## 1. Population Structure 2014 ## Source details **Canonical URL:** [1. Population Structure 2014](https://www.imf.org/-/media/files/publications/cr/2017/cr17191.pdf) ## Other formats - [Markdown version](/-/media/files/publications/cr/2017/cr17191.pdf.md) - [Structured JSON version](/-/media/files/publications/cr/2017/cr17191.pdf.json) --- ### Climate change risks - Vietnam is ranked among the five countries likely to be most affected by climate change due to long coastline, geographic location, and diverse topography and climates. - Over the last 50 years: - temperatures have increased twice as fast as the global average; - the sea level has risen by 20 centimeters; - the frequency and intensity of extreme weather events (drought, flood, salinization) have risen sharply. - Natural disasters result in 470 fatalities and cost 0.8 percent of GDP (annual average between 1990 and 2016). - Authorities’ climate change scenarios project by the end of the century: - sea levels are expected to rise by up to a meter; - sea waters would cover 40 percent of the Mekong Delta area, 3 percent of coastal provinces and 20 percent of Ho Chi Minh City; - direct impacts could affect 10–12 percent of the population and reduce GDP by 10 percent. - Sectoral vulnerabilities: - agriculture, aquaculture, energy, transportation and tourism are the most affected sectors. - a 1 degree Celsius rise in summer temperatures is expected to raise energy demand by 1 percent. - a 1 meter sea-level rise would submerge 10,000 kilometers of roads. - Social impacts: - lower agricultural incomes, erosion of coastal land and deterioration of arable land could reverse poverty reduction and accelerate rural-to-urban migration. - Ho Chi Minh City, with population approaches close to 10 million, is projected to be among the cities in the region most affected by sea-level rise and increased storm surges. - Agronomic projections and observed impacts: - In 2016, rice production fell by 1.1 million tons due to salinization in the Mekong Delta. - Rice, grown by nearly 80 percent of Vietnamese farmers on 45 percent of the agricultural land, could see production decrease by 9 million tons (13 percent) by 2050 absent adaptation. ### Environmental and resource sustainability - Growth model and natural capital: - Rapid industrialization and resource-intensive development have reduced Vietnam’s stock of natural capital through depletion of mineral and non-mineral resources and increased air, land and water pollution. - Legacy war-related issues (dioxin hotspots, hereditary diseases and birth defects) add urgency to environmental upgrades. - Adjusted Net Saving (ANS): - ANS corrects Net National Savings for resource depletion, environmental degradation, and education spending. - Despite strong education spending, intensive oil production and CO2 and particulate emissions caused ANS to be strongly lower than NNS; in recent years lower oil production reversed that trend, and in 2014, the gap was close to zero. - Pollution sources and incidents: - Vietnam is one of the largest users of fertilizer worldwide; intensive livestock production, shrimp aquaculture, and fisheries practices contribute to land and water pollution and mangrove destruction. - In 2016, a steel plant released large amounts of toxic chemical waste, killing fish along a long stretch of coast and reducing tourism in affected provinces by about 30 percent. - Air pollution and health: - Vietnam is among the top ten countries affected by air pollution; levels of fine particulate matter in large cities and industrial zones are comparable to China’s. - Studies in Ho Chi Minh City show associations between air pollution concentration and hospital admissions for lung and respiratory problems. - GHG emissions are expected to double between 2010 and 2020 and triple by 2030; the entire population is exposed to levels exceeding WHO guidelines. - Energy sector and emissions: - Vietnam is the 20th largest user of coal-fired plants globally. - In 2014, coal produced a quarter of electricity; the Power Development Plan (PDP VII) projects 55GW of coal-fueled power in 2030, up from 14GW today. - If implemented, coal would then amount for 53 percent of the country’s energy mix. - A Harvard report estimates fatalities from air pollution due to coal-fired plants would rise from 4,300 to 25,000 per year. - Vietnam is the most energy intensive economy in Southeast Asia (energy intensity metric presented). ### Vietnam’s initiatives and commitments - Strategic frameworks and targets: - National Climate Change Strategy (NCCS) (issued 2011) objectives to 2050: (a) improve climate change resilience through mitigation and monitoring; (b) strengthen adaptability by lowering risks; (c) adopt a more sustainable, low-carbon growth model. - National Green Growth Strategy (NGGS) (approved 2012) targets: improve energy efficiency, limit sectors generating pollution and natural resource degradation, develop sustainable infrastructure (transportation, energy, irrigation and urban works), and reduce GHG emissions. - Renewable energy and efficiency initiatives: - Renewable Energy Development Project (with World Bank) plans to increase share of renewable energy to 30 percent in 2030 by providing credit support and facilitating private participation. - General Electric investment: US$1.5 billion project to produce up to 1,000 megawatts of wind-generated electricity by 2018. - Other measures: forest plantation increases, energy-efficiency regulations for new office and commercial buildings, and development of urban public transportation networks. - International commitments: - By ratifying the Paris Agreement in 2016, Vietnam committed to reduce GHG emissions by 2030 via Nationally Determined Contributions (NDC): - an 8 percent reduction compared to business as usual financed by domestic resources; - a 25 percent reduction with international support. - Vietnam is dedicated to achieve the Sustainable Development Goals by 2030, many containing environmental and climate targets. - Institutional arrangements: - National Committee on Climate Change (NCCC) established 2012, chaired by the Prime Minister, to lead, coordinate, harmonize and monitor climate