## 1. Asia and Latin America: How Deep are their Linkages?

## Source details

**Canonical URL:** [1. Asia and Latin America: How Deep are their Linkages?](https://www.imf.org/-/media/files/publications/reo/whd/2018/october/wreo1018-background-papers.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/reo/whd/2018/october/wreo1018-background-papers.pdf.md)
- [Structured JSON version](/-/media/files/publications/reo/whd/2018/october/wreo1018-background-papers.pdf.json)

---

### Trade: key findings and implications
- LAC trade with Asia has increased over the last 20 years, at the expense of trade with North America.
- Asia is the second largest trading partner for LAC, both in terms of imports and exports; within Asia, China is the main destination.
- Major country roles:
  - Brazil, Chile, Mexico, and Argentina are the main LAC exporters to Asia.
  - Mexico is the main importer from Asia (partly explained by its role in global value chains).
- Composition of trade:
  - LAC exports to Asia are highly concentrated in a small number of commodities, particularly minerals (copper and iron ores) and vegetables (soybeans and soybean oil).
  - LAC imports from Asia are concentrated in machines and electronics (40 percent of total import from Asia to LAC).
- Structural drivers:
  - Complementarity between China’s historical demand for raw materials and LAC’s natural-resource endowment explains much of the LAC→Asia trade.
  - China’s exports of final goods to LAC reflect China’s abundant low-cost labor and manufacturing capabilities.
- Risks and opportunities from China’s rebalancing:
  - China’s shift from investment- to consumption-led growth may reduce demand for mineral intermediate inputs while maintaining demand for soybeans and meat products.
  - Chile and other Andean exporters are particularly exposed to a potential deceleration in Chinese demand for minerals.
- Policy recommendations:
  - Diversify exports.
  - Increase value added.
  - Implement policies supporting trade integration and reduced trade costs to facilitate market reorientation.

### Investment: magnitude, composition, and impediments
- Portfolio and FDI magnitudes:
  - In LAC, portfolio investment is heavily biased towards North America; only 2 percent is allocated to assets issued by Asian countries.
  - Assets issued by LAC countries represent only 1 percent of Asia’s portfolio.
  - Asia holds 4 percent of the total FDI stock in LAC; LAC contributes 1 percent to the stock of FDI in Asian countries.
  - These portfolio and FDI compositions have remained stable over time.
- Geographic and country concentration:
  - Japan and South Korea are the key players:
    - Japan and South Korea together represent 86 percent of LAC’s holdings of Asian securities, with Chile and Brazil being the main investors.
    - Japan and South Korea account for 89 percent of LAC securities held in Asia’s investment portfolio.
    - Asia’s FDI in LAC: 60 percent originates from Japan and 20 percent from South Korea; destination concentration: 46 percent in Brazil and 32 percent in Mexico.
    - In LAC investments in Asia, Singapore is the main destination, accounting for almost 60 percent of the total investment position.
- Nature and sectoral patterns of FDI:
  - Most FDI between Asia and LAC corresponds to greenfield investments; interregional merger and acquisition activity remains very limited.
  - Asian greenfield investments in South America (Andean region) focus on mining and petroleum.
  - Investments to Central America leverage low-cost labor and proximity to the U.S. in apparel and offshore business services.
  - Market-seeking investments (automotive, business electronics) target Brazil’s large domestic market; Mexico’s investments reflect its status as an export platform for U.S.-destined products.
  - LAC investments in Asia primarily target fast-growing consumer markets in financial, IT, and business services; manufacturing investments are very rare.
- Factors restricting interregional investment:
  - Limited promotional efforts between regions.
  - Small markets and undiversified exports in many LAC countries.
  - Unfavorable business environments and institutional factors; some Asian and LAC economies have among the most restrictive FDI regimes in the world (per OECD FDI Regulatory Restrictiveness Index).
  - Regulatory restrictions and FTAs without investment chapters.
  - Informational advantages of Asian investors toward Asian emerging markets reducing portfolio flows to LAC.

### Other financial linkages and contagion channels
- Direct banking linkages:
  - Direct cross-border banking linkages are very limited and mostly involve lending by Japanese, South Korean, and Taiwanese banks to LAC.
  - Japan accounts for about 80 percent of the entire bank exposure of Asia to LAC, concentrated in Brazil and Mexico and offshore financial centers such as Panama and Bermuda.
- Indirect linkages:
  - Indirect financial linkages (proxied by the fund competition index) are stronger than direct ones but still smaller than intra-regional linkages.
  - Significant indirect linkages exist between Asia and some LAC emerging markets (Argentina, Colombia, Panama) through common exposure to the U.S. banking system.
  - Intraregional competition for global funds is high (Asia-Asia and LAC-LAC); interregional competition (LAC-Asia) is low, with exceptions noted above.
- Contagion channel:
  - The “common bank lender channel” could transmit shocks across regions by forcing common global lenders to reduce exposure elsewhere to restore capital ratios.
- Role of China’s financial footprint:
  - Chinese banks’ foreign claims have been growing substantially over the last eight years (though modest relative to domestic claims).
  - Development and global expansion of largely state-owned Chinese banks may create opportunities for deeper financial integration.
  - Accumulated official loans from China Policy Banks (2005–17), as percent of GDP by country:
    - Venezuela: around 30 percent of GDP
    - Ecuador: 17 percent of GDP
    - Jamaica: 15 percent of GDP
    - Trinidad and Tobago: 12 percent of GDP
    - Argentina: 3 percent of GDP
    - Brazil: 2 percent of GDP
  - Such lending is large in U.S. dollars for some countries even if the percent of GDP is smaller.

