings (per cent; on an annual basis) Rate of Change of Food Price Index (per cent; on an annual basis) Sources: the NSI, BNB calculations. Source: the NSI. costs. Retention of the upward trend in these prices over the following months will be a prerequisite for an upward pressure on final consumer prices. Contribution of Major Sub-groups of Unprocessed Food to Overall inflation (percentage points; on an annual basis) In May 2018 the growth rate in producer prices of non-durable consumer goods decelerated from the end of 2017, reflecting declines in inflation of some food products in line with lower import prices. These developments were passed on to final consumer prices of processed food, their inflation slowing down to 1.6 per cent on an annual basis in May 2018 (from 2.7 per cent at the end of 2017). In May 2018 annual inflation in the unprocessed food group decreased to 1.5 per cent, from 3.0 per cent at the end of the prior year. This corresponded to the downward price dynamics in most of imported unprocessed food products over the first three months of the year (including fruit and vegetables). The sole exception to this trend was the sub-group of meat and meat products recording since the beginning of the year low year-on-year growth in both import and final consumer prices. The internal factors exerting an upward pressure on food prices involved higher production costs in agriculture measured by prices of goods and services intended for current consumption in agriculture. In the first five months of 2018 core inflation (including services and non-food prices) continued to follow its upward pattern since the end of 2017, reflecting mainly the increased Sources: the NSI, BNB calculations. Core inflation (per cent; on an annual basis) Sources: the NSI, BNB calculations. 55 Inflation

57 services prices. In May 2018 inflation in services rose to 4.0 per cent (1.2 per cent at the end of 2017), which was consistent with higher production costs of corporations in this sector and enhanced demand measured by the turnover index in services. 83 Broken down by services sub-component, short-stay accommodation services and package holidays and accommodation had the largest positive contributions to the overall inflation. Transport services also followed an upward price trend since the beginning of the year, mainly due to the higher airfares in the context of increased international oil prices. Catering contributed positively to the overall inflation, probably reflecting the upward dynamics in processed food prices. In the first five months of 2018 end prices of non-food goods continued to fall on an annual basis, though tending to slow down their decline rates (-0.6 per cent in May, from -0.9 per cent at the end of 2017). These developments corresponded to the downward dynamics in prices of imported finished products 84 amid the appreciation of the euro against the US dollar between January and May compared to the same period of Decreased non-food prices were mainly driven by durable goods (automobiles and television equipment) unlike non-durable goods prices which increased on an annual basis in sync with developments in the price component of retail trade volumes. In May 2018 nominal retail trade in non-food goods 86 posted annual growth of 5.6 per cent largely driven by both increased traded volumes and higher selling prices. 87 Since the beginning of 2018 administratively controlled prices continued to increase on an annual basis. Tobacco products had a significant positive contribution to the annual inflation, reflecting the increased excise duty on Contribution of Services and Major Services Subgroups to Overall Inflation (per cent; percentage points; on an annual basis) Sources: the NSI, BNB calculations. Contribution of Major Non-food Goods (Excluding Energy Products) Sub-groups to Overall Inflation (per cent; percentage points; on an annual basis) Sources: the NSI, BNB calculations. Annual Growth Rate of Nominal Retail Trade in Non-food Goods Excluding Trade in Motor Vehicles and Lubricants (per cent; percentage points; on an annual basis; working-day adjusted data) 83 For details, see the Behaviour of Firms and Competitiveness Section in Chapter The analysis employs data on imported goods prices in group 8 miscellaneous finished products n.e.c. under the SITC. 85 BNB estimates show that by 2014 approximately 27 per cent of final household consumption expenditure was spent on imported goods other than food and energy products (according to the world supply and use tables). 86 Retail trade in non-food goods excluding automobile fuels and lubricants. 87 For details, see the Behaviour of Firms and Competitiveness Section in Chapter 3. Sources: the NSI, BNB calculations. Economic Review 2/

58 Contribution of Major Sub-groups of Goods with Administratively Controlled and Tobacco Prices to Overall Inflation (per cent; percentage points; on an annual basis) Diffusion Index of Major Goods and Services Groups a) relative shares of declining HICP sub-indices on an annual basis (per cent) Sources: the NSI, BNB calculations. cigarettes since early 2018 for a third consecutive year. 88 Starting from April, administered price inflation decelerated slightly due to the exhausted base effect of raised regulated prices of natural gas, 89 heating and electricity 90 in April Since the beginning of 2018, the EWRC approved higher water supply prices in 14 regions, 91 thus contributing to the rise in administered prices. In May the diffusion index, showing the share of goods and services groups with declining prices on an annual basis, dropped to 19.4 per cent against 22.6 per cent at the end of This reflected primarily the decreased number of services sub-groups and industrial goods with declining prices. Concurrently, for the January to May period price growth was also reported in some sub-groups within each main group. Since early 2018, previous year s gradual acceleration in the number of goods and services subgroups reporting price rises has been retained, b) relative shares of increasing HICP sub-indices on an annual basis (per cent) Sources: the NSI, BNB calculations. 88 An increase in the specific excise duty from BGN 101 to BGN 109 per 1000 cigarettes, a decrease in the ad volarem excise duty from 27 to 25 per cent, and an increase in the minimum overall excise duty (specific and ad volarem) from BGN 168 to BGN 177 per 1000 cigarettes became effective from the beginning of For more details, see: 90 For more details, see: 91 For more details, see: 57 Inflation

59 thereby giving rise to expectations of continued upward HICP dynamics in the coming months. According to the NSI business surveys carried out, in June 2018, most managers in industry (88.0 per cent), services (91.2 per cent) and retail trade (83.5 per cent) expect selling prices to remain stable over the next three months. However, the balance of opinions in the three sectors was positive in the first six months of 2018, reflecting the larger share of managers expecting price rises as compared with those expecting price declines. In June 2018 NSI surveys showed a decline in the share of managers in services pointing to insufficient demand and sector competition as obstacles to their business activity compared to the end of 2017, with retention of these trends appearing to be a precondition for price increases in services. In the retail trade sector, which is the closest to end users in the supply chain, a decrease was also reported in the share of managers pointing to insufficient demand as a factor limiting their activity. In the second half of 2018 inflation is expected to increase gradually driven mainly by core inflation and energy products. Inflation in core HICP components will be primarily attributable to the services and projected improvements in labour market conditions and households consumer demand over the third and fourth quarters of the year. Energy prices are expected to continue rising in the second half of 2018 under the influence of the upward movement in international oil prices. Administratively controlled prices are also expected to make a positive contribution to the overall inflation, reflecting largely the effect of increased tobacco excise duties since the start of the year and, to a lesser extent, the January 2018 price rise in water supply services. Expectations of Selling Prices in Industry, Retail Trade and Services in the Next Three Months (balance of opinions, per cent, 6-month moving average) Source: the NSI. Share of Firms by Sectors Pointing to Insufficient Demand as a Factor Limiting Their Activity (relative share of all corporations, per cent) Sources: the NSI, BNB calculations. Share of Firms by Sectors Pointing to Sector Competition as a Factor Limiting Their Activity (relative share of all corporations, per cent, 6-month moving average) There are risks of higher than projected inflation to be recorded in the second half of After the cut-off date of our forecast, the EWRC announced on 1 July its decision to raise electricity, heating and gas supply prices by 2.03, 7.33 and per cent (national averages) Sources: the NSI, BNB calculations. Economic Review 2/