change and green growth program implementation. - Four policy implementation programs: (i) National Target Program to Respond to Climate Change (NTP-RCC); (ii) Support Program to Respond to Climate Change (SP-RCC); (iii) National Scientific and Technological Program on Climate Change; (iv) National Target Program on Energy Efficiency and Conservation (NTP-EE). ### Policy considerations and recommendations - Financing and coordination challenges: - Climate change financing is a critical challenge for public financial management; comprehensive costing and identification of financing for adaptation and mitigation projects remains difficult. - Numerous national strategies with sometimes conflicting objectives must be harmonized across national, provincial and sectoral levels. - Market-based and regulatory instruments: - Environmentally friendly tax and subsidy reforms and independent environmental regulation are important policy tools. - Pigouvian taxes and subsidies can reflect social costs and benefits, induce environmentally friendly behavior, and incentivize green investments. - Arms-length regulation is essential. - Energy- and fuel-pricing recommendations: - IMF Fiscal Affairs Department research estimates tax on gasoline in Vietnam should be increased by 0.46 US dollar per liter (0.43 US dollar for diesel) to fully price negative externalities from fuel consumption. - The government’s proposal to double the environmental protection tax on petroleum consumption—from an average of 0.11 to 0.24 US dollar per liter—is noted as a positive development but requires consideration of impacts on vulnerable populations and transparency on revenue use. - Electricity tariffs should be reviewed to account for externalities from energy production, incentivize investment in energy efficiency projects and stimulate private investment in renewable energy. - Private investment and business climate: - Measures to improve the business climate are critical to boost private and foreign investment, especially for large-scale infrastructure projects. ### Demographics (overview) - Vietnam has enjoyed a substantial demographic dividend in recent decades. - Rapid projected aging could weigh on economic growth and necessitate social security and fiscal reforms. - Managed well, the demographic transition can coincide with longer, healthier, more prosperous and more productive lives for Vietnamese people. *International Monetary Fund — Vietnam staff report content excerpt (May 23, 2017).* ### 1. Vietnam is a young country, but ### Demographic transition and projections - Median age: 26. - Largest age cohorts: between 20 and 34. - Declining fertility rates since the early 1970s and rising life expectancy will increase the population’s median age. - Population growth rate is projected to fall close to zero by 2050. - Working-age population share peaked in 2013 and is projected to decline over coming decades. - Share of population age 65 and older is projected to reach close to 3½ times the current level by 2050. - Projections are based on the UN's medium fertility variant (UN Population Division, World Population Prospects, 2015 revision). ### Speed of aging and income context - Number of years for the old-age dependency ratio to increase from 15 percent to 20 percent: - Vietnam: 8 years. - Comparative note: 26 years in Europe; more than 50 years in the United States. - Vietnam is at risk of becoming old before becoming rich; among selected economies, Vietnam’s working-age population reached its peak at the lowest relative per capita income level. ### Demographics as headwinds to growth - Baseline assumptions for estimating growth impact of demographic-induced labor force changes: - Unchanged total factor productivity (TFP) growth. - Unchanged age- and gender-specific labor force participation rates (and employment rates). - Constant capital-to-effective-labor ratio. - Labor force size changes are estimated to add close to ½ percentage point to average annual growth between 2020 and 2050. - The growth effect could fall to zero by mid-century. - On a per capita basis, a declining labor force size is estimated to subtract 0.1 percentage points from annual average growth between 2020 and 2050. ### Productivity and workforce aging - Empirical finding: an increase in the share of older workers (55-65 years) is associated with a significant reduction in labor productivity growth (sample of Asian and European countries). - For Vietnam: - Share of older workers projected to increase by close to 10 percentage points by 2050. - This could reduce growth by 0.2 percentage points per year. - Productivity impact differs by occupation: - Productivity tends to decline with age in physically demanding professions (factory workers, construction, agriculture). - Productivity may increase with age in professions such as lawyers, managers, and doctors. - Vietnam has a high share of its workforce in professions where productivity tends to decline with age. ### Labor force participation and reallocation - Current LFPRs are already high. - Scenario: a gradual increase in LFPRs by 6 percentage points by 2050 could increase growth by 0.2 percentage points per year. - LFPRs for men and women are among the highest in the region, limiting scope for increases. - Policy scope: - Increase female and older worker participation. - Tackle under- and low-productivity employment. - Structural issues: productivity outside the FDI sector remains very low; informality is high; 40 percent of the workforce is employed in agriculture. ### Fiscal challenges and pension system sustainability - Under current policies, age-related public expenditures (pensions and health care) are projected to increase by 8 percentage points of GDP by 2050, more than double the current level. - Most of the increase would be related to pensions. - Spending on education could fall with a declining number of young people, but this is unlikely to fully offset rising age-related expenditures. - On the revenue side, lower growth rates because of demographic change could weigh on fiscal revenues. Pension system features and issues: - Structure: defined benefit pay-as-you-go system covering public sector workers and formal