Based on Latin American Economic Outlook 2009 London February 2009 Latin America: The Fiesta is Over Javier Santiso Director and Chief Economist OECD Development Centre

The financial crisis and Latin America Historically , when the U.S. sneezes, Latin America catches cold

Current account channel: trade and remittances Remittances (%, annual growth) Source: OECD Development Centre based on World Bank, November 2008 Latin American Exports (by destination) Source: OECD Development Centre based on WITS data, 2008 65% The financial crisis and Latin America Source: OECD Development Centre based on WITS 2008.

Current account channel: Reduction of commodity prices and trade balance Terms of trade gains from 2003 through 2008 (%) Commodities Exports (% of total exports) The financial crisis and Latin America Source: OECD Development Centre based on WITS 2008. Source: OECD Development Centre based on JP Morgan, 2009. Source: OECD Development Centre based on National Statistics and WITS, 2009.

Current channel: Chinese trade has decreased sharply over the last months Source: OECD Development Centre, based on national sources and Thomson, 2009 The financial crisis and Latin America

The financial crisis and Latin America Financial channel: foreign banks in local markets Structure of Mexican banks (% of total assets) Source: OECD Development Centre based on World Bank, November 2008 Market share of Foreign Banks in local banking systems (% of total assets) Source: OECD Development Centre based on WITS data, 2008 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% Mexico Chile Venezuela Brazil Argentina Colombia Peru Source: OECD Development Centre based on National Sources and Credit Suisse, 2008 20% 24% 15% 20% 11% 5% 5% Mexican Banks BBVA Santander Citibank HSBC Scotiabank Other Foreign Source: OECD Development Centre based on National Sources and Credit Suisse, 2008

Initial impact has been limited Source: OECD Development Centre, based on IMF (WEO October 2008) Output gap (%) GDP (% annual growth) Source: OECD Development Centre, based on IMF (WEO October 2008) thanks to credible economic policies and high internal demand

However this has increased recently Inflation (% y-o-y) Industrial production (% annual growth; moving average) Source: OECD Development Centre, based on Thomson Datastream, 2009 Source: OECD Development Centre, based on Thomson Datastream, 2009 But a monetary stimulus could be possible in some Latin American countries

GDP growth forecasts are on the downside Slowdown is not necessarily recession GDP forecasts for 2009 and 2010 (% annual growth) Source: OECD Development Centre, based on Consensus Forecasts, 2009 Source: OECD Development Centre, based on Consensus Forecasts, 2009 GDP forecasts for 2009 in Latam (% annual growth)

Latin America remains highly vulnerable to international capital markets’ volatility … Nominal exchange rate (US$, basis 100 = Nov.07) Equity market (basis 100 = Nov.07) Source: OECD Development Centre based on Thomson Datastream, 2009 Source: OECD Development Centre based on Thomson Datastream, 2009

… but at least for the public sector, the sensibility has been reduced Sovereign bond spreads and “global risk aversion” Source: OECD Development Centre based on Thomson Datastream, 2009 Spreads are now much less sensitive to external shocks The VIX index (CBOE Volatility Index) is the implied volatility of S&P500 index options.

Capital markets are differentiating between countries and with respect to past sovereign debt crises Markets do not appear to be anticipating a crisis for ‘credible’ countries EMBI sovereign bond spreads (bp) Source: OECD Development Centre based on Thomson Datastream, 2009 Source: OECD Development Centre based on Latin American Economic Outlook, 2009 today Fees and Sovereign bond spreads (Primary market)

Why such confidence? Public debt management has improved “ Original Sin Index” External Public Debt in Latin America (% of GDP) Source: OECD Development Centre calculations based on Dealogic, 2009. External debt is falling and countries can increasingly borrow abroad in local currency Source: OECD Development Centre calculations based on World Bank and Consensus Forecast, 2009. 0,0 0,2 0,4 0,6 0,8 1,0 2002 2003 2004 2005 2006 2007 - 2008 Brazil Colombia Peru Uruguay

However refinancing risks exist in a context of illiquidity in the international markets Corporate Bond High-Yield Issuance (US$ mill.) Global Covered Bond Issuance (US$ mill.) Source: OECD Development Centre based on Dealogic, 2009. This is above all true for the Latin Corporates ( External maturities reach $32bn in 2009) Source: OECD Development Centre based on Dealogic, 2009.

Revenues and expenditures Fiscal policy volatility Latin America’s fiscal performance has much improved over recent years Fiscal policy measures Source: OECD Development Centre calculations based on the ECLAC ILPES Public Finance database for Latin America, and OECD General Government Accounts data for OECD countries . Source: OECD Development Centre calculations based on the ECLAC ILPES Public Finance database for Latin America, and OECD General Government Accounts data for OECD countries .

Fiscal policy volatility

Gini coefficients of income inequality, before and after taxes and transfers While taxes and transfers reduce the inequality by 19 Gini points in Europe, the difference is less than two Gini points in Latin America However, fiscal policy does little to reduce inequality in Latin America Source: OECD Development Centre (2008), based on Euromod (2008) for OECD countries and Goñi et al. (2008) for Latin America.

Democratic Consolidation in Latin America: Experts’ and Citizens’ Views Sources: BTI Index (2008) and Latinobarómetro (2007). Political context: democratic consolidation… Source: OECD Development Centre based on Nieto Parra and Santiso (2008) Number of presidential elections in Latin America The Latin American countries covered are Argentina, Brazil, Chile, Colombia, Ecuador, Mexico, Peru and Venezuela. For elections with a second (run-off) round, the date of the final round is used.

Capital Markets, Democracy and the Cost of Debt There are clear links between political processes, particularly elections, and instability in the debt and currency markets The political cycle and capital markets Real Exchange Rates around Elections Correlation Between Exchange Rate and Sovereign- bond Spreads During Elections Source: Nieto Parra and Santiso (2008a), based on Datastream database.

There are clear links between political processes, particularly elections, and instability in the debt and currency markets

Capital Markets, Democracy and the Cost of Debt Bank recommendations are downgraded prior to elections, but tend to move positively again once the uncertainty has passed The political cycle and capital markets Investment-Bank recommendations around elections Brazil 2002 and 2006: From Lula Preta to Lula de Mel Source: Nieto Parra and Santiso (2008a), based on Datastream database.

Bank recommendations are downgraded prior to elections, but tend to move positively again once the uncertainty has passed

Capital Markets, Democracy and the Cost of Debt Bank recommendations are downgraded prior to elections, but tend to move positively again once the uncertainty has passed The political cycle and capital markets Source: Nieto Parra and Santiso (2008a), based on Datastream database.

Bank recommendations are downgraded prior to elections, but tend to move positively again once the uncertainty has passed

Capital Markets, Democracy and the Cost of Debt The market’s fears have some justification: elections are indeed associated with fiscal volatility Are Capital-market Jitters around Elections Justified? Impact of Elections on Fiscal Policy, 1990-2006 (% of GDP) Note: The impact of elections on fiscal policy is calculated as the difference between the fiscal variable (as percentage of GDP) during the election year and non-election years. Source: Nieto Parra and Santiso (2008b, forthcoming)

The market’s fears have some justification: elections are indeed associated with fiscal volatility

Capital Markets, Democracy and the Cost of Debt Across individual Latin American countries, it exposes considerable variation Are Capital-market Jitters around Elections Justified? Impact of Elections on Fiscal Policy, 1990-2006 (% of GDP) Note: The impact of elections on fiscal policy is calculated as the difference between the fiscal variable (as percentage of GDP) during the election year and non-election years. Source: Nieto Parra and Santiso (2008b, forthcoming)

Across individual Latin American countries, it exposes considerable variation

Capital Markets, Democracy and the Cost of Debt The economic pronouncements made by Latin American electoral candidates differ markedly among countries Uncertainty around elections: the role of economic-policy Platforms Non-credible policies announced by Candidates, 1998-2008 Source: Nieto Parra and Santiso (2008a), based on The Economist (2008).

The economic pronouncements made by Latin American electoral candidates differ markedly among countries

Non-credible policies announced by Candidates, 1998-2008

Is Latin American Democracy Maturing In the eyes of the Capital Markets? The election effect has been much less marked since 2006 than before... Investment-bank recommendations around elections by year, Latin America There were elections in 2006 in all the large countries of Latin America other than Argentina. Source: Nieto Parra and Santiso (2008a)

The election effect has been much less marked since 2006 than before...

Investment-bank recommendations around elections

by year, Latin America

Is Latin American Democracy Maturing In the eyes of the Capital Markets? ...and is partially supported by higher primary surplus due to high GDP growth rather than spending restraint, and there has been no decrease in non-credible policy pronouncements (e.g., Argentina, Mexico, Peru and Venezuela). Impact of Presidential Elections on Fiscal Variables (% of GDP) Note: The impact of 2005 and 2006 elections on fiscal policy is calculated as the difference between the fiscal variable (as a proportion of GDP) during the election year and prior non-election years. Source: Nieto Parra and Santiso (2008b, forthcoming)

...and is partially supported by higher primary surplus due to high GDP growth rather than spending restraint, and there has been no decrease in non-credible policy pronouncements (e.g., Argentina, Mexico, Peru and Venezuela).

Impact of Presidential Elections on Fiscal Variables (% of GDP)

Public debt management and political cycles Debt levels still high, but considerable progress has been made. Good public debt management is linked to healthy development of the domestic bond markets in the region Sovereign bonds markets have been keenly sensitive to political events: Why? - Investors worry about pre-election fiscal expansion to attract voters (1990-2006, impact of general elections on fiscal expenditure in LA come close to 25% of GDP growth; virtually zero in OECD), with negative post-electoral consequences… - … and that candidates economic policy platforms are not credible The political dimension of public debt

Debt levels still high, but considerable progress has been made.

Good public debt management is linked to healthy development of the domestic bond markets in the region

Sovereign bonds markets have been keenly sensitive to political events: Why?

- Investors worry about pre-election fiscal expansion to attract voters (1990-2006, impact of general elections on fiscal expenditure in LA come close to 25% of GDP growth; virtually zero in OECD), with negative post-electoral consequences…

- … and that candidates economic policy platforms are not credible

Based on Latin American Economic Outlook 2009 London February 2009 Latin America: The Fiesta is Over Javier Santiso Director and Chief Economist OECD Development Centre