Convergence in the Baltics: From Boom to Bust?

Convergence in the Baltics: From Boom to Bust? Christoph Rosenberg IMF Office for Central Europe and the Baltics CASE Warsaw, September 3, 28

EU15: Population: 332.1 mn GDP PPP per capita: 27.6 Population: 1.3 mn GDP PPP per capita: 17,8 Population: 2.3 mn GDP PPP per capita: 14,4 Population: 3.4 mn GDP PPP per capita: 14,9

Outline What drove the Baltic boom and why is it coming to an end now? Vulnerabilities: external imbalances and foreign exchange mismatches Why the Baltics are different How to achieve a soft landing avoiding the Portuguese trap

Income convergence has been faster than in other new member states and the rest of the world Convergence in Emerging Europe and in the Rest of the World, 22 6 Average growth of PPP per capita GDP, 22 6 2 15 1 5 MD UA BY RO TR BA AL MK LV RS LT BG HU EE RU SK PL CZ 6 7 8 9 1 11 12 ln(ppp per capita GDP), 22 Rest of the world Europe Linear (rest of the world) Linear (Europe) HR 2 15 1 5 Source: IMF REO 27.

The boom has been fueled by very high credit growth Growth in GDP and credit to the private sector annual average 5 45 4 35 3 25 2 15 1 5 GDP Credit to the private sector 98-1-3 4-7 98-1-3 4-7 98-1-3 4-7 98-1-3 4-7 EST LAT LIT CE4 Source:Eurostat, national authorities.

which has inter alia been related to negative (perceived) real interest rate Domestic and foreign currency lending rates, deflated by wage growth, 27 5 3-month domestic currency rate deflated by wage growth 3-month Eurozone currency rate deflated by wage growth -5-1 -15-2 EST LAT LIT CZK HUN POL SVK Source: National authorities, Eurostat and IMF staff calculations.

Procyclical fiscal policies have added oil to the fire 3. 2. 1.. -1. -2. -3. -4. Change in Actual and Structural Fiscal Balances, 23 7 1/ Percent of GDP Change in structural balances Change in actual balances -5. LAT EST LIT CE4 Source: IMF, World Economic Outlook; and IMF staff calculations. 1/ The structural balances do not take into account spending related to pension reform and spending related to EU transfers.

And so have loose monetary policies Emerging Europe: Lending Interest Rate minus ideal suggested by Taylor Rule, 23 7 6 6 2 Southeastern Europe Central-eastern Europe 2-2 -2-6 Baltics -6-1 -1-14 -14 Other emerging Europe -18 23 24 25 26 27-18

The overheating has contributed to consumer price inflation 2 CPI, end of period percent 15 LAT 1 EST LIT 5 World CE4-5 Source: National authorities. 1997 1998 1999 2 21 22 23 24 25 26 27 Jul-8

and soaring real estate prices Residential property prices YoY change in local currency deflated by CPI, period averages 35 3 1-3 4-7 25 2 15 1 5 EST LAT LIT CE4 Source:UniCredit Group, Regional Overview.

In mid-27 the growth came to an 14 GDP growth annual percentage change abrupt halt 12 1 LAT 8 LIT 6 4 CE4 2 EST -2-4 95 96 97 98 99 1 2 3 4 5 6 7 8Q1 Source: National authorities.

mainly driven by a slow-down of credit growth engineered by Swedish parent banks 7 Credit to the private sector annual percentage change 6 5 4 3 2 EST CE4 LAT LIT 1-1 98 99 1 2 3 4 5 6 7 8Q2 Source: National authorities.

Two key vulnerabilities emerged during the economic boom: Large external imbalances (double-digit current account deficits, highly negative international investment position) Large currency mismatches (especially in the balance sheets of households and corporates)

The credit boom was mainly funded by a build-up of external debt 7 Change in external debt an and credit to the private sector, 22-27 in GDP percentage points Change in external private debt Change in credit to private sector 6 5 4 3 2 1-1 EST LAT LIT CE4 Sources: National authorities; and IMF staff calculations.