private sector workers. - Retirement ages: 60 for men and 55 for women. - Problems: - Generous general and occupational early retirement rules reduce effective retirement ages and increase unfunded liabilities. - Benefit indexation has regularly exceeded CPI. - High accrual rates yield replacement rates of 75 percent after 30 and 25 years of service for men and women, respectively. - Pension fund reserves are poorly invested, with returns below GDP growth. - System is inequitable across private and public sector employees and across the covered and uncovered population. - Suggested non-pension old-age insurance: land reform to give peasants title to land that can be traded and sold could be an effective old age insurance mechanism for large parts of the rural population. ### Policy implications and recommendations - Short-run and long-run policy priorities: - Raise the economy’s growth potential and sustain high growth in the next decades to offset demographic headwinds. - Foster faster productivity growth while reducing the factor intensity of growth. - Key reform areas: - SOE and banking reforms to achieve a level playing field for the private sector and improve allocation of capital. - Growth-friendly fiscal consolidation to strengthen public finances before rapid aging sets in. - Pension reforms in the near term should: - Raise retirement ages. - Address early retirement rules. - Reform accrual rates and indexation. ### Financial deepening, credit allocation, and growth (summary) - Vietnam’s financial system is bank-centered and dominated by the state. - 2016 statistics: - Banking sector assets amounted to 194 percent of GDP. - Banking sector accounted for more than 96 percent of financial sector assets (insurance companies: 3 percent; securities and fund management companies: 1 percent). - Four major state-owned credit banks (SOCBs) account for 45 percent of banking sector assets and provide half of total credit. - Stock market capitalization increased to 33 percent of GDP, from 27 percent in 2015. - Credit growth: - Averaged 24 percent in the last ten years. - Credit-to-GDP ratio expanded by 4.8 percentage point per annum between 2000 and 2015 and reached 124 percent of GDP at end-2016. - Consequences of rapid credit growth: - Productivity of credit and rates of return to investment have deteriorated. - Real credit grew almost three times faster than real GDP. - Credit targets by bank and by sector create inefficiencies in capital allocation. - Share of SOEs in total credit declined to 15 percent of total in 2016, but some poorly performing SOEs remain highly leveraged. - Growing share of loans going to real estate, financial and personal sectors; real estate investment expanded by 29 percent per year on average in real terms over the last decade while average annual output growth in the sector was 5.2 percent over the same period. - SME financing constraints: - In 2015, access to finance was the main business environment constraint for SMEs. - Only 29 percent of small enterprises (1–20 employees) have an active line of credit versus 57 percent of large firms (+100 employees). - Domestic SMEs compete with SOEs and large domestic enterprises with preferential access to resources; as a result, SME investment is subdued and largely internally financed. - Macro implications: - Credit misallocation and SME financing constraints contribute to a low share of investment in GDP (down by 10 percentage points since 2005). - Non-state sector investment has stagnated around 12 percent of GDP. - Public investment efficiency is hampered by tight fiscal space and lack of coordination in a fragmented general government; local governments account for about 80 percent of total state budget investment and often select projects without a strategic approach linked to national priorities. *International Monetary Fund — Vietnam staff report content excerpt (May 23, 2017).* ### 7. A new financial development index created by the Fund indicates that Vietnam’s ### Financial development index — findings - Vietnam’s financial sector development has improved during the last two decades but remains below emerging markets. - Components of the Financial Development Index reveal: - Financial depth and efficiency are at its peers’ level for both institutions and markets. - Access to financial institutions is low, owing to the relatively small amount of bank branches and ATMs per capita (respectively 4 and 24 per 100,000 adults). ### Access and capitalization implications for growth - Empirical analysis indicates a bell-shaped relationship between financial development and growth. - Financial development increases growth up to a certain point (between 0.45 and 0.7 on the FD index) after which further development has a negative impact. - Research found that the weakening effect on growth at higher levels of financial development can be attributed to financial deepening, rather than to higher access or greater efficiency. - For Vietnam specifically: - Estimates suggest that improving access to financial services to individuals and SMEs could help boost growth by an additional percentage point each year. ### Credit growth and financial stability — historical context and risks - Vietnam’s credit-to-GDP ratio is now close to levels reached during previous periods of macroeconomic instability, signaling potential risks ahead. - Past credit cycles led to deteriorating quality of bank balance sheets and higher inflation. - The most recent ones, in 2008 and 2011, were characterized by large shares of credit directed to SOEs and to the real estate sector. - They were followed by surges of inflation—over 20 percent—and engendered a sharp rise of NPLs and significant weaknesses in the banking sector, which are still weighing on the real economy. ### Credit-to-GDP: regression analysis and gap - Regression analysis indicates that credit in Vietnam is 40 percent of GDP greater than expected. - Method: a random-effects regression linking credit to fundamentals and using international panel data following Cottarelli and others (2004). - Fundamentals used: public debt, GDP per capita, inflation, capital account openness, bank regulation indicators, legal origin and a time trend. - Results imply an appropriate level of credit-to-GDP ratio would be around 80 percent and that the gap has widened in recent years. ### Policy recommendations and conclusions - Vietnam’s economy would reap large benefits from financial sector reforms. Specific reforms mentioned: - Bank and SOE reforms, for instance the phasing out of credit targets. - Lifting interest rate ceilings. - Establishment of a level playing field for access to credit. - Expected policy effects: - Such reforms would improve credit allocation, reduce market distortions, and foster private sector investment and productivity. - Deeper financial markets and institutions are also needed to finance investment, including in infrastructure. *Source: IMF staff estimates.