### Looking forward: trends and policy implications
- Linkages between Asia and LAC have grown and are expected to further intensify; China will likely remain a major driving force, though other Asian economies also matter.
- China’s rebalancing may alter export composition for LAC: demand for mineral intermediates may decline while consumption-driven demand for agricultural products may persist.
- Investment linkages have grown but remain weak; a more sophisticated and diversified Chinese financial system could play a larger role as an investor in LAC alongside Japan and South Korea.
- Policy implications:
  - LAC should diversify exports and increase value added.
  - Facilitate trade integration and reduce trade costs.
  - Address institutional and regulatory barriers and improve the business environment to attract more Asian FDI and portfolio flows.

### Key statistics and empirical magnitudes
- Portfolio investment allocated by LAC to Asian assets: 2 percent.
- Assets issued by LAC in Asia’s portfolios: 1 percent.
- Asia’s share of total FDI stock in LAC: 4 percent.
- LAC’s contribution to FDI stock in Asia: 1 percent.
- Machines and electronics share of imports from Asia to LAC: 40 percent.
- Japan/South Korea share of LAC’s holdings of Asian securities: 86 percent (combined).
- Japan/South Korea share of LAC securities in Asia’s portfolios: 89 percent.
- Asia’s FDI to LAC: origin concentration—60 percent from Japan, 20 percent from South Korea; destination concentration—46 percent in Brazil, 32 percent in Mexico.
- Singapore share of LAC investments in Asia: almost 60 percent.
- Chinese policy-bank accumulated lending (2005–17), percent of GDP: Venezuela 30 percent; Ecuador 17 percent; Jamaica 15 percent; Trinidad and Tobago 12 percent; Argentina 3 percent; Brazil 2 percent.
- Investment in LA6 countries is consistently below other emerging markets by more than 3 percent of GDP (difference reported as 3.22 percentage points).
- Savings as a determinant of investment: estimated savings coefficient oscillates between 0.2 and 0.5 across model specifications.
- LA6 counterfactual example: if LA6 countries had the same average saving rates as other EMs, LA6 investment/GDP would rise by 1.4 percentage points (calculated from a coefficient of 0.4 times the savings gap: 24–20.5).
- Regression excerpt (selected estimates reported):
  - Lagged investment: 0.166** (pval 0.0410)
  - Initial GDP per head: -0.000160*** (pval 2.00e-09)
  - GDP cycle (filter): 0.797*** (pval 0.000280)
  - Rule of Law (World Bank): 2.482*** (pval 5.12e-10)
  - Savings: 0.434*** (pval 0)
  - Observations cited: 557, 560, 271 (various specifications)

---

### *This Regional Economic Outlook background paper was prepared by Fabio Di Vittorio, Ana Lariau, Pedro Rodriguez, and Carlos Goncalves.*

### 2. INVESTMENT AND SAVINGS IN LATIN AMERICA

### Savings and public sector role: key findings
- All LA6 countries save less than their age structure would suggest. LA6 = Brazil, Chile, Colombia, Mexico, Peru, Uruguay.
- Around half of the difference in total savings (LA6 vs. comparator) is due to low public savings.
- In the EMs subgroup, public savings are at the same level they were in 1995, whereas in LA6 the average is still 2 percentage points below.
- Most variation in net public savings over time is driven by oscillations in public savings, while public investment is relatively stable (and slightly procyclical).

### Empirical evidence on public savings and total savings
- Table 3 estimates (Public Savings on Total Savings):
  - Public Savings coefficient: 0.724*** (0)
  - GDP cycle coefficient: 0.325* (0.0868)
  - GDP per capita coefficient: 0.000111*** (1.48e-05)
  - POP_1564_pc coefficient: 0.310*** (8.97e-06)
  - Observations: 544
  - R-squared: 0.441
- Interpretation:
  - Higher public savings lead to higher total savings, although the effect is less than one-to-one.
  - Offsetting behavior by private agents appears weak in practice.

### Historical behavior during commodity price boom and policy implication
- During the 10-year terms-of-trade boom starting in 2003, governments generally did not save enough out of the windfall; only Chile and Peru increased net public savings in the 2003–06 period.
- The end of the commodity super-cycle around 2013–14 caused a sharp fall in real income; except for Mexico, the larger share of the adjustment was shouldered by the private sector.
- Policy recommendation: Authorities should aim at increasing public savings to raise total savings and pave the way for higher investment and long-run economic growth.

---

### 3. LABOR MARKET DYNAMICS IN LATIN AMERICA

### Aggregate labor market trends and social gains
- Average unemployment across the region fell from 10.4 percent in 2000 to a low of 6.1 percent in 2014.
- Labor informality declined by close to 20 percentage points over the same broad period.
- Higher labor incomes, especially for low income households, contributed to lower poverty and inequality; gains were pronounced for commodity exporters during the commodity price boom.