The overheating of the economy led to exceptionally high current account deficits. 1 5 Current account deficit in percent of GDP Emerging Asia Latin America -5-1 -15-2 -25 CE4 LIT EST LAT 2 1 2 3 4 5 6 7 8Q1 Source: National authorities.

Since the onset of the global credit crisis, markets have viewed countries with large imbalances as more risky 35 3 Ukraine Change in CDS since July 27 25 2 15 1 5 Latvia Bulgaria Iceland Romania Estonia Lithuania South Africa Turkey Hungary Mexico Poland Brazil Greece Slovakia Portugal Czech Republic Russia China -3-25 -2-15 -1-5 5 1 15 C/A deficit, 27 Source: Bloomberg.

Determining CA sustainability in the new member states (Rahman, IMF WP 8/92) Macroeconomic balance approach: estimate current account norms based on macro fundamentals (59 countries, annual data 1971-27) CA norm = f (fiscal balance, demographics, net foreign assets position, relative per capita income, etc.) CA norm in the new member states are higher than elsewhere because: Large share of dependent population (=> lower national savings => higher CA) Lower net foreign asset position (greater dependence on foreign capital for growth)

But CA deficits in the Baltics still exceed what would be consistent with macro fundamentals Current Account Balances and Model Predictions Percent of GDP -2-2 -4-4 -6-6 -8-8 -1-12 Current account model predictions, emerging Europe -1-12 -14 Current account balances, emerging Europe, average 23-7 -14-16 -18 Latvia Estonia Bulgaria Romania Current account model predictions, Asia Lithuania Hungary Slovak Rep. Czech Rep. Poland Slovenia -16-18 Source:Rahman 28.

What drives the divergence from CA norms? Cyclical or temporary factors: output gap, capital inflows due to EU accession Structural factors: export composition, cost competitiveness in manufacturing sector Policy-related factors: exchange rate regime, financial sector policies

The Baltics have built up large currency mismatches Net FX position, 27 percent of GDP 2 1-1 -2-3 -4-5 EST LAT LIT CZK POL HUN Change in net FX position, 22-27, percent of GDP 2 1-1 -2-3 -4 EST LAT LIT CZK POL HUN Sectoral net FX position, 27, percent of GDP 4 2 Government Banks Other sectors -2-4 -6-8 EST LAT LIT CZK POL HU N Source: National authorities, IMF staff calculations.

especially in the household sector Latvia: Fx liabilities and assets of the household sector Assets Liabilities percent of GDP percent of GDP 4 4 35 35 3 3 25 25 2 2 15 15 1 1 5 5 27q1 27q3 2q1 2q3 21q1 21q3 22q1 22q3 23q1 23q3 24q1 24q3 25q1 25q3 26q1 26q3 27q1 27q3 28q1 2q1 2q3 21q1 21q3 22q1 22q3 23q1 23q3 24q1 24q3 25q1 25q3 26q1 26q3 28q1 Source: CNB.

Euroization is higher than in other new member states Credit to GDP ratio in local currency vs foreign currency Chart 4. Credit-to-GDP ratio in local currency vs foreign currency (in %, year 27) in %, year 27 8. 8. 7. Latvia Estonia Group A 7. credit-to-gdp ratio in foreign currency 6. 5. 4. 3. 2. Group C Romania Croatia** Slovenia* Bulgaria Lithuania Hungary 6. 5. 4. 3. 2. Group B Poland 1. Slovakia 1. *26, **3q27 Czech Republic Source: national authorities and IMF staff calculations. 45'... 1. 2. 3. 4. 5. 6. 7. 8. credit-to-gdp ratio in local currency Source: National authorities, IMF staff calculations.