* ### 18. The government’s recapitalization policy is evaluated by macroeconomic performance and bank profitability measures ### Model setup and measurement - Aggregate output, ܻ, and aggregate productivity, ܼ, are defined as ݕൌܻ ௛ ݕ൅ ௟ ݖൌ ௛ ݇ ௛ ݖ൅ ௟ ݇ ௟ and ൌܼ ݕ ௛ ݕ൅ ௟ ݇ ௛ ݇൅ ௟ . - Note: aggregate productivity increases if bank credit is more allocated to the H sector becauseݖ ௛ ݖ൐ ௟ ; credit misallocation to the L sector lowers aggregate output via decreasing ܼ. - Bank profitability measured by net interest margin (NIM) for loans and deposits: ൌ ܯܫܰ ܴ ௅.௛ ݇ ௛ ܴ൅ ௅,௟ ݇ ௟ ݇ ௛ ݇൅ ௟ ܴെ ஽. - Macroeconomic consequences of credit misallocation analyzed via comparative statics on these variables. ### Equilibrium NPL resolution and recapitalization - Equilibrium amount of NPL resolution, ࢞, is increasing with respect to the government’s capital injection, ࢋ ࡳ . - Government can facilitate NPL resolution by increasing public recapitalization, ݁ ீ . - Full NPL resolution corresponds to (ൌ1ݔ); policy implications are discussed separately for: - (i) second-best: partial NPL resolution (൏1ݔ), and - (ii) first-best: full NPL resolution (ൌ1ݔ). ### The case of partial NPL resolution (൏1ݔ, the second best) - Credit misallocation in partial resolution lowers aggregate productivity and entails growth cost. - As banks’ capital buffers increase, credit to the L sector proportionally decreases, but some L firms continue operating via forbearance lending. - Credit to the H sector is constant with respect to recapitalization when ൏1ݔ. - Aggregate output, ܻ, is slightly decreasing with respect to recapitalization during NPL resolution. - NIM increases as recapitalization proceeds in this stage; deposit rates constant; asset returns increase as bank reallocates credit from L to H sector; profitability rises. ### The case of full NPL resolution (ൌ1ݔ, the first best) - After full NPL resolution, credit to H sector increases with recapitalization. - Equilibrium lending rate ܴ ௅,௛ ∗ decreases with ߣ, so lending rate and credit to H sector ݇ ௛ increase with ݁ ீ . - Bank uses capital buffers to extend new lending rather than absorb NPL losses; NIM declines along with recapitalization. - Aggregate output increases with recapitalization after completing NPL resolution; ܻ response to ݁ ீ is V-shaped, kinked at ൌ1ݔ. - Further increases in recapitalization become ineffective once capital constraint becomes slack: ߣ = 0 implies lending rates, credit amounts, output, and profitability measures all constant with respect to ݁ ீ . ### Government policy on optimal amount of recapitalization (political economy equilibrium) - Government budget constraint: ݁ൌܶ ீ ,ܩ where ܶ is tax revenue, ݁ ீ is recapitalization, and ܩ is other government expenditures. - Government maximizes social welfare: ݕߙ൅ܻൌܷ ௟ ݂ߚ൅ ሺ ܩ ሻ - Optimal recapitalization cases: - (i) zero capitalization (ࢋ ࡳ ൌ૙), - (ii) partial capitalization (ࢋ ࡳ ൐૙ and ࣅ൐૙), and - (iii) full capitalization (ࢋ ࡳ ൐૙ and ࣅൌ૙). - Political economy implications: - Higher ߙ and ߚ make NPL resolution by recapitalization less likely. NPL resolution via public recapitalization will not take place if government has special interest to help SOEs (high ߙ) or values other expenditures more (high ߚ). - Because ݂(⋅) is concave (݂ ᇱᇱ (⋅)൏0), when tax revenues ܶ are high, government tends to choose capitalization by public money. ### Empirical evidence on credit misallocation and SOEs in Vietnam (2005–2015) - Dataset: annual Worldscope listed firms’ financial data, 2005–2015. Data cleaning: drop Telecommunications, Utilities, Financials (ICB: 6000-9000); remove outliers: roa outside [−0.17, 0.38], intr negative or >40%, d_debt >100%, d_ta >50%. - Variable definitions: - dum_soe = 1 if state share > 10 percent, 0 otherwise. - roa = EBIT/total asset. - roa_soe = dum_soe * roa. - intr = interest expenses/total debt. - d_debt = Year-on-year growth rate of total debt. - Findings on loan rates between SOEs and non-SOEs: - Estimation result: β ଵ ൌె0.0044 ሺݐ ݁ݑ݈ܽݒ: େ2.30ሻ. Interpretation: SOEs obtain credit at lower rates on average even after controlling for profitability and firm size. - Loan rates and profitability (ex-ante and ex-post): - For unprofitable SOEs (bad firm sample), relationship between loan rates and profitability is biased positive (i.e., β ଷ ൐0), implying underperforming SOEs obtain additional loans at lower loan rates. - Loan growth and profitability (forbearance lending test): - For SOEs, particularly SOEs in bad firm sample, relationship is biased negative compared with private firms (β ଶ ൏0), and no clear relationship between loan growth and profitability is observed for SOEs (β ଵ ൅β ଶ close to zero). - Interpretation: some SOEs obtain forbearance lending despite underperforming profitability. - Selected coefficients (Table summaries): - Table 1 (Loan Rates and Profitability) — Dependent Variable: inter t - roa t-1: -0.064*** | -0.154*** | -0.111* - roa t: 0.148*** | 0.216*** | 0.105** - roa_soe t-1: 0.015 | 0.030 | 0.183** - roa_soe t: 0.050 | -0.043 | 0.069 - Table 2 (Debt Growth and Profitability) — Dependent Variable: d_debt t - roa t-1: 0.731*** | 0.693 | 1.768*** - roa_soe t-1: -0.800** | -1.421* | -1.252** - Note: *, **, *** indicate significance at 10 percent, 5 percent, and 1 percent levels respectively. ### Conclusion and policy implications - Extent of bank recapitalization is a political economy equilibrium reflecting government’s special interest in low-productivity sector (SOEs) and its budget tightness. - Partial or full recapitalization may be chosen depending on parameters (ߙ, ߚ) and fiscal space (ܶ). - First-best (fast NPL resolution and full recapitalization) improves credit allocation and growth relative to second-best (partial recapitalization). - First-best requires government fiscal space (marginal cost of funds below a threshold); fiscal capacity is key for financial stability. - Empirical analysis of Vietnam supports the importance of policy regarding credit misallocation between SOEs and non-SOEs, showing evidence of preferential loan rates and forbearance lending to SOEs. *Source: IMF staff estimates and analysis contained in the provided chapter.* ### 1. Population Structure 2014 __________________________________________________________ 9 ### 1. Population Structure 2014 ### Climate change risks - Vietnam is ranked among the five countries likely to be most affected by climate change due to long coastline, geographic location, and diverse topography and climates. - Over the last 50 years: - temperatures have increased twice as fast as the global average; - the sea level has risen by 20 centimeters; - the frequency and intensity of extreme weather events (drought, flood, salinization) have risen sharply. - Natural disasters result in 470 fatalities and cost 0.8 percent of GDP (annual average between 1990 and 2016). - Authorities’ climate change scenarios project by the end of the century: - sea levels are expected to rise by up to a meter; - sea waters would cover 40 percent of the Mekong Delta area, 3 percent of coastal provinces and 20 percent of Ho Chi Minh City; - direct impacts could affect 10–12 percent of the population and reduce GDP by 10 percent. - Sectoral vulnerabilities highlighted: - agriculture, aquaculture, energy, transportation and tourism are the most affected sectors. - a 1 degree Celsius rise in summer temperatures is expected to raise energy demand by 1 percent. - a 1 meter sea-level rise would submerge 10,000 kilometers of roads. - Social impacts: - lower agricultural incomes, erosion of coastal land and deterioration of arable land could reverse poverty reduction and accelerate rural-to-urban migration. - Ho Chi Minh City, with population approaches close to 10 million, is projected to be among the cities in the region most affected by sea-level rise and increased storm surges. - Agronomic projections and observed impacts: - In 2016, rice production fell by 1.1 million tons due to salinization in the Mekong Delta. - Rice, grown by nearly 80 percent of Vietnamese farmers on 45 percent of the agricultural land, could see production decrease by 9 million tons (13 percent) by 2050 absent adaptation. ### Environmental and resource sustainability - Growth model and natural capital: - Rapid industrialization and resource-intensive development have reduced Vietnam’s stock of natural capital through depletion of mineral and non-mineral resources and increased air, land and water pollution. - Legacy war-related issues (dioxin hotspots, hereditary diseases and birth defects) add urgency to environmental upgrades. - Adjusted Net Saving (ANS): - ANS corrects Net National Savings for resource depletion, environmental degradation, and education spending. - Despite strong education spending, intensive oil production and CO2 and particulate emissions caused ANS to be strongly lower than NNS; in recent years lower oil production reversed that trend, and in 2014, the gap was close to zero. - Pollution sources and incidents: - Vietnam is one of the largest users of fertilizer worldwide; intensive livestock production, shrimp aquaculture, and fisheries practices contribute to land and water pollution and mangrove destruction. - In 2016, a steel plant released large amounts of toxic chemical waste, killing fish along a long stretch of coast and reducing tourism in affected provinces by about 30 percent. - Air pollution and health: - Vietnam is among the top ten countries affected by air pollution; levels of fine particulate matter in large cities and industrial zones are comparable to China’s. - Studies in Ho Chi Minh City show associations between air pollution concentration and hospital admissions for lung and respiratory problems. - GHG emissions are expected to double between 2010 and 2020 and triple by 2030; the entire population is exposed to levels exceeding WHO guidelines. - Energy sector and emissions: - Vietnam is the 20th largest user of coal-fired plants globally. - In 2014, coal produced a quarter of electricity; the Power Development Plan (PDP VII) projects 55GW of coal-fueled power in 2030, up from 14GW today. - If implemented, coal would then amount for 53 percent of the country’s energy mix. - A Harvard report estimates fatalities from air pollution due to coal-fired plants would rise from 4,300 to 25,000 per year. - Vietnam is the most energy intensive economy in Southeast Asia (energy intensity metric presented). ### Vietnam’s initiatives and commitments - Strategic frameworks and targets: - National Climate Change Strategy (NCCS) (issued 2011) objectives to 2050: (a) improve climate change resilience through mitigation and monitoring; (b) strengthen adaptability by lowering risks; (c) adopt a more sustainable, low-carbon growth model. - National Green Growth Strategy (NGGS) (approved 2012) targets: improve energy efficiency, limit sectors generating pollution and natural resource degradation, develop sustainable infrastructure (transportation, energy, irrigation and urban works), and reduce GHG emissions. - Renewable energy and efficiency initiatives: - Renewable Energy Development Project (with World Bank) plans to increase share of renewable energy to 30 percent in 2030 by providing credit support and facilitating private participation. - General Electric investment: US$1.5 billion project to produce up to 1,000 megawatts of wind-generated electricity by 2018. - Other measures: forest plantation increases, energy-efficiency regulations for new office and commercial buildings, and development of urban public transportation networks. - International commitments: - By ratifying the Paris Agreement in 2016, Vietnam committed to reduce GHG emissions by 2030 via Nationally Determined Contributions (NDC): - an 8 percent reduction compared to business as usual financed by domestic resources; - a 25 percent reduction with international support. - Vietnam is dedicated to achieve the Sustainable Development Goals by 2030, many containing environmental and climate targets. - Institutional arrangements: - National Committee on Climate Change (NCCC) established 2012, chaired by the Prime Minister, to lead, coordinate, harmonize and monitor climate change and green growth program implementation. - Four policy implementation programs: (i) National Target Program to Respond to Climate Change (NTP-RCC); (ii) Support Program to Respond to Climate Change (SP-RCC); (iii) National Scientific and Technological Program on Climate Change; (iv) National Target Program on Energy Efficiency and Conservation (NTP-EE). ### Policy considerations and recommendations - Financing and coordination challenges: - Climate change financing is a critical challenge for public financial management; comprehensive costing and identification of financing for adaptation and mitigation projects remains difficult. - Numerous national strategies with sometimes conflicting objectives must be harmonized across national, provincial and sectoral levels. - Market-based and regulatory instruments: - Environmentally friendly tax and subsidy reforms and independent environmental regulation are important policy tools. - Pigouvian taxes and subsidies can reflect social costs and benefits, induce environmentally friendly behavior, and incentivize green investments. - Arms-length regulation is essential. - Energy- and fuel-pricing recommendations: - IMF Fiscal Affairs Department research estimates tax on gasoline in Vietnam should be increased by 0.46 US dollar per liter (0.43 US dollar for diesel) to fully price negative externalities from fuel consumption. - The government’s proposal to double the environmental protection tax on petroleum consumption—from an average of 0.11 to 0.24 US dollar per liter—is noted as a positive development but requires consideration of impacts on vulnerable populations and transparency on revenue use. - Electricity tariffs should be reviewed to account for externalities from energy production, incentivize investment in energy efficiency projects and stimulate private investment in renewable energy. - Private investment and business climate: - Measures to improve the business climate are critical to boost private and foreign investment, especially for large-scale infrastructure projects. ### Demographics (overview) - Vietnam has enjoyed a substantial demographic dividend in recent decades. - Rapid projected aging could weigh on economic growth and necessitate social security and fiscal reforms. - Managed well, the demographic transition can coincide with longer, healthier, more prosperous and more productive lives for Vietnamese people. *International Monetary Fund — Vietnam staff report content excerpt (May 23, 2017).* ### 1. Vietnam is a young country, but ### 1. Vietnam is a young country, but ### Demographic transition and projections - Median age: 26. - Largest age cohorts: between 20 and 34. - Declining fertility rates since the early 1970s and rising life expectancy will increase the population’s median age. - Population growth rate is projected to fall close to zero by 2050. - Working-age population share peaked in 2013 and is projected to decline over coming decades. - Share of population age 65 and older is projected to reach close to 3½ times the current level by 2050. - Projections are based on the UN's medium fertility variant (UN Population Division, World Population Prospects, 2015 revision). ### Speed of aging and income context - Number of years for the old-age dependency ratio to increase from 15 percent to 20 percent: - Vietnam: 8 years. - Comparative note: 26 years in Europe; more than 50 years in the United States. - Vietnam is at risk of becoming old before becoming rich; among selected economies, Vietnam’s working-age population reached its peak at the lowest relative per capita income level. - Figure 5 context: per capita income at purchasing power parity relative to the United States at the peak of working-age population share — Vietnam ranks lowest among listed economies. ### Demographics as headwinds to growth - Baseline assumptions for estimating growth impact of demographic-induced labor force changes: - Unchanged total factor productivity (TFP) growth. - Unchanged age- and gender-specific labor force participation rates (and employment rates). - Constant capital-to-effective-labor ratio. - Labor force size changes are estimated to add close to ½ percentage point to average annual growth between 2020 and 2050 (Figure 6). - The growth effect could fall to zero by mid-century. - On a per capita basis, a declining labor force size is estimated to subtract 0.1 percentage points from annual average growth between 2020 and 2050. ### Productivity and workforce aging - Empirical finding: an increase in the share of older workers (55-65 years) is associated with a significant reduction in labor productivity growth (sample of Asian and European countries). - For Vietnam: - Share of older workers projected to increase by close to 10 percentage points by 2050. - This could reduce growth by 0.2 percentage points per year. - Productivity impact differs by occupation: - Productivity tends to decline with age in physically demanding professions (factory workers, construction, agriculture). - Productivity may increase with age in professions such as lawyers, managers, and doctors. - Vietnam has a high share of its workforce in professions where productivity tends to decline with age. ### Labor force participation and reallocation - Current LFPRs (labor force participation rates) are already high. - Scenario: a gradual increase in LFPRs by 6 percentage points by 2050 could increase growth by 0.2 percentage points per year. - LFPRs for men and women are among the highest in the region, limiting scope for increases. - Policy scope: - Increase female and older worker participation. - Tackle under- and low-productivity employment. - Structural issues: productivity outside the FDI sector remains very low; informality is high; 40 percent of the workforce is employed in agriculture. ### Fiscal challenges and pension system sustainability - Under current policies, age-related public expenditures (pensions and health care) are projected to increase by 8 percentage points of GDP by 2050, more than double the current level. - Most of the increase would be related to pensions. - Spending on education could fall with a declining number of young people, but this is unlikely to fully offset rising age-related expenditures. - On the revenue side, lower growth rates because of demographic change could weigh on fiscal revenues. Pension system features and issues: - Structure: defined benefit pay-as-you-go system covering public sector workers and formal private sector workers. - Retirement ages: 60 for men and 55 for women. - Problems: - Generous general and occupational early retirement rules reduce effective retirement ages and increase unfunded liabilities. - Benefit indexation has regularly exceeded CPI. - High accrual rates yield replacement rates of 75 percent after 30 and 25 years of service for men and women, respectively. - Pension fund reserves are poorly invested, with returns below GDP growth. - System is inequitable across private and public sector employees and across the covered and uncovered population. - Suggested non-pension old-age insurance: land reform to give peasants title to land that can be traded and sold could be an effective old age insurance mechanism for large parts of the rural population. ### Policy implications and recommendations - Short-run and long-run policy priorities: - Raise the economy’s growth potential and sustain high growth in the next decades to offset demographic headwinds. - Foster faster productivity growth while reducing the factor intensity of growth. - Key reform areas: - SOE and banking reforms to achieve a level playing field for the private sector and improve allocation of capital. - Growth-friendly fiscal consolidation to strengthen public finances before rapid aging sets in. - Pension reforms in the near term should: - Raise retirement ages. - Address early retirement rules. - Reform accrual rates and indexation. ### Financial deepening, credit allocation, and growth (summary of linked section) - Vietnam’s financial system is bank-centered and dominated by the state. - 2016 statistics: - Banking sector assets amounted to 194 percent of GDP. - Banking sector accounted for more than 96 percent of financial sector assets (insurance companies: 3 percent; securities and fund management companies: 1 percent). - Four major state-owned credit banks (SOCBs) account for 45 percent of banking sector assets and provide half of total credit. - Stock market capitalization increased to 33 percent of GDP, from 27 percent in 2015. - Credit growth: - Averaged 24 percent in the last ten years. - Credit-to-GDP ratio expanded by 4.8 percentage point per annum between 2000 and 2015 and reached 124 percent of GDP at end-2016. - Consequences of rapid credit growth: - Productivity of credit and rates of return to investment have deteriorated. - Real credit grew almost three times faster than real GDP. - Credit targets by bank and by sector create inefficiencies in capital allocation. - Share of SOEs in total credit declined to 15 percent of total in 2016, but some poorly performing SOEs remain highly leveraged. - Growing share of loans going to real estate, financial and personal sectors; real estate investment expanded by 29 percent per year on average in real terms over the last decade while average annual output growth in the sector was 5.2 percent over the same period. - SME financing constraints: - In 2015, access to finance was the main business environment constraint for SMEs. - Only 29 percent of small enterprises (1–20 employees) have an active line of credit versus 57 percent of large firms (+100 employees). - Domestic SMEs compete with SOEs and large domestic enterprises with preferential access to resources; as a result, SME investment is subdued and largely internally financed. - Macro implications: - Credit misallocation and SME financing constraints contribute to a low share of investment in GDP (down by 10 percentage points since 2005). - Non-state sector investment has stagnated around 12 percent of GDP. - Public investment efficiency is hampered by tight fiscal space and lack of coordination in a fragmented general government; local governments account for about 80 percent of total state budget investment and often select projects without a strategic approach linked to national priorities. *Source: cr17191 - 1. Vietnam is a young country, but (IMF).* ### 7. A new financial development index created by the Fund indicates that Vietnam’s ### 7. A new financial development index created by the Fund indicates that Vietnam’s ### Financial development index — findings - Vietnam’s financial sector development has improved during the last two decades but remains below emerging markets. - Components of the Financial Development Index reveal: - Financial depth and efficiency are at its peers’ level for both institutions and markets. - Access to financial institutions is low, owing to the relatively small amount of bank branches and ATMs per capita (respectively 4 and 24 per 100,000 adults). ### Access and capitalization implications for growth - Empirical analysis indicates a bell-shaped relationship between financial development and growth. - Financial development increases growth up to a certain point (between 0.45 and 0.7 on the FD index) after which further development has a negative impact. - Research found that the weakening effect on growth at higher levels of financial development can be attributed to financial deepening, rather than to higher access or greater efficiency. - For Vietnam specifically: - Estimates suggest that improving access to financial services to individuals and SMEs could help boost growth by an additional percentage point each year. ### Credit growth and financial stability — historical context and risks - Vietnam’s credit-to-GDP ratio is now close to levels reached during previous periods of macroeconomic instability, signaling potential risks ahead. - Past credit cycles led to deteriorating quality of bank balance sheets and higher inflation. - The most recent ones, in 2008 and 2011, were characterized by large shares of credit directed to SOEs and to the real estate sector. - They were followed by surges of inflation—over 20 percent—and engendered a sharp rise of NPLs and significant weaknesses in the banking sector, which are still weighing on the real economy. ### Credit-to-GDP: regression analysis and gap - Regression analysis indicates that credit in Vietnam is 40 percent of GDP greater than expected. - Method: a random-effects regression linking credit to fundamentals and using international panel data following Cottarelli and others (2004). - Fundamentals used: public debt, GDP per capita, inflation, capital account openness, bank regulation indicators, legal origin and a time trend. - Results imply an appropriate level of credit-to-GDP ratio would be around 80 percent and that the gap has widened in recent years. ### Policy recommendations and conclusions - Vietnam’s economy would reap large benefits from financial sector reforms. Specific reforms mentioned: - Bank and SOE reforms, for instance the phasing out of credit targets. - Lifting interest rate ceilings. - Establishment of a level playing field for access to credit. - Expected policy effects: - Such reforms would improve credit allocation, reduce market distortions, and foster private sector investment and productivity. - Deeper financial markets and institutions are also needed to finance investment, including in infrastructure. ### NPL resolution, recapitalization, and credit misallocation — empirical and model insights - Legacy of NPLs, scarcity of funds for recapitalization and resulting credit misallocation between favored sector (SOEs) and the rest of the economy (non-SOEs) is an important policy issue in Vietnam, weighing on efficiency and growth. - International experience on NPL resolution by AMCs (e.g., VAMC) highlights advantages and success factors: - Advantages: economies of scale for efficient NPL resolution; specialization allowing banks to focus on new lending. - Common factors for successful AMCs: (1) enough legal powers, (2) pricing NPLs at market prices with strict supervisory policies, and (3) ample capital buffers. - Examples of institutional powers and pricing approaches from international AMCs are described (e.g., Danaharta, KAMCO). - Weak capital buffers make banks reluctant to resolve NPLs; recapitalization by public funds is one option but purchasing NPLs at above-market prices can conceal public costs. - The paper develops a two-sector banking model (H sector and L sector) with emphasis on bank capital buffers and political economy constraints on recapitalization: - Model setup and key equations describe loan demand functions, bank balance sheet identity, NPL resolution choice x, and capital constraint .݀ߛ൒݁ - Optimality conditions yield deposit rate, lending rates, and equilibrium NPL resolution x*. - Result: x* is an increasing function of government capital injection, ݁ீ — i.e., capital injection by the government can help banks resolve more NPLs. - Calibrated comparative statics (parameter values chosen for illustration): - Capital ratio must be higher than 9 percent, which means ߛ = 0.1. - LGD is 30 percent, ߶ = 0.3. - The NPL in the initial period, ݇ത௟, is set so that the NPL ratio is equal to 12 percent. - Market return ܴெ and return in the good sector ݖ௛ are set to 1.06 based on the average growth rate in Vietnam. - Elasticity of deposit demand ߳ and curvature of production function ߠ are chosen so that the deposit and lending rate are equal to 5 percent and 9 percent, respectively. - Return in the bad sector ݖ௟ is set to 0.7 for illustration. - Initial value of equity, ݁̅, chosen so that the resolution of NPLs is 20 percent without capital injection, ݁ீ = 0. *Source: IMF staff estimates.* ### 18. The government’s recapitalization policy is evaluated by macroeconomic ### 18. The government’s recapitalization policy is evaluated by macroeconomic performance and bank profitability measures ### Model setup and measurement - Aggregate output, ܻ, and aggregate productivity, ܼ, are defined as ݕൌܻ ௛ ݕ൅ ௟ ݖൌ ௛ ݇ ௛ ݖ൅ ௟ ݇ ௟ and ൌܼ ݕ ௛ ݕ൅ ௟ ݇ ௛ ݇൅ ௟ . - Note: aggregate productivity increases if bank credit is more allocated to the H sector becauseݖ ௛ ݖ൐ ௟ ; credit misallocation to the L sector lowers aggregate output via decreasing ܼ. - Bank profitability measured by net interest margin (NIM) for loans and deposits: ൌ ܯܫܰ ܴ ௅.௛ ݇ ௛ ܴ൅ ௅,௟ ݇ ௟ ݇ ௛ ݇൅ ௟ ܴെ ஽. - Macroeconomic consequences of credit misallocation analyzed via comparative statics on these variables. ### Equilibrium NPL resolution and recapitalization - Equilibrium amount of NPL resolution, ࢞, is increasing with respect to the government’s capital injection, ࢋ ࡳ . - Under the capital constraint, the bank needs enough initial capital buffers, ݁ ஻ ݁൅ ீ , for NPL resolution to absorb credit losses; otherwise NPL resolution forces large credit reductions to meet capital requirement. - Government can facilitate NPL resolution by increasing public recapitalization, ݁ ீ . - Full NPL resolution corresponds to (ൌ1ݔ); policy implications are discussed separately for: - (i) second-best: partial NPL resolution (൏1ݔ), and - (ii) first-best: full NPL resolution (ൌ1ݔ). - Note: there is also a no NPL resolution case (x = 0) when bank capital buffer is very thin; this extreme case is not considered further. ### The case of partial NPL resolution (൏1ݔ, the second best) - Credit misallocation in partial resolution lowers aggregate productivity and entails growth cost. - Equilibrium credit across sectors: as banks’ capital buffers increase, credit to the L sector proportionally decreases, but some L firms continue operating via forbearance lending. Givenݖ ௛ ݖ൐ ௟ , aggregate productivity remains below full-resolution case. - Credit to the H sector is constant with respect to recapitalization when ൏1ݔ (Figure 1–2). - Lagrange multiplier for the capital constraint is constant at, ൌߣ ܴ ெ ሺ ߶1െ ሻ ݖെ ௟ ߛെ߶ ൐0. - Equilibrium lending rate and amount of credit constant with respect to ݁ ீ . - Aggregate output, ