DO THE DRIVERS OF LOAN DOLLARISATION DIFFER BETWEEN CESEE AND LATIN AMERICA? A META-ANALYSIS. Documentos de Trabajo N.º PDF

Description
DO THE DRIVERS OF LOAN DOLLARISATION DIFFER BETWEEN CESEE AND LATIN AMERICA? A META-ANALYSIS 2014 Mariya Hake, Fernando López-Vicente and Luis Molina Documentos de Trabajo N.º 1406 DO THE DRIVERS OF LOAN

Please download to get full document.

View again

of 20
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Creative Writing

Publish on:

Views: 32 | Pages: 20

Extension: PDF | Download: 0

Share
Transcript
DO THE DRIVERS OF LOAN DOLLARISATION DIFFER BETWEEN CESEE AND LATIN AMERICA? A META-ANALYSIS 2014 Mariya Hake, Fernando López-Vicente and Luis Molina Documentos de Trabajo N.º 1406 DO THE DRIVERS OF LOAN DOLLARISATION DIFFER BETWEEN CESEE AND LATIN AMERICA? A META-ANALYSIS DO THE DRIVERS OF LOAN DOLLARISATION DIFFER BETWEEN CESEE AND LATIN AMERICA? A META-ANALYSIS (*) Mariya Hake OESTERREICHISCHE NATIONALBANK Fernando López-Vicente and Luis Molina (**) BANCO DE ESPAÑA (*) This article has been previously published in Focus on European Economic Integration (FEEI) Q1 / 14. (**) Oesterreichische Nationalbank, Foreign Research Division, (corresponding author). Banco de España, International Economics Division, The authors wish to thank two anonymous referees as well as Peter Backé, Markus Eller and Thomas Gruber (all OeNB), Jarko Fidrmuc (Zeppelin University, Germany) and the participants at an internal seminar at the Banco de España in December 2013 for their helpful comments and suggestions. Documentos de Trabajo. N.º The Working Paper Series seeks to disseminate original research in economics and fi nance. All papers have been anonymously refereed. By publishing these papers, the Banco de España aims to contribute to economic analysis and, in particular, to knowledge of the Spanish economy and its international environment. The opinions and analyses in the Working Paper Series are the responsibility of the authors and, therefore, do not necessarily coincide with those of the Banco de España or the Eurosystem. The Banco de España disseminates its main reports and most of its publications via the INTERNET at the following website: Reproduction for educational and non-commercial purposes is permitted provided that the source is acknowledged. BANCO DE ESPAÑA, Madrid, 2014 ISSN: (on line) Abstract In this paper we compare the determinants of loan dollarisation in two emerging market regions, namely Central, Eastern and Southeastern Europe (CESEE) and Latin America, by means of a meta-analysis of 32 studies that provide around 1,200 estimated coefficients for six drivers of foreign currency lending. One common pattern we identify is that macroeconomic instability (as expressed by inflation volatility) and banks funding in foreign currency play a significant role in explaining loan dollarisation in both regions. By contrast, the interest rate differential appears to be a key determinant only in Latin America, while the positive impact of exchange rate volatility on dollarisation implies a more prominent role for supply factors in the CESEE region. While the robustness of the results has been verified, our meta-analysis shows that estimates reported in the literature tend to be influenced by study characteristics such as the methodology applied and the data used. Keywords: foreign currency loans, CESEE, Latin America, meta-regression, random effects maximum likelihood. JEL Classification: C5, E52, F31, O57, P20. Resumen En el presente artículo se comparan los determinantes de la dolarización de los préstamos bancarios en dos regiones, América Latina y Europa Central y Oriental, con una metodología de metaanálisis a partir de 32 estudios previos sobre el tema, que proporcionan unos coeficientes estimados para las 6 principales variables macroeconómicas que influyen sobre la aparición de préstamos en moneda extranjera en las economías. Ambas regiones presentan patrones comunes (la estabilidad macroeconómica y la financiación de los bancos en moneda extranjera juegan un papel relevante a la hora de explicar la dolarización de los préstamos), pero también claras diferencias, como el hecho de que el diferencial de tipos de interés sea una variable determinante tan solo en América Latina, mientras que en Europa Central y Oriental tenga mayor influencia la volatilidad del tipo de cambio. Estas diferencias apuntarían a una mayor prominencia de los factores de demanda en el caso de América Latina, y de oferta en Europa del Este, a la hora de determinar la concesión de un préstamo en moneda extranjera. Palabras clave: préstamos en moneda extranjera, América Latina, Europa del Este, metaanálisis. Códigos JEL: C5, E52, F31, O57, P20. During the 1980s and 1990s, high levels of inflation, wide interest rate spreads, local currency depreciation and the low credibility of domestic economic policies as well as chronic monetary financing of budget deficits prompted massive portfolio shifts into dollardenominated assets and liabilities in most Latin American countries (Galindo and Leiderman, 2005). One decade later, a similar