### Post-boom resilience and country-specific outcomes
- Unemployment increased again to 8.8 percent in 2017 (driven to a large extent by Brazil).
- In Chile, Colombia, and Peru, slowdown in labor demand growth was broadly offset by slowdown in labor supply growth (with LFP essentially constant), keeping unemployment roughly constant.
- In Argentina and Brazil, output nearly stagnated between 2012 and 2017, contributing to unemployment increases.
- Brazil lost 4 million formal jobs between 2015 and 2017 and the unemployment rate nearly doubled to 13.7 percent.
- Evidence of labor hoarding (a fall in labor productivity) was present, especially in Brazil.

### Decomposition, Okun’s law, and responsiveness
- Unemployment changes decomposed into contributions from working-age population, labor force participation, output, and labor productivity.
- Okun’s law specification: ut − ut−1 = c + p(yt − yt−1) + εt, where p is the Okun coefficient.
- Estimations show a more muted response of unemployment (and employment) to GDP growth in many Latin American countries than in advanced economies or other emerging markets.
- Employment responds less to output growth in most Latin American countries than in advanced economies or other emerging markets.

### Labor market rigidities and margins of adjustment
- Latin American labor markets remain among the most rigid in the world despite recent reforms; hiring and firing practices are particularly restrictive.
- Rigidity effects:
  - Restrictive hiring/firing reduces firms’ willingness to increase employment when growth picks up and limits ability to lay off workers in downturns.
  - Firms may prefer to reduce productivity rather than adjust employment, contributing to weaker Okun responses.
- Preliminary local projection estimates indicate a larger decline in unemployment during expansions when hiring and firing practices are more flexible.
- Informal labor acts as a key margin of adjustment in the presence of high formal labor market rigidities:
  - Economies with higher labor formality tend to have larger Okun coefficients.
  - In the LA6, higher formality rates are associated with larger declines in unemployment and larger increases in employment during booms.

### Policy implication for labor markets
- Further reforms to enhance labor market flexibility could allow for more dynamic and responsive labor markets, while contributing to reducing informality and increasing productivity.

---

### 4. The Role of Foreign Exchange Intervention in Latin America’s Inflation-Targeting Countries

### Overview and motivation
- Many central banks in the region with inflation targeting (IT) responded to external shocks with foreign exchange intervention (FXI).
- Stated motives for FXI: contain excessive FX volatility and build up international reserve buffers for self-insurance.
- FXI can contain disruptive effects but should not substitute for necessary policy adjustments.
- IT is not incompatible with FXI when domestic and foreign assets are not perfect substitutes; interest rates and FXI can tackle inflation expectations and external balance respectively.

### Communication, rules-based approaches, and credibility
- Credibility of IT can be preserved with strong communication that explains intentions on both monetary policy and FXI.
- Clear FXI communication and data transparency help markets understand objectives and maintain priority of IT.
- Most countries in the region have followed a rules-based approach to FXI.
- Example — Chile:
  - Intervened only four times since adopting IT (2001, 2002–03, 2008, and 2011).
  - 2011 announcement: program of (auction-based) FX purchases of $12 billion through daily purchases of $50 billion over 240 days.
  - In Chile most impact of intervention observed right after announcements—typically on the same day.

### Rules-based interventions, instruments, and transparency
- Mexico used U.S. dollar put options between 1996 and 2001 to increase international reserves by over $12 billion.
- Colombia used similar mechanisms during November 1999 to September 2002, March to August 2003, and March to May 2008, accumulating over $2.8 billion.
- Transparent communication and provision of intervention data helped avoid the notion of an exchange rate target level.

### Use of derivatives, market deepening, and measured effectiveness
- Colombia used “volatility options”:
  - Put (call) auctions with maturity of one month exercised when strike exchange rate was lower (higher) than its 20-day moving average minus (plus) 5 percent.
  - Volume during 1999–2001: $190 million (about 50 percent of the daily market volume); later reduced to $180 million and threshold to 4 percent.
  - Estimated effect: every purchase of $1 million → domestic currency depreciated by 0.01 percent for three weeks.
- Brazil:
  - Announced in August 2013 daily sales of $500 million in swaps contracts until year-end; program extended to mid-2014.
  - By early 2015 accumulated swap exposure: $108 billion (about 1/3 of the stock of international reserves).
  - Estimate: around ¼ percent change in the exchange rate per every $1 billion purchase or sale.
- Mexico non-deliverable forward program (since 2017):
  - Original program auctions announced for up to $20 billion.
  - Initial allotment: $1 billion in March 2017.
  - Total after additional auctions: $4.5 billion.
  - December 2017 additional placement: $500 million dollars.

### FXI and financial stability in dollarized economies
- In highly dollarized economies (Costa Rica, Peru, Uruguay) large dollarized liabilities create currency mismatches that can generate financial stability risks—justifying FXI.
- Peru’s lean-against-the-wind strategy (two-sided discretionary spot interventions) market shares in stress periods:
  - Purchasing up to 25 percent of the spot market turnover in 2006.
  - Selling 13 percent of market turnover between June 2008 and February 2009.
  - Selling 11 percent of turnover in 2016.
- As a result, Peru’s exchange rate volatility has been much smaller than Chile’s.