60 from the beginning of July, 92 implying higher than projected growth in administratively set prices. In addition, if oil prices increase at rates higher than expected, this may result in a faster than projected rise in fuel prices along with price increases in the other components of the consumer basket, reflecting higher production costs of corporations. In the first quarter of 2018 the house price index (HPI) continued to increase, though tending to decelerate its growth rate. In early 2018 its annual growth was 7.1 per cent, from 8.7 per cent on average in The increase in house prices was due to the higher prices of existing and, to a lower extent, new dwellings. Higher house prices in the first quarter were recorded in five of the six largest cities (with a population of over 120,000). Despite these developments, in early 2018 house prices still remained 21.6 per cent lower than the maximum values recorded in the third quarter of The upward house price dynamics was in line with the improved consumer confidence and increased disposable income of households driven by favourable labour market developments. Relatively low levels of new time deposit rates were also a driver for the enhanced household demand for housing as they motivate households to look for alternative investment opportunities in the form of real estate purchases. The annual percentage rate of charge (APRC) on new mortgage loans continued to decline, reaching 4.1 per cent in the first quarter (against 4.4 per cent on average for 2017). 93 Banks policy towards easing credit standards and conditions for households also acted in this direction. 94 Rate of Change of House Price Index (per cent; on an annual basis) Source: the NSI. Rate of Change in the House Price Index in Cities of More Than 120,000 Inhabitants over the First Quarter of 2018 (per cent; on an annual basis) Source: the NSI. In addition, house price dynamics also influenced the behaviour of firms in the construction sector. In early 2018 real value added in this sector reported year-on-year growth of 9.3 per cent (from 5.9 per cent on average for 2017). The continued increase in house prices at rates 92 For details, see: 93 Data on the APRC on housing loans to households are averages weighted by the relevant volumes of new business for a 12-month period. 94 For further details, see the Monetary and Credit Aggregates Section in Chapter Inflation

61 higher than those of the firms construction cost index, favourable financing conditions and improved household attitudes toward home buying appear to be factors for a further increase of economic activity in the construction sector over the following months. Higher supply of new residential dwellings is expected to exert a dampening effect on house price developments in the next quarters. An indication of this trend is the 37.1 per cent year-on-year rise of the area measured in square metres in new residential building permits issued in the first quarter of 2018 (against 33.1 per cent at the end of 2017). Construction Production Dynamics and New Residential Building Permits Issued (number of buildings; seasonally adjusted data) (2015 = 100; seasonally adjusted data) Sources: the NSI, BNB calculations. Key Indicators of the Housing Market Developments in Bulgaria (per cent; year-on-year change) Prices Indicators II III IV I House price index, total New dwellings Existing dwellings Inflation (HICP) House rentals, paid by tenants (HICP) Lending New housing loans Annual percentage rate of charge on new housing loans (per cent, at end of period) Housing loans balances Construction and investment Permits issued for the construction of new residential buildings (square meters) Value added in construction (at average annual prices for 2010) Constriction production index, building construction Fixed capital investment, residential buildings Construction cost index for new buildings Note: Data on the APRC on housing loans to households are averages weighted by the relevant volumes of new loans for a 12-month period. Sources: the NSI, the BNB, Eurostat. Economic Review 2/

62 Rates of Change in Major Goods and Services Groups Prices and Contributions of These Groups to Inflation Inflation accumulated as of May 2017 (December 2016 = 100) Inflation accumulated as of May 2018 (December 2017 = 100) Annual rate of inflation as of May 2018 (May 2017 = 100) Inflation (per cent) inflation rate by group (per cent) contribution (percentage points) inflation rate by group (per cent) contribution (percentage points) inflation rate by group (per cent) contribution (percentage points) Food Processed food Unprocessed food Services Catering Transport services Telecommunication services Other services Energy products Transport fuels Industrial goods Goods and services with administratively controlled prices* Tobacco products * The index of goods and services with administratively controlled prices is calculated through weighting the relevant elementary aggregates in the consumer basket. Sources: the NSI, BNB calculations. 61 Inflation

63 5. BULGARIAN NATIONAL BANK FORECAST OF KEY MACROECONOMIC INDICATORS FOR The BNB forecast of key macroeconomic indicators is based on data published as of 15 June ECB, EC and IMF assumptions on the global economic developments and dynamics of major commodity group prices in international markets as of 22 May 2018 are used. Based on the ECB, EC and IMF forecasts external demand for Bulgarian goods and services is expected to continue growing tending to slow down its growth rates. The baseline scenario of external demand neither takes into account the effects of the US foreign trade measures launched in the beginning of June 2018, nor a possible escalation of international trade conflicts during the forecast period. International prices are expected to increase significantly in In the period non-energy product prices are expected to continue increasing, while petroleum prices are anticipated to slightly decrease due to projected global supply growth of petroleum products. As regards the assumptions about external environment developments, there are risks of lower growth in external demand due to the uncertainty surrounding the assessments of the effects of the US trade policy on global trade developments and the uncertainty about the economic growth in some important for Bulgaria trading partners. Regarding the forecast of crude oil prices there are risks of a hike during the period due to a possible decline in global oil inventories, lower than expected US oil production and worsening of the global geopolitical situation. The improved global economic environment contributed to Bulgaria s stable economic growth. According to the macroeconomic forecast between 2018 and 2020 the economic growth rate will be stable with a positive contribution of domestic demand (private consumption, private investment and a significant increase in public investment). Private consumption growth is supported by the improved labour market: the unemployment rate is decreasing against the background of an increasing economic activity rate, with relatively high growth of wages being a motivating factor for inclusion in the labour force. Increasing domestic demand pushed up the demand for import goods and services. As a result net exports will have a negative contribution to real GDP growth despite the expected steady growth in exports. These developments will change the current account balance from a surplus to a deficit in the period, with inflows of capital transfers from the EU and from foreign direct investments exceeding significantly this deficit. Inflation, influenced largely by international market price dynamics, is expected to accelerate its rate in 2018 and then to slow down and stabilise at 2 per cent annually. Risks to the macroeconomic outlook are related mainly to the uncertainty of assumptions about the external environment. An internal factor creating uncertainty to the outlook is the public investment dynamics which is determined by the EU fund absorption rate. Forecast The BNB forecast of key macroeconomic indicators is based on data published as of 15 June ECB, EC and IMF assumptions on global economic developments and the dynamics of international prices of major commodity groups as of 22 May 2018 are used. Based on these assumptions global economic growth in 2018 and 2019 is expected to remain close to that reported in 2017 supported by still favourable funding conditions. Given the gradually weakening positive effect of current cyclical factors in advanced economies and exhaustion of the positive effect of US fiscal stimuli on growth at the end of the forecast horizon, global economic growth in 2020 is expected to slow down compared to In accordance with these forecasts for the period external demand for Bulgarian goods and services will continue growing, the rate of growth gradually moderating, mostly in line with the projected lower growth in the EU Member States, Bulgaria's main trading partners. The baseline scenario of external demand neither takes into account the effects of the US foreign trade measures Economic Review 2/