process resulting in a buildup of large stocks of financial assets and liabilities in foreign currency was observed in the European transition economies. While such dollarization 1 may help reduce capital flight, curb inflation expectations and induce macroeconomic stabilization, it may also limit the independence of monetary policy and create systemic vulnerabilities in financial and nonfinancial sectors. The potential adverse effects of dollarization are amplified when firms and households hold unhedged liabilities, in particular bank loans, in foreign currency: this exacerbates credit default risk and currency mismatch and thus creates potential threats to financial stability. Moreover, evidence from emerging economies in general and from Latin America and CESEE in particular reveals that, unless addressed, dollarization tends to be a persistent phenomenon. Yet to be able to achieve dedollarization (i.e. reduce foreign currency-denominated assets) policymakers need to be aware of the key underlying drivers and understand above all whether dollarization was induced by demand- or supply-side factors (EBRD, 2010). The literature on dollarization has identified major determinants of foreign currency lending in emerging market economies, reflecting both demand- and supply-side factors and the interaction between them. These factors include the interest rate differential, the inflation rate and exchange rate depreciation; the volatility of inflation and of the exchange rate as well as the ratio between the two variables (the so-called minimum variance portfolio ratio MVP ratio); and banks funding in foreign currency. 2 At the same time, empirical studies on both Latin America and CESEE have remained rather inconclusive and the results diverge to some extent depending on the countries analyzed, the time period considered or the estimation method used. Against this backdrop, this paper aims to first analyze the main drivers of loan dollarization (i.e. foreign currency lending by banks in the domestic market) in CESEE and Latin America, and to establish whether loan dollarization has been a supply- or a demanddriven process. In a second step, we investigate whether and how the drivers of loan dollarization differ between the two regions. Such a comparison should allow us (i) to identify typical patterns and idiosyncratic factors characterizing dollarization; and (ii) to deduce policy lessons for CESEE from the way dollarization and its consequences were handled earlier in Latin American countries. For that purpose, we conduct a metaregression analysis to condense the findings of previous empirical studies and establish the true effect size across datasets (Stanley and Jarrel, 1989). Our findings suggest that loan dollarization was indeed driven by different factors in CESEE and Latin America. In Latin America, unlike in CESEE, the interest rate spread had a positive and significant impact on dollarization whereas exchange rate volatility had a negative impact, which would imply that Latin American dollarization was demand-driven. Hence, a 1. Dollarization is the (total or partial) replacement of the domestic currency by any foreign currency as a store of value, unit of account or medium of exchange within the domestic economy. Dollarization frequently involves the U.S. dollar, which is widespread in Latin American countries, while the CESEE countries have extensively used the euro and the Swiss franc. In this paper we analyze the dollarization of banks' financial assets, specifically lending to the private nonfinancial sector by banks in the domestic market. 2. We should underline that the literature has identified region-specific factors which might influence the degree of dollarization. In particular, the EU accession perspective and the euro adoption perspective of the CESEE countries have been shown to play a key role (e.g. Rosenberg and Tirpak, 2008). However, in our study we focus on determinants of foreign currency lending which are common to both regions and have a sufficient number of coefficients. BANCO DE ESPAÑA 7 DOCUMENTO DE TRABAJO N.º 1406 rise in exchange rate volatility would make foreign currency loans less attractive for borrowers. In CESEE in contrast, exchange rate volatility had a positive impact, making risk-averse lenders more willing to supply foreign currency loans in order to match their foreign currency positions and reduce their currency risk. In both regions, loan dollarization was, moreover, heavily driven by macroeconomic instability, as reflected by inflation volatility, and banks funding in foreign currency. This paper is structured as follows. Section 1 provides descriptive evidence of financial dollarization, both on the assets and liabilities side in Latin America and CESEE. Section 2 presents a literature review of the determinants of foreign currency lending aimed at identifying the most common explanatory factors at the macroeconomic level. Section 3 describes the meta-analysis framework used to estimate the true size effect of the drivers of loan dollarization. Section 4 discusses the metaregression results and checks their robustness. The last section concludes. BANCO DE ESPAÑA 8 DOCUMENTO DE TRABAJO N.º 1406 1 Descriptive Evidence on Financial Dollarization in Latin America and CESEE 3 Although dollarization has been reduced successfully by some countries in both regions, 4 it tends to be a persistent phenomenon and has indeed been rising in some economies. Yet there are some striking differences between the two regions. First, the degree of currency substitution is higher on average in CESEE than in Latin America, both on the assets and the liabilities side (see charts 1 and 2). Share of Foreign Currency Loans Chart % of total loans Source: National central banks. Note: The data refer to loans to the private nonfinancial sector and are adjusted for exchange rate developments (using January 2008 exchange rates). Data for Brazil and Colombia are not available. In CESEE, 60% of private sector loans and 40% of private sector deposits were denominated in foreign currency in 2012, compared with only 27% and 24%, respectively, in Latin America. The lower dollarization levels in some countries in Latin America are, however, the result of policy or market intervention: In 2001, around 50% of total loans and deposits were denominated in U.S. dollars (or even around 70% in some countries, e.g. Peru and Uruguay). For instance, Argentina officially pesified (dedollarized) and indexed foreign currency loans and deposits after the 2001 crisis. Brazil, Chile, Mexico and Colombia imposed restrictions on holding foreign currency loans, introduced financial instruments indexed to exchange rate and inflation developments, or even implemented government policies to dedollarize public sector liabilities. 5 In Latin America, both loan and deposit dollarization hence decreased constantly from 2000 onward and somewhat stabilized at lower levels during the recent crisis. In contrast, dollarization in CESEE was increasing steadily before the 2008/2009 crisis, fueled by both the EU accession perspective and increasing external funding as well as demand factors (Beckmann, Scheiber and Stix, 2011). The share of foreign currency loans in CESEE continued to increase even after the onset of the 2008/2009 crisis in all countries but 3. In the context of this paper, the CESЕE region includes the eight CESEE EU Member States which have not yet adopted the euro (i.e. Bulgaria, the Czech Republic, Croatia, Hungary, Latvia, Lithuania, Poland and Romania and two (potential) EU candidate countries (i.e. Albania and Serbia). Latin America includes seven countries: Argentina, Brazil, México, Chile, Colombia, Peru and Uruguay. 4. The list of success stories includes Brazil, Chile, Colombia, Mexico and Poland (EBRD, 2010). 5. See Gallego et al. (2010). BANCO DE ESPAÑA 9 DOCUMENTO DE TRABAJO N.º 1406 the Czech Republic, Croatia and Albania. Indeed, the crisis seems to have pushed up dollarization in some CESEE countries. On average, loan dollarization increased by 13 percentage points in the region as a whole between 2008 and Second, the degree of regional divergence differs as well. In Latin America, the share of foreign currency loans in total loans outstanding in 2012 ranged from 11% (in Argentina and Mexico) to around a 50% (in Peru and Uruguay), while the respective shares in CESEE ranged from 10% (the Czech Republic) to close to 90% (Latvia). 6 Furthermore, in CESEE, the share of foreign currency deposits was as high as 60% to 75% in the majority of the countries analyzed, with only one country (the Czech Republic) exhibiting a share clearly below 15% of total deposits. In contrast, in Latin America, five of the seven countries analyzed registered a ratio below 15%. Share of Foreign Currency Deposits % of total deposits Chart Source: National central banks. Note: The data refer to deposits made by the private nonfinancial sector and are adjusted for exchange rate developments (using January 2008 exchange Third, regarding potential drivers of loan dollarization, a major difference between the two regions is the degree of currency mismatch in the respective banking systems (i.e. the difference between the level of loans and deposits in foreign currency as a share of GDP; see chart 3 7 ). The banking systems in CESEE as defined here tend to be dollarized more heavily on the assets side than on the liabilities side. The currency mismatch is high and positive, having evolved over time from 1% of GDP on average in the early 2000s to around 15% in 2008, due to an extraordinary increase of foreign currency loans. From 2008 onwards dollarization decreased strongly as the crisis affected both foreign currency loan demand and supply, especially in countries like Hungary. Only in Albania and the Czech Republic is the sign of the mismatch negative (i.e. foreign currency deposits exceed foreign currency loans). In Latin America in contrast, the cross-country correlation between U.S. dollar loans and U.S. dollar deposits was close to 1 in 2012, following a decline during the 2000s. Within Latin America, Uruguay is an outlier, with a negative currency mismatch of 40% of GDP in 2012, 6. On January 1, 2014, Latvia became the 18 th euro area member state. 