### Effectiveness conclusions and policy recommendations
- FXI can be effectively integrated into an IT policy framework.
- FXI tends to be more effective when:
  - Interventions are transparent and preannounced with strong communication and data transparency.
  - Derivatives intervention appears as effective or, at worst, slightly less effective than spot intervention.
- Policy recommendations:
  - Further enhance transparency and communication of FXI policy so the market internalizes the central bank FXI reaction function.
  - Rely on announcements to reduce the need for actual intervention, increase intervention effectiveness, and reduce its cost.
  - Strengthening FXI communication and transparency helps enhance central bank credibility and IT effectiveness.

---

### 5. PRODUCTIVITY IN LATIN AMERICA

### Misallocation measures and international comparisons
- Hsieh and Klenow measure suggests potential manufacturing-sector productivity increases if misallocation wedges were eliminated:
  - Brazil: 123 percent
  - Colombia: 98 percent
  - Mexico: 130 percent
- Note: counterfactual is stringent; misallocation levels in developed economies hover between 40 percent and 60 percent.

### Misallocation across sectors
- Degree of misallocation varies considerably across sectors within countries.
- Misallocation tends to be more severe in nontradable “services” compared to manufacturing and agriculture, suggesting lack of competition may be important.

### Empirical results (regression highlights)
- Time to pay taxes:
  - Full sample: 0.000288* (0.000146)
  - Excludes Outliers: 0.000151*** (4.97e-05)
- Flexibility of wage determination:
  - Full sample: -0.0410 (0.0376)
  - Excludes Outliers: -0.0543 (0.0374)
- Population:
  - Full sample: 2.52e-10** (1.12e-10)
  - Excludes Outliers: 3.09e-10*** (8.49e-11)
- Constant:
  - Full sample: 0.954*** (0.179)
  - Excludes Outliers: 1.014*** (0.177)
- Observations: 63 (Full sample), 60 (Excludes Outliers)
- R-squared: 0.120 (Full sample), 0.107 (Excludes Outliers)

### Policy discussion and recommendations
- Misallocation varies across and within countries, indicating both macroeconomic and microeconomic distortions explain poor productivity and non-convergence.
- Policy-induced wedges that may allow small and unproductive firms to survive include:
  - (a) targeted subsidies
  - (b) national content laws and lack of import competition
  - (c) poorly functioning credit markets
  - (d) tax regimes favoring small firms
  - (e) government credit lines at below market interest rates to “selected” firms/sectors

--- 

*This Regional Economic Outlook background paper was prepared by Fabio Di Vittorio, Ana Lariau, Pedro Rodriguez, and Carlos Goncalves.*

### 1. Asia and Latin America: How Deep are their Linkages? 1

### 1. Asia and Latin America: How Deep are their Linkages?

### Overview
- Asia and Latin America and the Caribbean (LAC) have become increasingly connected over the last 20 years, with the process being asymmetric: trade linkages have deepened considerably, while interregional investment and financial connections remain limited.
- The analysis focuses on three types of linkages between Asia and LAC: trade, investment, and financial.

### Trade: key findings and implications
- LAC trade with Asia has increased over the last 20 years, at the expense of trade with North America.
- Asia is the second largest trading partner for LAC, both in terms of imports and exports; within Asia, China is the main destination.
- Major country roles:
  - Brazil, Chile, Mexico, and Argentina are the main LAC exporters to Asia.
  - Mexico is the main importer from Asia (partly explained by its role in global value chains).
- Composition of trade:
  - LAC exports to Asia are highly concentrated in a small number of commodities, particularly minerals (copper and iron ores) and vegetables (soybeans and soybean oil).
  - LAC imports from Asia are concentrated in machines and electronics (40 percent of total import from Asia to LAC).
- Structural drivers:
  - Complementarity between China’s historical demand for raw materials (to support investment-led growth) and LAC’s natural-resource endowment explains much of the LAC→Asia trade.
  - China’s exports of final goods to LAC reflect China’s abundant low-cost labor and manufacturing capabilities.
- Risks and opportunities from China’s rebalancing:
  - China’s shift from investment- to consumption-led growth may reduce demand for mineral intermediate inputs while maintaining demand for soybeans and meat products.
  - Chile and other Andean exporters are particularly exposed to a potential deceleration in Chinese demand for minerals.
  - Policy recommendations to manage this transition: diversify exports, increase value added, and implement policies supporting trade integration and reduced trade costs to facilitate market reorientation.