What drives euroization in the new member states? (Rosenberg and Tirpak, IMF WP 8/173) Hypothesis: Share of the fx loans depends on: interest rate differential availability of fx-denominated funding through foreign parent banks GDP level currency regime EU/EMR2 membership openness (export/gdp, remittances) regulatory measures

Index of regulatory measures: Index = policy i, t i, t Policies to discourage foreign currency borrowing (policy i,t ) Monitoring FX risk Disclosure FX risks to customers Tightening eligibility criteria for FX borrowing Higher risk weights/provisioning/reserve requirements depending on banks' FX exposure Ceilings on banks' FX exposure Score.2.4.6.8 1.

Empirical investigation fxloans =α+β irdiff +β loantodep +β openness +β restrict + X +ε it, 1 it, 2 it, 3 it, 4 it, it, it, Xi,t: variable tested, but not included in the preferred model: GDP per capita, asset share of foreign banks, size of the economy, EBRD index of banking sector reform, actual exchange rate volatility, and dummies for ERM2 and EU membership Table 3. Estimated coefficients from our preferred model Domestic banks only Incl. cross-border loans A B Interest rate differential.136***.194*** Loan-to-deposit ratio.17195***.7128** Openness.116*.14416** FX restriction index (lagged) -.1317* -.653 Note: *; **; *** refers to significance at 1%, 5%, and 1% level respectively. Discussion on the model specification is in the Appendix II. Time dummies are included. Source: Authors' calculations.

Interest rate differentials matter, especially in the Baltics where exchange rate risk is perceived to be low 1 8 Lithuania: Figure 12. Lithuania: Share of Share foreign of foreign currency currency loans loans (in EUR) (in EUR) on housing on housing loans loans and and interest rate differential New housing loans, in% (LHS) Total housing loans, in % (LHS) Interest rate spread over EUR loans, in p.p. (RHS) 2. 1.5 6 1..5 4. 2 -.5 Source: Bank of Lithuania and IMF staff calculations. Oct-4 Jan-5 Apr-5 Jul-5 Oct-5 Jan-6 Apr-6 Jul-6 Oct-6 Jan-7 Apr-7 Jul-7 Oct-7 Jan-8-1.

Preliminary conclusions: Euroization is a byproduct of convergence. EU membership boosts foreign exchange borrowing through multiple channels: it offers better access to foreign funds in a fully liberalized environment of capital flows, it provides natural hedging opportunities, through increasing trade openness, it may boost private sector s confidence in exchange rate stability and imminent euro adoption. Regulatory measures have limited effectiveness due to opportunities to borrow directly from abroad (i.e., for corporations).

Will the Baltics follow the Portuguese path? Rapid convergence after joining the EU Since 21, sub-par growth and high C/A deficits Wage increase exceeded productivity growth and undermined competitiveness Or relive the Asia experience? Strong growth, driven by foreign borrowing In 1997, sudden stop of capital inflows, fixed currency regimes collapsed Painful recession aggravated by currency mismatches, but relatively fast recovery

What makes the Baltics different? EU membership has created circumstances that are different from the textbook case: Very open capital markets, role of Nordic parent banks Labor is highly mobile (emigration) Strong commitment to currency pegs, perceived to be backed by EU

Trade and financial integration has accelerated 16 14 Estonia 12 1 8 1996-2 21-26 6 4 2 Change in trade openness 1/ Change in financial deepening 2/ Change in financial openness 3/ 1 8 6 4 1996-2 21-26 Latvia 2-2 Change in trade openness 1/ Change in financial deepening 2/ Change in financial openness 3/ 6 5 4 3 1996-2 21-26 Lithuania 2 1-1 -2 Change in trade openness 1/ Change in financial deepening 2/ Change in financial openness 3/ Sources: World Economic Outlook; International Financial Statistics; country authorities; and Fund staff calculations. 1/ Trade openness is defined as total trade (imports plus exports) in percent of GDP. 2. Financial deepening is defined as cedit to private sector in percent of GDP. 3. Financial openness is defined as the sum of total external asset and toal external liabilities in percent of GDP.

Foreign bank presence offers protection against a sudden Asia-style withdrawal Asset share of foreign owned banks percent EST LAT LIT POL CZK HUN SVK 26 99.1 62.9 91.8 74.3 84.7 82.9 97. Source: EBRD.

but it also exposes the Baltics to common-lender contagion risk 1% Concentration of Emerging Europe Exposure to Western Europe, 27 (Percent) 8% 6% 4% 2% % BA BY AL SK HR RO MD CZ UA HU BU RU PL MK LV LT EE Austria Italy Germany France Sweden Switzerland Netherlands Other Source: Bank for International Settlements, Quarterly Review, June 28. Note: Country names are abbreviated according to the ISO standard codes. 1/ Emerging Europe exposure to western European banks is defined as the share of the reporting banks in each western European country in the total outstanding claims on a given emerging European country (both bank and nonbank sectors). For example, about 42 percent of Croatia's exposures to Western European reporting banks is owed to Austrian banks, 38 percent to Italian banks, 13 percent to French banks, etc. For the Baltic countries, 85 percent or more of exposures to the reporting banks is owed to Swedish banks.

Labor markets are flexible, but emigration has led to tight supply conditions 7 Employment rate (in percent of working age population) EST LAT 65 LIT 6 CE4 55 2 21 22 23 24 25 26 27 Source: Eorostat

...and, until recently, made it more difficult to contain real wage growth. Annual real wage growth percent 2 15 LT 1 EE 5 CE4 LV -5-1 1999 2 21 22 23 24 25 26 27 28Q1 Source: Eurostat.

The fixed exchange rate regime have weathered the loss in confidence well 13 12 11 Baltics: Exchange rate per euro, 27-28 Estonia Spot and Forw ard Rates 1/ (Central parity f 15.6466=1) Spot 3-month 6-month 1 99 16 15 14 13 12 11 1 99 98 Central parity Central parity-1% Latvian Spot and Forw ard Rates 1/ (Central parity f.7284=1) Central parity+1% Spot 3-month 6-month 2-Jan 2-Feb 2-Mar 2-Apr 2-May 2-Jun 2-Jul 2-Aug 2-Sep 2-Oct 2-Nov 2-Dec 2-Jan 2-Feb 2-Mar 2-Apr 2-May 2-Jun 2-Jul 2-Aug 12 11.5 11 1.5 1 99.5 99 Lithuanian Spot and Forw ard Rates 1/ (Peg of 3.4528=1) Spot 3-month 6-month 2-Jan 2-Feb 2-Mar 2-Apr 2-May 2-Jun 2-Jul 2-Aug 2-Sep 2-Oct 2-Nov 2-Dec 2-Jan 2-Feb 2-Mar 2-Apr 2-May 2-Jun 2-Jul 2-Aug 2-Jan 2-Feb 2-Mar 2-Apr 2-May 2-Jun 2-Jul 2-Aug 2-Sep 2-Oct 2-Nov 2-Dec 2-Jan 2-Feb 2-Mar 2-Apr 2-May 2-Jun 2-Jul 2-Aug Source: Bloomberg. 1/ Calculated cross rate as the product of currency US$-euro rate, using average of bid and ask prices.

How can a soft landing be achieved? 1. Fiscal policies should not seek to offset a contraction in demand by discretionary easing (this was Portugal s mistake). 2. Facilitate the switch of production and investment from non-tradables (retail, real estate) to tradables (manufacturing, tourism). 3. Wages should be flexible. Inward migration to the Baltics could help. 4. Strengthen financial supervision, crossborder cooperation and financial safety nets. Ensure that banks maintain adequate capital and liquidity.

The global twin crisis a blessing in disguise The credit crunch reduces the access to cheap financing and serves as a reminder that no boom lasts forever Dampen expectations with regard to rapid wage growth, employment, large-scale public investment, speedy euro adoption The inflation spike helps to reduce real wages (if nominal wage growth can be contained)

Thank you! Dr. Christoph B. Rosenberg Senior Regional Representative International Monetary Fund Regional Office for Central Europe and Baltics Ul. Zielna 37C -18 Warszawa Tel.: + 48 22 338 67 E-mail: cee-office@imf.org Visit our website at: www.imf.org/cee