64 launched in the beginning of June 2018, nor a possible deepening of international trade conflicts over the projection horizon. Year-on-year prices of energy products increased significantly in the first five months of 2018 in both US dollars and euro irrespective of the movements in the EUR/USD exchange rate. This reflected the lower supply due to temporarily discontinued oil extraction in the USA, Libya, Venezuela and in the North Sea and the stronger demand as a result of global economic growth. The average annual price of petroleum products in 2018 is expected to increase on Over the remaining projection horizon supply is anticipated to increase due to the agreement between OPEC and other oil producers on increasing oil production compared to the February 2018 commitment, which prompted a decline in international oil prices. Non-energy product prices exhibited a divergent dynamics both in euro and US dollars in the first five months of 2018 reflecting the appreciated euro against the US dollar. Higher non-energy product prices in US dollars were a result of increased agricultural commodity and metal prices. Given the increased global demand for metals the upward price dynamics was underpinned by factors limiting supply such as China s measures intended to reduce air pollution implying contraction of some productions and US sanctions imposed on the second largest aluminium producer in April. Rebalancing of demand and supply in non-energy products is expected to continue to put upward pressure on prices in US dollars and euro over the projection horizon. In 2018 economic growth in Bulgaria is expected to accelerate on 2017, with domestic demand contributing most substantially due to sustained private consumption growth and the anticipated acceleration of government consumption and gross fixed capital formation, including a sizeable increase in public investment. The negative contribution of net exports is expected to slightly increase compared with 2017, which will reflect accelerated imports amid high growth rates of exported goods and services. In the period real GDP growth in Bulgaria is expected to remain at around 4 per cent, with domestic demand being the main factor behind its dynamics. Private consumption growth in real terms will slightly slow down from 2017 but will continue to have a strong positive contribution to the overall economic dynamics. Household disposable income will further increase due to improved labour market conditions, which will result in lower unemployment rates and higher employment and wages. The expected slower rise in the number of employed due to demographic factors will be the reason behind the gradual private consumption slowdown until the end of the projection horizon. The positive sentiment of firms, expected enhanced demand for goods and services and high capacity utilisation rate create conditions for keeping the private investment growth rate high. Government consumption is also expected to increase in the period mostly due to government s plans for wage increases in the education sector according to the latest revision of the medium-term budget projections. Following the expected significant increase in public investment in 2018 the forecast points to a further rise at a slower rate in the reflecting the dynamics of costs under EU programmes. Projected growth of external demand combined with the increasing competitiveness of the Bulgarian economy will contribute to the acceleration of export growth. Given the expected significant increase in domestic demand the growth rate of imports of goods and services will outpace that of exports. Therefore, net exports will have a negative contribution to overall growth in the economy. In 2018 growth in real exports of goods and services is anticipated to accelerate its rate compared to 2017, recovering growth in exports of services and slightly slowing down growth in exports of goods. The return to growth in exports of services in 2018 will be driven by the expected increase in exports of other services 95 and a further increase in revenue from tourism. Between 2019 and 2020 exports of goods are expected to gradually decelerate on 2017, reflecting the projected slowdown in external demand. Though at slower rates, exports of goods and services will continue increasing faster than 95 Services other than transport, travel, manufacturing services on physical inputs and maintenance and repair services. 63 BNB Forecast

65 external demand due to competitive advantages of Bulgarian exporters, which will help increase Bulgaria s market share in the world trade. Imports of goods and services will accelerate their rate of real growth in 2018 in line with the projected domestic demand growth (in particular, the expected strong increase in government investment). In 2019 and 2020 growth in imports is anticipated to moderate on 2018 and to follow the domestic demand dynamics. As a result, net exports are expected to have a negative contribution to real GDP growth over the projection horizon, most notably in In 2018 terms of trade (measuring the change in prices of exported goods against the change in prices of imported goods) for Bulgaria are expected to deteriorate, which coupled with faster real growth of imports than exports of goods will contribute to higher trade deficit as a share of GDP compared with The trade deficit will continue to increase until the end of the projection horizon, with the expected improvement in the terms of trade having a partially offsetting effect. The services trade balance in 2018 is anticipated to record a higher surplus as a share of GDP compared with 2017 due to the higher projected rise in exports than imports of services. Between 2019 and 2020 the surplus in services trade as a share of GDP is expected to stabilise. Enhanced economic activity in Bulgaria is expected to boost companies profits which will result in a gradual increase in dividends paid to non-residents, and correspondingly, to a rise in the deficit of the balance of payments primary income account at the end of the forecast horizon. In 2018 net transfers on the balance of payments secondary income account are projected to decrease as a share of GDP compared with 2017 and to remain at these levels in the period. As a result of the component dynamics the surplus on the balance of payments current account will decrease as a share of GDP in 2018 compared with 2017, and in 2019 and 2020 the current account balance will turn from surplus to a gradually increasing deficit. The improvement in labour market conditions is expected to encourage the inclusion of the population in the labour force and growth in employment. However, this will be insufficient to fully offset the effect of negative demographic developments. Unemployment is expected to decrease over the review period but at a slower rate. Labour productivity growth will accelerate as a result of the expected slight rise in employment and relatively constant dynamics of real GDP. The compensation per employee will continue to increase at a relatively high rate as a result of constraints in labour supply. Faster growth of labour productivity than compensation per employee will prompt a gradual slowdown in unit labour cost growth over the forecast horizon. The mid-2017 upward trend in international oil prices will be a factor behind higher harmonised index of consumer prices in the following months. Core inflation is also expected to follow a trend toward a gradual increase due mainly to service price developments, and to make a significant positive contribution to the overall inflation. Administratively controlled prices will be another group contributing positively to the overall inflation mainly as a result of raised tobacco products excise duties since early 2018 and higher prices of water supply services over the period. In line with the expected stabilisation of international oil prices in 2019 and 2020, inflation is anticipated to fall slightly in the coming years. As regards deposit dynamics in the Bulgarian banking system, their growth rate is expected to remain at around 6 7 per cent over the forecast horizon, reflecting employment and wage rises and further contributing to private consumption and savings growth. Expectations of sustained low deposit interest rates over the forecast horizon will reduce deposit growth. Interest rates on loans are anticipated to follow the euro area interest rate cycle, with market expectations for three-month EURIBOR level to remain broadly unchanged in 2018 and slightly increase in the period. Strengthening of credit activity is expected in case of higher growth of loans to households in 2019 due to the institution reclassification effect 96. The growth rate of loans to non-financial corporations will 96 As of April 2018 other monetary financial institutions sector includes BNP Paribas Personal Finance S.A. branch Bulgaria, with balances reclassified from other financial intermediaries sector. The reclassification results from the transformation through merger of BNP Paribas Personal Finance EAD specialised in lending and reported to-date in other financial intermediaries sector for the purposes of monetary statistics. Economic Review 2/

66 exceed that of loans to households mainly as a result of the payment of loans extended under the National Program for Energy Efficiency of Multi-family Residential Buildings and significant decrease in the volume of new loans under this programme. Forecast revisions The projections for real GDP growth in 2018 have been revised downwards, while those for 2019 remain unchanged compared to the forecast published in the Economic Review, issue 4 of The positive sentiment observed in the first half of 2018 coupled with improving labour market conditions resulting in expected employment and wage growth at comparatively high rates led to upward revision of private consumption for both 2018 and The anticipated growth of gross capital formation has also been revised upwards reflecting higher projected volume of government investment compared to the previous forecast. In line with the dynamics of domestic demand components, imports of goods and services are expected to grow at higher rates than estimated in the previous forecast. Expected growth rates of exports of goods and services in 2018 and 2019 remain similar to the previous ones, and, according to the revision, will result in more negative contribution of the net exports contribution to GDP growth over the forecast horizon. Significant increases in international oil prices are the main reason behind the upward revision of inflation for 2018 and Private consumption and labour costs growth has been revised upwards over the forecast horizon suggesting higher inflation in services for the period compared with the previous forecast. Expectations of higher food inflation reflect mainly the revised assumptions for growth of international food prices in euro. GDP and Inflation Forecast Revisions (21 June 2018 vis-ў-vis 28 December 2017) Annual rate of change, per cent Forecast as of 21 June 2018 Forecast as of 28 December 2017 Revision (percentage points) GDP at constant prices Private consumption Government consumption Gross fixed capital formation Exports (goods and services) Imports (goods and services) HICP at end of period Core inflation Energy Food Goods and services with administratively controlled prices and tobacco products Risks to the Forecast As regards the economic growth, there are risks of lower than expected GDP growth rate reflecting the uncertainty in the outlook for external environment and public investments. As regards external environment, the following risks have been identified: higher than expected oil price rises due to supply lagging behind demand and decline in external demand reflecting the negative impact of US trade policy on global trade and worsened economic situation in some of Bulgaria s major trading partners. Despite the expected fall in oil prices in 2019 and 2020 it is possible to keep the upward trend in case of lower than expected US production, worsening of the global geopolitical 65 BNB Forecast

67 situation and higher than projected decline in global oil inventories. The response of the parties concerned to the imposition of higher tariffs on certain imported goods and the implementation of other protectionist measures by the USA, could result in deceleration of global trade growth and decline in external demand for Bulgarian goods and services. The impact of this risk for Bulgaria could be significant if the growth of its major trading partners is affected. Heightened uncertainty surrounding Brexit along with potential deterioration of the economic situation in some of Bulgaria s major trading partners also represent a risk of decline in the external demand for Bulgarian goods and services. Due to possible delay in government investment plans implementation in 2018 for administrative reasons there is a risk of lower growth of public investment with negative effect on real GDP growth. The uncertainty over the outlook for a particular indicator may be graphically illustrated by means of the so-called fan chart. Chart bands depicted in a particular colour set an interval in which with a certain probability the projected value is expected to fall (for further details see the note to the chart on GDP growth). Each interval widens with the increase in the forecast horizon, reflecting the increasing uncertainty Expected Annual Rate of Change of Real GDP Note: The fan chart shows the expert views of the forecasters on the uncertainty around the projected value based on probability distribution. The reporting period shows revisions of GDP growth estimates. The middle band of the chart, depicted in the darkest colour, includes the central projection and the probability distribution shows 20 per cent probability for the actual value to fall in this band in each of the quarters. If neighbouring bands (in the same brighter colour) are added to the middle band, there would be a 40 per cent coverage of the probability mass. Thus, by adding each same colour couple of bands, the probability for the value to fall there would be increased by 20 percentage points to reach 80 per cent. The probability for the value to remain outside the coloured part of the chart is 20 per cent based on the distribution chosen. Expected Annual Rate of Change in Inflation at the End of the Period Note: The fan chart shows the expert views of the forecasters on the uncertainty around the projected value based on probability distribution. The middle band of the chart, depicted in the darkest colour, includes the central projection and the probability distribution shows 20 per cent probability for the actual value to fall in this band in each of the quarters. If neighbouring bands (in the same brighter colour) are added to the middle band, there would be a 40 per cent coverage of the probability mass. Thus, by adding each same colour couple of bands, the probability for the value to fall there would be increased by 20 percentage points to reach 80 per cent. The probability for the value to remain outside the coloured part of the chart is 20 per cent based on the distribution chosen. further into the future. The fan chart on the annual GDP growth shows that there is a 60 per cent probability for annual GDP growth to range from 3.4 to 4.6 per cent for There are risks of higher than projected inflation to occur in case of unforeseen rises in administratively controlled prices and international oil prices. Higher than expected oil prices presuppose a faster than projected rise in fuel prices along with other groups of goods and services which are indirectly affected by fuel prices (transport services, natural gas and heating prices). A risk of higher than projected inflation arises from the possibility of higher than expected increase in unit labour costs and the pass-through on final prices and in particular on services. The balance of risks regarding inflation displayed in the fan chart below shows that there is a 60 per cent probability for annual HIPC growth to be from 1.5 to 3.7 per cent at the end of A key source of uncertainty related to the balance of payment projection is the external environment: external demand and international commodity price dynamics which could have a significant effect on Economic Review 2/

68 the trade balance and current account accordingly. If any of the described risks regarding the external environment materialises, a higher than projected trade balance deficit and consequently current account balance deterioration could be expected. As regards credit aggregates and deposits, there is a risk that credit growth proves to be lower than projected. Lower than expected external demand will result in a decline in the domestic demand for loans due to the high degree of openness of the Bulgarian economy. Additional risk of slowdown in credit growth stems from the uncertainty regarding bank policies to sell credit portfolios and write-off nonperforming loans. Uncertainty is also reported in respect of the future impact of National Programme for Energy Efficiency of Multi-family Residential Buildings on the growth rates of household loans. Forecast of Key Macroeconomic Indicators for (per cent) Annual rate of change GDP at constant prices Private consumption Government consumption Gross fixed capital formation Exports (goods and services) Imports (goods and services) HICP at end of period Core inflation Energy Food Goods and services with administratively controlled prices and tobacco products Employment Unit labour costs Labour productivity Unemployment rate (share of labour force) Claims on non-government sector Claims on corporations Claims on households Deposits of the non-government sector Share of GDP Balance of payments current account Trade balance Services, net Primary income, net Secondary income, net Annual rate of change External assumptions External demand Average annual Brent oil price (in USD) Average annual price of non-energy products (in USD) Brent oil price at the end of period (in USD) Price of non-energy products at the end of period (in USD) Data refer to non-financial corporations. 67 BNB Forecast

69 RESEARCH TOPICS METHODS FOR ASSESSMENT OF THE FINANCIAL CYCLE IN BULGARIA The study of the financial cycle and the way it interacts with the business cycle is crucial for taking informed decisions when applying the macro-prudential policy instruments of the central bank. The financial cycle arises from the self-reinforcing interaction between the financial constraints of economic agents and the change in their risk perceptions and risk tolerance. 1 As a result of this interaction, the financial markets undergo periods of ups and downs. Taking timely measures of the macro-prudential policy is important in limiting the cyclical systemic risk accumulation in the financial system and the adverse impact of financial crises on economic growth. The study of the phases of the financial cycle and their link to the business cycle has been a subject of particular attention since the onset of the global financial crisis of , which is yet another crisis, proving the strong adverse impact of fluctuations in the financial system on economic activity. The set of macro-prudential supervision instruments 2 is expected to be an effective means of limiting the cyclical accumulation of risk in the financial system. Regulation (EU) No 575/ and Directive 2013/36/ЕU 4 envisage the counter-cyclical capital buffer as one of the instruments to limit the cyclical accumulation of risk in the financial system. Recommendation (ESRB/2014/1) 5 of the European Systemic Risk Board (ESRB) places the focus on the extent to which the credit-to-gdp ratio deviates from its long-term trend in the assessment of the financial cycle, recommending the use of additional indicators to better substantiate the analysis. Two alternative approaches for assessing the financial cycle dominate in the academic literature. The first one is based on the assumption that the length of the financial cycle is within predefined ranges, most often between 8 and 30 years, and this assumption is used to derive an estimate of the financial cycle and study its dynamics and characteristics. Within this approach, univariate frequency filters are most often used to derive trend and cycle information from the time series, such as Hodrick and Prescott filter 6 (HP filter) and Christiano and Fitzgerald filter 7 (band-pass filter). Another method used within the first approach is the turning point analysis 8, based on the identification of turning points in the time series using a predefined algorithm to identify these points. 1 Borio, C. The financial cycle and macroeconomics: What have we learnt? Journal of Banking & Finance, Vol. 45, 2014, pp For further information on the business cycles see: Mitchell, W. C. Business Cycles: The Problem and Its Setting. New York: National Bureau of Economic Research; 1927; Burns, A. F. and W. C. Mitchell. Measuring Business Cycles. New York: National Bureau of Economic Research, Vol. 2, 1946; Ganev, К. Business cycles. Theories and models. Sofia University Press St. Kliment Ohridski, For more information on macro-prudential supervision instruments see: index.htm 3 Regulation (EU) No 575/2013 of the European Parliament and of the Council of 26 June 2013 on prudential requirements for credit institutions and investment firms and amending Regulation (EU) No 648/26. 4 Directive 2013/36/EU of the European Parliament and of the Council of 26 June 2013 on access to the activity of credit institutions and the prudential supervision of credit institutions and investment firms, amending Directive 2002/87/EC and repealing Directives 2006/48/EC and 2006/49/EC. 5 Recommendation of the European Systemic Risk Board of 18 June 2014 on guidance for setting countercyclical buffer rates. 6 Hodrick, R. and E. C. Prescott. Postwar U.S. Business Cycles: An Empirical Investigation. Journal of Money, Credit, and Banking. 1997, Vol. 29 (1), pp Christiano, L. and T. Fitzgerald. The Band Pass Filter. International Economic Review, 2003, Vol. 44, pp See Harding, D. and A. Pagan. Dissecting the cycle: a methodological investigation. Journal of Monetary Economics, 2002, Vol. 49 (2), pp ; Claessens, S., M. A. Kose and M. E. Terrones. How do business and financial cycles interact?, Journal of International Еconomics, 2012, Vol. 87 (1), pp Economic Review 2/

70 Within the second approach no constraints are imposed on the length of the financial cycle. An example of such an approach is a structural unobserved components model 9, based on the estimation of a system of equations that characterise the individual cyclical and trend components of a set of observable variables. The advantage of using this type of model is that its built-in design allows for greater flexibility in the decomposition of the time series without imposing limits on the length of the financial cycle. Another example of applying this approach is the so-called wavelet analysis 10, within which the time series are treated only as a function both in time and frequency. This method allows for the exploring the synchronisation and coherence of the cycles over several time series. The aim of this study is to assess the phases of the financial cycle in the Bulgarian economy for the period Two specific methods belonging to the two alternative approaches are applied using multivariate models based on a set of indicators to measure the financial cycle. The first one is based on principal component analysis to extract the cyclical components by using the band-pass filter and then to combine them into a single measure of the financial cycle. When applying this method it is assumed that the length of the financial cycle ranges between 8 and 30 years The second method relies on a structural unobserved components model, which does not impose any restrictions on the length of the financial cycle. Band-pass Filter and Principal Components Analysis Following the literature on financial cycles and taking into account the Recommendations of the ESRB 11 on the indicators that may complement the main indicator to assess the phases of the financial cycle, i.e. the credit to GDP ratio deviation from its long-term trend, the assessment of the financial cycle is based on aggregation of the cyclical component information of the following set of indicators: 1) Measures of credit developments and private sector debt burden. Excessive growth of lending to firms and households and, respectively, high private sector indebtedness create risks of inability to repay the debt obligations in the downside phase of the cycle. 2) Measures of potential overvaluation of property. Rapid growth in property prices is one of the factors considered in the literature to accompany and accelerate the onset of most financial crises. During the expansion phase of the financial cycle cheap financing can push demand and prices above a sustainable level, which can stimulate further credit expansion as a result of the wealth effect on consumers and the increased value of collateral. 3) Measures of external imbalances. High and persistent current account deficits are usually associated with an excessive accumulation of external debt funding the overheating of the economy. 4) Interest rate spreads. In the upside phase of the business cycle, the favourable macroeconomic environment, the positive attitude of economic agents and growing income and profits contribute to a reduction in risk perceptions, resulting in declining interest rate spread. At the same time, the reduced risk assessment can stimulate the financing of more risky projects, which can turn out to be unprofitable in the downside phase of the cycle, generating losses for creditors. 5) Measures of the strength of banks balance sheets The low level of capital adequacy and the high ratio of loans to attracted deposits reduce the ability of banks to respond to shocks. 9 See Runstler, G. and M. Vlekke. Business, housing and credit cycles. Working Paper Series, European Central Bank, 2016, N 1915, p See Kunovac, D., M. Mandler and M. Scharnagl. Financial Cycles in Euro Area Economies: A Cross-Country Perspective, Deutsche Bundesbank Discussion Paper, 2018, N 4, p See footnote Research Topics

71 Table 1. Indicators Used to Obtain a Measure of the Financial Cycle in Bulgaria Group Indicators Source Measures of credit developments and private sector debt burden Measures of potential overvaluation of property prices Credit to NFC-to-GDP (Domestic credit + external debt*) Credit to households-to-gdp Credit to NFC (annual growth rate) Credit to households (annual growth rate) House Price Index (annual growth rate) BNB: Monetary statistics, Gross external debt; NSI BNB: Monetary statistics; NSI BNB: Monetary statistics BNB: Monetary statistics Measures of external imbalances Current account balance-to-gdp BNB: Balance of payments; NSI Interest rate spreads Measures of the strength of banks balance sheets Spread between interest rates on new loans to NFC and 3-Month EURIBOR Spread between interest rates on new loans to households and 3-Month EURIBOR Capital-to-asset ratio (leverage) Profits-to-assets ratio Loan-to-deposit ratio NSI BNB: Interest rate statistics,; ECB BNB: Interest rate statistics; ECB BNB: Banking supervision BNB: Banking supervision BNB: Banking supervision *External debt includes the terms other sectors and direct investment: intercompany lending from the Gross external debt statistics The indicators used to obtain an estimate of the financial cycle are displayed in Table 1. All data are quarterly and cover the period 1999 Q Q4. The indicators are normalised 12 to ensure the comparability of their units which is a standard approach in the academic literature on financial cycle. In addition to normalisation some of the indicators are multiplied by a coefficient of -1, so that an increase of the respective variable indicates accumulation of risk and a decline is indicative of a materialisation or reduction of risk. Indicators that are multiplied by a coefficient of -1 are the current account ratio as a ratio to GDP, interest rate spreads and capital to bank system assets ratio. After the normalisation procedure the so-called band-pass filter is applied to extract only the cyclical component of each series. This type of filters are two-sided univariate frequency filters designed to isolate and extract cycles of certain lengths in the series. The frequency band of the filter which defines the upper and lower boundary of the cycle length to be extracted is set in advance. There is a common view in the literature that in contrast to the business cycle which is a shorter-term process, the financial cycle is rather a mid-term length process. 13 When extracting cyclical components of indicators used to obtain an estimate of the financial cycle by means of a band-pass filter, a frequency of 32 to 120 quarters is usually used in the academic literature and therefore a band-pass filter with the same frequency has been used in this study. Chart 1 displays the extracted cyclical components of the selected indicators to obtain an estimate of the financial cycle. A conclusion could be made that in 2017 an accumulation of risk is observed in respect of growing house prices, declining interest spreads, rising bank profits and acceleration of lending growth rate to households. Due to the normalisation of the indicators the crossing of the null axis can be interpreted as a boundary between the phases of risk accumulation (in case of positive null axis deviation) and risk materialisation (in case of negative null axis deviation). Once the cyclical components of the selected indicators are extracted, the principal component analysis is used in order to obtain an aggregate measure of the financial cycle. This method is designed to extract common components of a set of variables and represents a statistical procedure that transforms a number of (possibly) correlated variables into a set of uncorrelated variables 12 The normalisation procedure for each time series consists of extracting the mean of the studied period from each observation and dividing it by the standard deviation. 13 See for example Drehmann, M., C. Borio and K. Tsatsaronis. Characterising the Financial Cycle: Don t Lose Sight of The Medium Term!, BIS Working Paper, 2012, 380, p. 37; Aikman, D., A. G. Haldane and B. D. Nelson. Curbing the credit cycle, The Economic Journal, 2015, Vol. 125 (585), pp Economic Review 2/

72 called principal components. The principal components are linear combinations of the original variables weighted by their contribution to explain the variance in all series. Chart 1. Cyclical Components of the Indicators Used to Obtain an Estimate of the Financial Cycle To extract the principal components of the selected indicators which are considered to characterise the financial cycle, it is necessary to group them in a specific way. To that end, nine groups of indicators have been established. The common indicators are extracted from each of these groups with the first principal indicator being of major interest for the analysis which explains the greater part of the variance in each group. The first group (FC1) includes only the cyclical components of the two ratios of credit to the GDP, i.e. credit to non-financial corporations as a ratio to GDP (including bank credit and external financing) Sources: the BNB, the NSI, the ECB, own calculations. and credit to households as a ratio to GDP. The use of the deviation of credit-to-gdp from its long-term trend as a core indicator to obtain a measure of the cyclical movements in the financial system has a long history in the literature. 14 To the above two ratios, the annual growth of credit to non-financial corporations and to households respectively is added in the second group (FC2) to account for different pattern of credit developments from that of the real GDP. 15 In the third group (FC3) the spreads between interest rates on new loans to households and NFC with the 3-month Euribor are added to account for funding conditions and risk premiums. 16 As a next indicator property prices which are commonly used in the literature to take account of risk assessment in the economy are included. 17 The inclusion of the current account balance as a ratio to GDP in the fifth group (FC5) aims at accounting for potential external imbalances. 18 Though not very common in the literature on financial cycles the last four groups include bank balance sheets indicators with the last group being limited only to these indicators. 19 Table 2 gives information on the way the indicators are grouped, and Table 3 shows the results of the principal component analysis. 14 See Schularick, M. and A. M. Taylor. Credit booms gone bust: Monetary policy, leverage cycles, and financial crises, , American Economic Review, Vol. 102 (2), 2012, pp ; Jordá, Ó., M. Schularick and A. M. Taylor. When Credit Bites Back, Journal of Money, Credit and Banking, 2013, Vol. 45 (s2), pp. 3 28; Dell Ariccia, G., D. Igan, L. Laeven and H. Tong. Policies for Macrofinancial Stability: Dealing with Credit Booms and Busts. In: Financial crises: Causes, consequences, and policy responses, Washington: IMF, 2014, pp ; Aikman, D. et al. Op.cit. 15 See Schularick, M. and A. M. Taylor. Credit booms gone bust:..; Jordá, Ó., M. Schularick and A. M. Taylor. When Credit Bites Back See Giese, J., H. Andersen, O. Bush, C. Castro, M. Farag and S. Kapadia. The Credit-To-GDP Gap And Complementary Indicatrs For Macroprudential Policy: Evidence From The UK, International Journal of Finance & Economics, Vol. 19 (1), 2014, pp ; Plasil, M., T. Konecny, J. Seidler and P. Hlavac. In the Quest of Measuring the Financial Cycle. Technical report. Czech National Bank, Research Department, 2015; English, W. B. Interest rate risk and bank net interest margins, BIS Quarterly Review, Vol. 10, 2002, pp See Borio, C. The financial cycle...; Drehmann, M., C. Borio and K. Tsatsaronis. Characterizing the Financial ; Claessens, S., M. A. Kose and M. E. Terrones. How do business ; Dees, S. Credit, asset prices and business cycles at the global level, Economic Modelling, Vol. 54, 2016, pp See Giese, J. et al. Op. cit.; Plasil, M. et al., Op. cit.; Comunale, M. and J. Hessel. Current Account Imbalances in the Euro Area: Competitiveness or Financial Cycle, DNB Working Paper, 443, 2014, р See Giese, J. et al. Op. cit.; Stremmel, H. Capturing the financial cycle in Europe, ECB Working Paper, 1811, 2015, p Research Topics

73 Table 2. Groups of Indicators Used to Obtain an Aggregate Measure of the Financial Cycle Financial cycle measure Indicators included (cyclical components) FC1 FC2 FC3 FC4 FC5 FC6 FC7 FC8 FC9 Credit to NFC-to-GDP : : : : : : : : Credit to households-to-gdp : : : : : : : : Credit to NFC (y-o-y) : : : : : : : Credit to households (y-o-y) : : : : : : : Spread, interest rates on credits to NFC : : : : : : Spread, interest rates on credits to households : : : : : : House price index (y-o-y) : : : : : Current account balance-to-gdp : : : : Capital-to-asset ratio : : : : Profits-to-assets ratio Loan-to-deposit ratio : : : : : Table 3 shows how much of the variance in each group of indicators is explained by the first principal component. In the case of the first potential measure of the financial cycle (FC1), which consists of only two indicators, i.e. credit to non-financial corporations and credit to households both as a ratio to GDP, the first principal component (PCA1) explains about 98 per cent of the variance. Such a result can be expected when the number of variables in the group is small and when the variables share common dynamics. Table 3. Results of the Principal Component Analysis FC1 FC2 FC3 FC4 FC5 FC6 FC7 FC8 FC9 PCA Source: own calculations. When adding complementary indicators to the credit-to-gdp gap, the variance explained by the first principal component expectedly declines. The comparison between the different measures of the financial cycle shows that except FC1, most of the variation that is explained by the first principal component is present in the sixth group FC6 (66 per cent). FC6 has also the advantage of containing indicators from all groups defined earlier: measures of credit developments and private sector debt burden, interest rate spreads, measures of potential overvaluation of property, measures of external imbalances and measures of the Chart 2. Indicators Used to Obtain a Measure of the Financial Cycle in Bulgaria Source: own calculations. strength of banks balance sheets. This means that they can be more informative about cyclical movements in the financial system than the credit-to-gdp gap alone. Graphical illustration of the nine potential measures of the financial cycle in Bulgaria is displayed in Chart 2. When only the deviation of the credit-to-gdp ratio from its long-term trend is used (FC1), it can be concluded that the peak of the financial cycle occurred at the end of Expanding the scope of the observed variables leads to different conclusions, namely that the peak of the financial cycle occurred earlier than Q and that in 2017 the Bulgarian economy was no any longer in the phase of risk materialisation, but in a phase of cyclical risk accumulation. The volatility (measured by Economic Review 2/

74 the standard deviation), the peak and the troughs of the financial cycle according to nine potential measures are displayed in Table 4. Table 4. Volatility, Maximum (Peak) and Minimum (Trough) of the Financial Cycle in the Different Indicators of the Financial Cycle in Bulgaria FC1 FC2 FC3 FC4 FC5 FC6 FC7 FC8 FC9 Volatility Minimum Timing 2002 Q Q Q Q Q Q Q Q Q1 Maximum Timing 2009 Q Q Q Q Q Q Q Q Q4 Source: own calculations. Table 4 shows that the volatility increases gradually starting from FC1 and moving to FC8. The only exception is potential measure FC9, where only the three indicators for the strength of banks balance sheets are included. According to FC1 and FC2, which include credit-to-gdp ratio and credit growth rates, the peak of the cycle lies between 2009 Q4 and 2010 Q2. At the same time the trough of the cycle lies somewhere between 2002 Q1 and 2002 Q4. The inclusion of interest rate spreads and the house price index in the aggregate measure of the financial cycle FC4 leads to a significant backward shift of the peak to 2006 Q1. According to financial cycle measures FC5 to FC8 the estimate of the cycle peak coincides 2006 Q3. Even when taking into account only indicators for the strength of banks balance sheets, the picture does not change much and the peak and trough of the cycle lie somewhere close to those indicated by financial cycle measures FC5 to FC8. To decide which is the most appropriate measure of the financial cycle among various alternatives, the so-called concordance index is applied. 20 It allows to measure the degree of synchronicity between the different indicators used to obtain the aggregate measure of the financial cycle and the aggregate measure itself. The concordance index measures the time periods in which two series are in the same phase (expansion or contraction) relative to all periods for which there are observations. If both series are in the same phase all the time, the concordance index would be 100 per cent and if both series are in the opposite phase all of the time, the index would be 0 per cent. It allows to measure the degree of synchronicity between the different indicators used to obtain the aggregate measure of the financial cycle and the aggregate measure itself. This is done by first calculating the concordance between each indicator in a group and the aggregate measure and then taking the average. The results are displayed in Table 5. Table 5. Concordance Index Indicators included (cyclical components) Financial cycle measure FC1 FC2 FC3 FC4 FC5 FC6 FC7 FC8 FC9 Credit to NFC-to-GDP : : : : : : : : Credit to households-to-gdp : : : : : : : : Credit to NFC (y-o-y) : : : : : : : Credit to households (y-o-y) : : : : : : : Spread, interest rates on credits to NFC : : : : : : Spread, interest rates on credits to households : : : : : : House price index (y-o-y) : : : : : Current account balance-to-gdp : : : : Capital-to-asset ratio : : : : Profits-to-assets ratio Loan-to-deposit ratio : : : : : Concordance index (per cent) Source: own calculations. 20 Harding, D. and A. Pagan, p The concordance index is also used by Stremmel (2015) to assess alternative measures of the financial cycle. 73 Research Topics

75 The degree of concordance between both credit-to-gdp ratios and the aggregate measure FC1 is 97 per cent. This result can be explained by the fact that the two indicators included in the first group share common dynamics during the study period. By including further indicators the degree of concordance between the individual indicators and the aggregate measure declines as expected. This is the case till FC5 is reached, where the concordance index starts growing again and reaches its highest value in the potential measures of the financial cycle FC6 and FC7 78 per cent, which means that the seven, respectively the eight indicators included in the sixth, respectively in the seventh group and the aggregate financial cycle measure extracted from them co-move 78 per cent of the time. The results indicate that with the exception of FC1, which is extracted only from the credit-to-gdp ratios, the best choice of financial cycle measure would be FC6 or FC7, where a broader set of indicators have been taken into consideration in the estimate. The use of an aggregate index based on a wider set of variables which show expansion or contraction phase of the cycle at the same time should take precedence over the use of an indicator such as the credit-to-gdp gap when making decisions on macro-prudential policy. As a whole, these results are in line with those obtained from the principal component analysis. Besides FC6 and FC7, generally all measures from FC5 to FC9 display very close estimates of the degree of concordance between the different indicators included in the respective group and the aggregate financial cycle measure which has been extracted from them. To test the robustness of the results obtained, all possible combinations of the selected indicators in the respective aggregate measures have been tried 21. The calculated concordance indices between the different indicators in each possible combination and the aggregate measures obtained from them do not change the conclusions made so far with financial measures FC6 and FC7 still displaying the greatest concordance, i.e. 78 per cent. As a next step in the analysis the chosen aggregate financial cycle measure FC6 is compared with an estimated business cycle for Bulgaria 22 in order to verify whether there is concordance between them over time. The dynamics of both cycles is displayed in Chart 3. Chart 3. Business Cycle, Financial Cycle Potential Measure in Bulgaria FC6 and Alternatives Note: The gray lines show all possible alternatives to the selected potential measure of the financial cycle in Bulgaria (FC6), derived by combining the selected indicators. Source: own calculations. The graphical investigation of the dynamics of the chosen measure of the financial cycle (FC6) and business cycle leads to the conclusion that both cycles are well synchronised. Another observation which can be made is that both cycles reached the trough at approximately the same time. This was the case first in the second half of 2000 and the beginning of 2001 and the second time at the end of In the expansion phase the financial cycle reached the peak in the second half of 2006, whereas the peak of the business cycle occurred 6 7 quarters later, around the middle of The same conclusion may be drawn when not only the dynamics of the chosen measure of the financial cycle FC6, is considered, but also all alternative measures from FC5 to FC9. It could also be argued that according to the chosen measure of the financial cycle in 2017 the economy is entering the phase of initial cyclical risk accumulation. As already mentioned, due to the normali- 21 In this procedure the indicators measuring credit developments and house prices which are standard in the literature on financial cycles are fixed and the rest of the variables vary. 22 The business cycle estimate has been carried out by means of unobservable components model based on Cobb-Douglas production function. For more details see: Economic Review 2/

76 sation of the indicators used to obtain the synthetic measure of the financial cycle, the crossing of the null axis can be interpreted as a boundary between the phases of cyclical risk accumulation and risk materialisation. An idea of the degree of synchronisation between the chosen measure of the financial cycle and the business cycle may be obtained through the calculation of the respective concordance index between the two cycles. The results of this calculation are displayed in Table 6. For the purposes of comparison the table also shows the respective concordance indices between the business cycle and the remaining potential measures of the financial cycle. Table 6. Concordance Index between the Potential Measures of the Financial Cycle and the Business Cycle in Bulgaria Financial cycle measure Concordance with the output gap (per cent) FC1 49 FC2 48 FC3 71 FC4 65 FC5 72 FC6 71 FC7 72 FC8 71 FC9 71 Source: own calculations. According to Table 6 potential financial cycle measurements FC5 to FC9 display the highest degree of concordance with the business cycle. The calculated concordance index shows that the chosen measure of the financial cycle FC6 and business cycle are in the same phase (expansion or contraction) 71 per cent of the time. In comparison, the potential measure FC1 co-moves with the business cycle only 49 per cent of the time of the studied period. Structural Unobserved Components Model The aim of the structural unobserved components model is to examine the robustness of the financial cycle estimate without imposing any restrictions on the length of the financial cycle. In this model it is assumed that the variation of one observable series, for example real GDP, is a combination of the variation of a group of unobservable components. 23 The observable series can be decomposed into its latent (unobservable) components, i.e. trend and cycle. Each of the latent components is in turn presented in the form of an equation the coefficients of which are assessed within the model. As a result a system of equations is created the combination of which determines the common variation of a specific series. The univariate model decomposes one series while the multivariate model creates a system of equations characterising the individual components of all included observable series. The advantage of using this type of model is that its built-in design allows for greater flexibility in the decomposition of the estimated series. The model used in this study adheres to the methodology described in the literature. Each variable decomposes into a trend, a stochastic cycle component, and a white noise component. The system of equations is as follows: y t = m t + c t + e t (1) m t = b t + m t-1 + v t (2) b t = b t-1 + u t (3) c t = f (j t, j t *) (4) where y t is vector of observable variables, m t is vector of relevant variable trends c t is vector of cyclical component and e t is the error item. The trend m t is defined by equations (2) and (3) and is expressed within the local linear trend model 24. The cyclical component of each individual 23 Harvey, A. C. Forecasting Structural Time Series and the Kalman Filter. Cambridge: Cambridge University Press, The local linear trend model allows for a flexible specification of the trend component in the series. Where the standard deviation of v t > 0 and standard deviation of u t = 0, the trend has the characteristic of a random walk with a constant drift, while where u t > 0 the drift is stochastic; where standard deviation of v t = 0 and standard deviation of u t > 0 the trend is smooth and is defined as an Integrated random walk; where estimated standard deviation of u t = 0 and v t = 0 the trend is a straight line with slope b 1 and initial value m Research Topics

77 series c t is a linear combination of stochastic cycles of all series included in the model and the individual stochastic cycles (j t, j t *) are modelled as a bivariate autoregressive process parametrised in terms of its length and persistence. Within the system of equations correlation between shocks influencing the individual series latent component is assumed to be possible. The model allows for a change in cyclical component phases and extended cyclical component dynamics 25. The defined latent components and the parameters characterising them are estimated using Kalman filter based on maximum likelihood. The use of the model for decomposition of a specific series in its version including a set of observable variables presents several advantages. Parameters determining the dynamics of series trend and cyclical component are estimated within the model and are not limited by predefined values. The estimate used information contained in all series. The model takes into account the correlation between latent components of individual series and their similar dynamics. The variables examined within the model are as follows: seasonally-adjusted GDP 26, total bank credit 27 and house prices 28 in the period. All variables are deflated by the GDP deflator. The decomposition of each series into the respective latent components, determining its dynamics, is estimated both on the basis of univariate model specification, i.e. Independently of the remaining series, and a model that includes the three variables in the same time within a common system. 29 When estimating the decomposition of the three series in a common system, the parameters characterising the stochastic cyclical component of the series for the credit and house prices, are set to an equal value. This results from the assumption that they have similar dynamics and together constitute a measure of the financial cycle. The individual cyclical component of each Chart 4. Estimate of the Cyclical Component of Real GDP in Bulgaria (percentage points) Sources: the NSI, own calculations. Chart 5. Estimate of the Cyclical Component of Real Credit in Bulgaria (percentage points) Sources: the BNB, own calculations. Chart 6. Estimate of the Cyclical Component of Real House Prices in Bulgaria (percentage points) Sources: the NSI, own calculations. 25 Runstler, G. and M. Vlekke. Business, housing and credit cycles. ECB Working Paper Series, no. 1915, 2016, p NSI data. 27 This refers to bank claims on the non-government sector based on BNB monetary statistics. 28 NSI data. 29 The estimate of the model has been done using code provided within the framework of ECB Working Group on Econometric Modelling. Economic Review 2/

78 of the three series then constitutes a linear combination of two stochastic cycles of different length and persistence. The length of the first stochastic cycle which measures the business cycle is estimated to be five years, while the second one measuring the financial cycle, is estimated to be 14 years accordingly. This confirms the assumption that the financial cycle length is between 8 and 30 years on which the band-pass filter estimate is based. It is important to point out that the dynamics of the cyclical components of each of the three series results from the dynamics of two abstract stochastic cycles of different frequencies. Intuitively, this could mean for example that the business cycle, defined as the deviation of real GDP from its potential level, is driven by two cycles passing through the economy at different frequencies. Charts 4, 5 and 6 display the estimates of the cyclical components of the series used. The three estimates are based on the univariate and multivariate unobserved components models and Hodrick-Prescott filter respectively. The multivariate estimate of the financial cycle based on cyclical components of the credit and house prices shows that in 2017 as regards credit, the phase of risk materialisation is coming to its end, while, as regards house prices, an initial phase of cyclical risk accumulation is observed. This confirms to a larger extent the estimate of the financial cycle current phase, made by using band-pass filter and principal component analysis. The univariate estimates show that in 2017 the credit and house price cycle is at the boundary between the phases of risk accumulation and risk materialisation. As regards to the real GDP according to the multivariate estimate the business cycle has been in the expansion phase since the end of 2014, but in 2017 there was no overheating of the economy yet. Concurrently, the univariate estimate indicates that since the end of 2016 the economy is entering the phase of overheating. Conclusion This study presents an assessment of the financial cycle in the Bulgarian economy through two different approaches used in the academic literature. The results show that the most commonly used indicator for measuring the financial cycle, namely the credit-to-gdp gap, gives rather limited information on the dynamics of the financial cycle in Bulgaria. The use of a wider set of indicators covering not only credit dynamics but also private sector debt burden, interest rate spreads, housing price growth, current account deficit as well as indicators for the sustainability of the banking system shows, that in 2017 an initial phase of accumulation of cyclical risk in the economy is observed. This assessment is also confirmed by an alternative approach that uses a structural model including GDP, bank credit and house prices, deflated by the GDP deflator and estimated within an unobserved components framework. The assessment of the phases of the financial cycle in Bulgaria through the aggregation of a wide range of indicators could serve to identify the appropriate moments for the build-up of buffers in the financial system when the cycle is in the upward phase and respectively for their release in the downside phase of the cycle, which will make the system more robust and capable of responding flexibly to shocks. 77 Research Topics

79 ISSN Elements of the 20 lev banknote, issues 1999 and 2007, are used in cover design.