### Investment: magnitude, composition, and impediments
- Portfolio and FDI magnitudes:
  - In LAC, portfolio investment is heavily biased towards North America; only 2 percent is allocated to assets issued by Asian countries.
  - Assets issued by LAC countries represent only 1 percent of Asia’s portfolio.
  - Asia holds 4 percent of the total FDI stock in LAC; LAC contributes 1 percent to the stock of FDI in Asian countries.
  - These portfolio and FDI compositions have remained stable over time.
- Geographic and country concentration:
  - Japan and South Korea are the key players for portfolio investment and FDI between Asia and LAC:
    - Japan and South Korea together represent 86 percent of LAC’s holdings of Asian securities, with Chile and Brazil being the main investors.
    - Japan and South Korea account for 89 percent of LAC securities held in Asia’s investment portfolio.
    - Asia’s FDI in LAC is geographically concentrated: 60 percent originates from Japan and 20 percent from South Korea; destination concentration: 46 percent in Brazil and 32 percent in Mexico.
    - In LAC investments in Asia, Singapore is the main destination, accounting for almost 60 percent of the total investment position.
- Nature of FDI:
  - Most FDI between Asia and LAC corresponds to greenfield investments; interregional merger and acquisition activity remains very limited.
  - Sectoral patterns:
    - Asian greenfield investments in South America (Andean region) focus on mining and petroleum.
    - Investments to Central America leverage low-cost labor and proximity to the U.S. in apparel and offshore business services.
    - Market-seeking investments (automotive, business electronics) target Brazil’s large domestic market; Mexico’s investments reflect its status as an export platform for U.S.-destined products.
    - LAC investments in Asia primarily target fast-growing consumer markets in financial, IT, and business services; manufacturing investments are very rare.
- Factors restricting interregional investment:
  - Limited promotional efforts between regions.
  - Small markets and undiversified exports in many LAC countries.
  - Unfavorable business environments and institutional factors; some Asian and LAC economies have among the most restrictive FDI regimes in the world (per OECD FDI Regulatory Restrictiveness Index).
  - Regulatory restrictions and FTAs without investment chapters that limit investor incentives.
  - Informational advantages of Asian investors toward Asian emerging markets may reduce portfolio flows to LAC.

### Other financial linkages and contagion channels
- Direct banking linkages:
  - Direct financial linkages via cross-border banking are very limited and mostly involve lending by Japanese, South Korean, and Taiwanese banks to LAC.
  - Japan accounts for about 80 percent of the entire bank exposure of Asia to LAC, concentrated in Brazil and Mexico and offshore financial centers such as Panama and Bermuda.
- Indirect linkages:
  - Indirect financial linkages (proxied by the fund competition index) are stronger than direct ones but still smaller than intra-regional linkages.
  - Significant indirect linkages exist between Asia and some LAC emerging markets (Argentina, Colombia, Panama) through common exposure to the U.S. banking system.
  - The heatmap analysis indicates intraregional (Asia-Asia and LAC-LAC) competition for global funds is high, while interregional (LAC-Asia) competition is low, with exceptions noted above.
- Contagion channel:
  - The “common bank lender channel” could transmit shocks across regions: losses in one country/region could force a common global lender to reduce exposure elsewhere to restore capital ratios—this channel was strong during the Mexican and Asian crises in the 1990s.
- Role of China’s financial footprint:
  - Chinese banks’ foreign claims have been growing substantially over the last eight years (though modest relative to domestic claims).
  - Development and global expansion of largely state-owned Chinese banks may create opportunities for deeper financial integration between the regions.
  - Accumulated official loans from China Policy Banks (2005–17), as percent of GDP by country:
    - Venezuela: around 30 percent of GDP
    - Ecuador: 17 percent of GDP
    - Jamaica: 15 percent of GDP
    - Trinidad and Tobago: 12 percent of GDP
    - Argentina: 3 percent of GDP
    - Brazil: 2 percent of GDP
  - Such lending is large in U.S. dollars for some countries even if the percent of GDP is smaller.

### Looking forward: trends and policy implications
- Linkages between Asia and LAC have grown and are expected to further intensify; China will likely remain a major driving force, though other Asian economies also matter.
- China’s rebalancing may alter export composition for LAC: demand for mineral intermediates may decline while consumption-driven demand for agricultural products may persist.
- Investment linkages have grown but remain weak; a more sophisticated and diversified Chinese financial system could play a larger role as an investor in LAC alongside Japan and South Korea.
- Policy implications emphasized in the analysis:
  - LAC countries should diversify exports and increase value added to rebalance relations with China.
  - Policies to facilitate trade integration and reduce trade costs are important to allow LAC firms to seize new market opportunities.
  - Addressing institutional and regulatory barriers and improving the business environment could help attract more Asian FDI and portfolio flows.

### Key statistics and empirical magnitudes (as reported)
- Portfolio investment allocated by LAC to Asian assets: 2 percent.
- Assets issued by LAC in Asia’s portfolios: 1 percent.
- Asia’s share of total FDI stock in LAC: 4 percent.
- LAC’s contribution to FDI stock in Asia: 1 percent.
- Machines and electronics share of imports from Asia to LAC: 40 percent.
- Japan/South Korea share of LAC’s holdings of Asian securities: 86 percent (combined).
- Japan/South Korea share of LAC securities in Asia’s portfolios: 89 percent.
- Asia’s FDI to LAC: origin concentration—60 percent from Japan, 20 percent from South Korea; destination concentration—46 percent in Brazil, 32 percent in Mexico.
- Singapore share of LAC investments in Asia: almost 60 percent.
- Chinese policy-bank accumulated lending (2005–17), percent of GDP: Venezuela 30 percent; Ecuador 17 percent; Jamaica 15 percent; Trinidad and Tobago 12 percent; Argentina 3 percent; Brazil 2 percent.
- Investment in LA6 countries is consistently below other emerging markets by more than 3 percent of GDP (difference reported as 3.22 percentage points).
- Savings as a determinant of investment: estimated savings coefficient oscillates between 0.2 and 0.5 across model specifications.
- Example from LA6 counterfactual: if LA6 countries had the same average saving rates as other EMs, LA6 investment/GDP would rise by 1.4 percentage points (calculated from a coefficient of 0.4 times the savings gap: 24–20.5).
- Regression excerpt (selected estimates reported):
  - Lagged investment: 0.166** (pval 0.0410)
  - Initial GDP per head: -0.000160*** (pval 2.00e-09)
  - GDP cycle (filter): 0.797*** (pval 0.000280)
  - Rule of Law (World Bank): 2.482*** (pval 5.12e-10)
  - Savings: 0.434*** (pval 0)
  - Observations cited: 557, 560, 271 (various specifications)

*This Regional Economic Outlook background paper was prepared by Fabio Di Vittorio, Ana Lariau, Pedro Rodriguez, and Carlos Goncalves.*

### 2.  INVESTMENT AND SAVINGS IN LATIN AMERICA

### 2.  INVESTMENT AND SAVINGS IN LATIN AMERICA

### Savings and public sector role: key findings
- All LA6 countries save less than their age structure would suggest (Figure 3). LA6 = Brazil, Chile, Colombia, Mexico, Peru, Uruguay.
- Around half of the difference in total savings (LA6 vs. comparator) is due to low public savings (Figure 4).
- In the emerging markets (EMs) subgroup, public savings are at the same level they were in 1995, whereas in LA6 the average is still 2 percentage points below.
- Most of the variation in net public savings over time is driven by oscillations in public savings, while public investment is relatively stable (and slightly procyclical) (Figure 5).

### Empirical evidence on public savings and total savings
- Table 3 (Public Savings on Total Savings) estimates:
  - Public Savings coefficient: 0.724*** (0)
  - GDP cycle coefficient: 0.325* (0.0868)
  - GDP per capita coefficient: 0.000111*** (1.48e-05)
  - POP_1564_pc coefficient: 0.310*** (8.97e-06)
  - Observations: 544
  - R-squared: 0.441
- Interpretation: Higher public savings lead to higher total savings, although the effect is less than one-to-one. This finding is robust after controlling for the country’s GDP cycles, global cycles, GDP per capita, and age structure.
- Offsetting behavior by private agents (anticipating lower future taxes and reducing private savings) appears weak in practice.

### Historical behavior during commodity price boom and policy implication
- During the 10-year terms-of-trade boom starting in 2003, governments in the region generally did not save enough out of the windfall; only Chile and Peru increased net public savings in the 2003–06 period (Figure 6).
- The end of the commodity super-cycle around 2013–14 caused a sharp fall in real income; with the exception of Mexico, the larger share of the adjustment was shouldered by the private sector.
- Policy recommendation: Authorities should aim at increasing public savings to raise total savings and pave the way for higher investment and long-run economic growth.

---

### 3.  LABOR MARKET DYNAMICS IN LATIN AMERICA

### Aggregate labor market trends and social gains
- Average unemployment across the region fell from 10.4 percent in 2000 to a low of 6.1 percent in 2014.
- Labor informality declined by close to 20 percentage points over the same broad period.
- Higher labor incomes, especially for low income households, contributed to lower poverty and inequality; these gains were pronounced for commodity exporters during the commodity price boom.

### Post-boom resilience and country-specific outcomes
- Unemployment increased again to 8.8 percent in 2017 (driven to a large extent by Brazil).
- In Chile, Colombia, and Peru, the slowdown in labor demand growth was broadly offset by a slowdown in labor supply growth (with LFP essentially constant), keeping unemployment roughly constant.
- In Argentina and Brazil, output nearly stagnated between 2012 and 2017, removing positive demand impacts on unemployment and contributing to unemployment increases.
- Brazil lost 4 million formal jobs between 2015 and 2017 and the unemployment rate nearly doubled to 13.7 percent.
- Evidence of labor hoarding (a fall in labor productivity) was present, especially in Brazil, which limited increases in unemployment during recessions.

### Decomposition, Okun’s law, and responsiveness
- Decomposition of unemployment changes identifies contributions from working-age population, labor force participation, output, and labor productivity.
- Okun’s law specification used: ut − ut−1 = c + p(yt − yt−1) + εt, where p is the Okun coefficient.
- Estimations show a more muted response of unemployment (and employment) to GDP growth in many Latin American countries than in advanced economies or other emerging market economies (Figure 3).
- Employment responds less to output growth in most Latin American countries than in advanced economies or other emerging markets.

### Labor market rigidities and margins of adjustment
- Latin American labor markets remain among the most rigid in the world despite recent reforms; hiring and firing practices are particularly restrictive (Table 1 summary metrics cited).
- Rigidity effects:
  - Restrictive hiring/firing reduces firms’ willingness to increase employment when growth picks up and limits ability to lay off workers in downturns.
  - Firms may prefer to reduce productivity rather than adjust employment, contributing to weaker Okun responses.
- Preliminary local projection estimates indicate a larger decline in unemployment during expansions when hiring and firing practices are more flexible (Figure 4).
- Informal labor acts as a key margin of adjustment in the presence of high formal labor market rigidities:
  - Economies with higher labor formality tend to have larger Okun coefficients.
  - In the LA6, higher formality rates are associated with larger declines in unemployment and larger increases in employment during booms (Figure 5).

### Policy implication for labor markets
- Further reforms to enhance labor market flexibility could allow for more dynamic and responsive labor markets, while contributing to reducing informality and increasing productivity.

*IMF Regional Economic Outlook background paper: Western Hemisphere (chapter sections on “Investment and Savings in Latin America” and “Labor Market Dynamics in Latin America”).*

### 4.  The Role of Foreign Exchange Intervention in

### 4.  The Role of Foreign Exchange Intervention in Latin America’s Inflation-Targeting Countries

### Overview and motivation
- Many central banks in the region with inflation targeting (IT) responded to external shocks with foreign exchange intervention (FXI).
- Stated motives for FXI: contain excessive foreign exchange (FX) volatility and build up international reserve buffers for self-insurance.
- Rationale: disorderly market conditions could lead to exchange rate movements that amplify shocks, unanchor inflation expectations, produce one-sided and illiquid markets, and encourage herd-like behavior—ultimately undermining macroeconomic and financial stability.
- FXI can contain disruptive effects but should not substitute for necessary policy adjustments.
- IT is not incompatible with FXI when domestic and foreign assets are not perfect substitutes; two instruments (interest rates and FXI) can tackle two objectives (anchoring inflation expectations and ensuring external balance) (Ghosh and others (2016)).

### Communication, rules-based approaches, and credibility
- The credibility of an IT framework can be preserved with a strong communication strategy that explains authorities’ intentions on both monetary policy and FXI.
- Clear FXI communication policy combined with data transparency helps markets understand objectives sought with FXI while maintaining the priority of IT, enabling market participants to internalize the central bank’s FXI reaction function.
- Most countries in the region have followed a rules-based approach to FXI, facilitating market understanding of intervention and its subordination to interest rate policy.
- Example — Chile:
  - Intervened only four times since adopting IT (2001, 2002–03, 2008, and 2011).
  - In all FXI cases the central bank’s communique stressed explicitly that interventions were subordinated to the macroeconomic framework.
  - In the 2011 event the central bank announced a program of (auction-based) FX purchases of $12 billion through daily purchases of $50 billion over 240 days; the central bank further stated that subordination was to a monetary policy with exchange rate flexibility.
  - In Chile most of the impact of the intervention was observed right after announcements—typically on the same day.

### Rules-based interventions, instruments, and transparency
- Mexico used U.S. dollar put options between 1996 and 2001 to increase international reserves by over $12 billion; the FX Commission gave financial institutions the right to sell U.S. dollars to the central bank the following month if the option’s strike price the day before was below its 20-day moving average.
- Colombia used the same mechanism during November 1999 to September 2002, March to August 2003, and March to May 2008, accumulating over $2.8 billion.
- Transparent communication of intervention rules and provision of intervention data helped avoid the notion of an exchange rate target level and allowed market participants to factor in the central bank’s reaction function to protracted exchange rate movements.

### Use of derivatives, market deepening, and measured effectiveness
- Colombia used “volatility options” to mitigate exchange rate volatility without impacting trend:
  - Put (call) auctions with a maturity of one month exercised when the strike exchange rate was lower (higher) than its 20-day moving average minus (plus) 5 percent.
  - During 1999–2001 the volume of these options was $190 million (about 50 percent of the daily market volume); later the volume was reduced to $180 million and the threshold to 4 percent.
  - Effectiveness estimate: for every purchase of $1 million the domestic currency depreciated by 0.01 percent for three weeks.
- Brazil announced in August 2013 daily sales of $500 million in swaps contracts until year-end; program extended to mid-2014. By early 2015 the central bank had a $108 billion accumulated swap exposure (about 1/3 of the stock of international reserves).
  - Some estimates: around ¼ percent change in the exchange rate per every $1 billion purchase or sale.
  - Most of the impact took place on the date of the announcement; implementation lags imply estimates are typically not statistically significant.
- Mexico provided non-deliverable forward contracts since 2017 to intervene without sacrificing reserves and to satisfy investors and corporate hedging needs:
  - Original program auctions were announced for up to $20 billion.
  - After an initial allotment of $1 billion in March of that year, additional auctions brought the total to $4.5 billion.
  - In December 2017, an additional $500 million dollars were placed.

### FXI and financial stability in dollarized economies
- In highly dollarized economies (Costa Rica, Peru, Uruguay) large dollarized liabilities create currency mismatches that can generate financial stability risks from excessive exchange rate volatility—justifying FXI.
- Peru followed a lean-against-the-wind strategy through mostly two-sided discretionary spot interventions:
  - In periods of stress, volumes were as high as purchasing 25 percent of the spot market turnover in 2006, selling 13 percent of market turnover between June 2008 and February 2009, and 11 percent of the turnover in 2016.
  - As a result, Peru’s exchange rate volatility has been much smaller than Chile’s (another mineral exporter subject to similar external shocks).

### Effectiveness conclusions and policy recommendations
- Latin America’s experience suggests FXI can be effectively integrated into an IT policy framework.
- FXI tends to be more effective when:
  - Interventions are transparent and preannounced on the back of a strong FXI communication policy and data transparency.
  - Derivatives intervention appears as effective or, at worst, slightly less effective than spot intervention.
- Policy recommendations emphasized:
  - Further enhance transparency and communication of FXI policy so the market better internalizes the central bank FXI reaction function.
  - Rely on announcements to reduce the need for actual intervention, increase intervention effectiveness, and reduce its cost.
  - Strengthening FXI communication and transparency helps enhance central bank credibility and its inflation-targeting effectiveness.

### Key quantitative figures and examples (preserve exact values)
- Chile: FX interventions in 2001, 2002–03, 2008, and 2011.
- Chile 2011 announcement: program of (auction-based) FX purchases of $12 billion through daily purchases of $50 billion over 240 days.
- Mexico: increased international reserves by over $12 billion via U.S. dollar put options (1996–2001).
- Colombia: accumulated over $2.8 billion using put options (1999–2002, 2003, 2008).
- Colombia volatility options:
  - Volume during 1999–2001: $190 million (about 50 percent of the daily market volume); later reduced to $180 million and threshold to 4 percent.
  - Estimated effect: every purchase of $1 million → domestic currency depreciated by 0.01 percent for three weeks.
- Brazil:
  - Announced daily sales of $500 million in swaps in August 2013.
  - By early 2015 accumulated swap exposure: $108 billion (about 1/3 of the stock of international reserves).
  - Estimate: around ¼ percent change in the exchange rate per every $1 billion purchase or sale.
- Mexico non-deliverable forwards (since 2017):
  - Original auctions up to $20 billion.
  - Initial allotment: $1 billion in March 2017.
  - Total after additional auctions: $4.5 billion.
  - December 2017 additional placement: $500 million dollars.
- Peru intervention market shares in stress periods:
  - Purchasing up to 25 percent of the spot market turnover in 2006.
  - Selling 13 percent of market turnover between June 2008 and February 2009.
  - Selling 11 percent of turnover in 2016.

*Regional Economic Outlook background paper: Western Hemisphere — "The Role of Foreign Exchange Intervention in Latin America’s Inflation-Targeting Countries."*

### 5.  PRODUCTIVITY IN LATIN AMERICA

### 5.  PRODUCTIVITY IN LATIN AMERICA

### Misallocation measures and international comparisons
- A measure of misallocation proposed by Hsieh and Klenow suggests that misallocation in the manufacturing sector would generate productivity increases of:
  - Brazil: 123 percent
  - Colombia: 98 percent
  - Mexico: 130 percent
- These percentages were obtained by computing how much productivity would increase in the manufacturing sector if the wedges that generate misallocation were eliminated.
- The methodology constructs a counterfactual exercise in which all distortions are driven to zero; the text notes this is a stringent counterfactual and that misallocation levels in developed economies hover between 40 percent and 60 percent, providing a more realistic benchmark.

### Misallocation across sectors
- The degree of misallocation varies considerably across sectors within a given country.
- Misallocation tends to be more severe in nontradable sectors—“services”—as compared to manufacturing and agriculture, suggesting that lack of competition may be an important factor behind poor allocation of productive resources.
- Table 5 (Orbis database; IMF staff calculations, following the Hsieh and Klenow (2009) methodology) documents large heterogeneity in percent of misallocation across NACE 4-digit sectors, with several services sectors showing particularly high percentages.

### Empirical results (regression highlights from table)
- Regression results reported (robust standard errors in parentheses; significance: *** p<0.01, ** p<0.05, * p<0.1):
  - Time to pay taxes:
    - Full sample: 0.000288* (0.000146)
    - Excludes Outliers: 0.000151*** (4.97e-05)
  - Flexibility of wage determination:
    - Full sample: -0.0410 (0.0376)
    - Excludes Outliers: -0.0543 (0.0374)
  - Population:
    - Full sample: 2.52e-10** (1.12e-10)
    - Excludes Outliers: 3.09e-10*** (8.49e-11)
  - Constant:
    - Full sample: 0.954*** (0.179)
    - Excludes Outliers: 1.014*** (0.177)
  - Observations: 63 (Full sample), 60 (Excludes Outliers)
  - R-squared: 0.120 (Full sample), 0.107 (Excludes Outliers)

### Policy discussion and recommendations
- Misallocation of resources varies substantially across countries and within countries, indicating that both macroeconomic and microeconomic distortions explain lackluster productivity performance and non-convergence in LAC countries.
- Many economic policies common in Latin America are in principle conducive to poor allocation of productive resources. Policy-induced wedges that may allow small and unproductive firms to survive include:
  - (a) targeted subsidies
  - (b) national content laws and lack of import competition
  - (c) poorly functioning credit markets
  - (d) tax regimes favoring small firms
  - (e) government credit lines at below market interest rates to “selected” firms/sectors

### References
- Chang-Tai Hsieh and Peter Klenow, “Misallocation and Manufacturing TFP in China and India,” Quarterly Journal of Economics, Vol 74, 2009.

*Regional Economic Outlook background paper: Western Hemisphere — 5. PRODUCTIVITY IN LATIN AMERICA*

---


_Source: https://www.imf.org/-/media/files/publications/reo/whd/2018/october/wreo1018-background-papers.pdf_