7. Yet we do not have data on assets and liabilities different from loans and deposits in foreign currency held by banks. If we account for those other assets and liabilities, the currency mismatch may be amplified or reduced. For instance, banks may hedge net short positions in loans-deposits with long positions in other dollar-denominated assets and, therefore, match their foreign currency positions, reducing or at least balancing the indirect exchange rate induced risk. BANCO DE ESPAÑA 10 DOCUMENTO DE TRABAJO N.º 1406 reflecting the absorption of substantial amounts of U.S. dollar deposits from Argentina after the crisis in the early 2000s. Dollarization Mismatch between Foreign Currency Loans and Deposits Chart % of GDP Source: National central banks. Note: The mismatch is measured as the difference between foreign currency loans and foreign currency deposits as a % of GDP. Data for Brazil and Colombia are not available. Fourth, the degree of dollarization is also reflected by foreign currency holdings abroad and the issuance of foreign currency debt in international markets. Such offshore dollarization is seen as less damaging than domestic dollarization, since the default risk is transferred to foreign institutions, although it usually reveals deficiencies in the domestic credit markets and distrust in the banking system. Yet for most of the CESEE countries offshore deposits represent only a small fraction of total deposits and have decreased in the sample period. In Latin America, offshore deposits are more relevant but have also decreased from the early 2000s (chart 4). Corporate issuance of foreign currency debt has gained relevance and grown exponentially in both Latin America and CESEE, as the accommodative stance of monetary policy in developed countries has sharply reduced funding costs in international Offshore Deposits Chart % of total deposits Source: BIS BANCO DE ESPAÑA 11 DOCUMENTO DE TRABAJO N.º 1406 markets for foreign currency loans in domestic markets. The pattern in the two regions is very similar: an increase of corporate issuance in international markets and in foreign currency. In absolute figures, the importance of foreign funding sources remains limited for these economies, though (around 2% of GDP and 5% of total bank credit in both regions). 8 Finally, the countries in the two regions differ somewhat with respect to exchange rate and inflation rate developments and volatilities as well as with regard to the interest rate differential (i.e. the difference between the price of loans in foreign and in domestic currency). 9 Interestingly, while the interest rate differential (chart 5) has stabilized or decreased in some countries with a high degree of dollarization in both regions (e.g. Peru and Uruguay; Croatia and Albania), it remains at elevated levels of up to 10 percentage points difference in other highly dollarized countries in both regions (e.g. Serbia and Argentina), not least due to the persistently high inflation rates in these countries. Inflation volatility has decreased in all countries under review since 2005 (chart 6), with the exception of Latvia, which nevertheless registered very low inflation rates and even some episodes of deflation in recent years. Going further, although the majority of countries have seen their exchange rates appreciate since 2001, partly explained by the increase in income per capita and related to the Balassa- Samuelson effect, some differences arise in terms of exchange rate volatility, which decreased strongly in CESEE countries and has increased slightly in those Latin American countries with inflation targeting. 10 Differential between Interest Rates for Local Currency and Foreign Currency Operations Interest Rate Differential: Deposits Interest Rate Differential: Lending Percentage points Percentage points Chart Albania Bulgaria Czech Republic Croatia Hungary Latvia Lithuania Poland Romania Serbia Argentina Mexico Chile Peru Uruguay Albania Bulgaria Czech Republic Croatia Hungary Latvia Lithuania Poland Romania Serbia Argentina Mexico Chile Peru Uruguay Source: IMF International Financial Statistics and national central banks. 8. Data for fixed income issuance come from the Dealogic database and cover all corporate bonds and medium-term notes placed by domestic firms and sovereigns in domestic and international markets. 9. The majority of studies included in section 4 use as a proxy for the interest rate differential a somewhat different calculation, the difference between the domestic interest rate and the U.S. or euro area interest rate, probably as it is difficult to recover long time series data for these differentials, and as some of the domestic markets for foreign currency loans or deposits were developed only from 2000 onwards. 10. Inflation and exchange rate volatility can be calculated in different ways. The papers included in the next section use both rolling standard deviations of inflation rates or volatility extracted using statistical models like GARC
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks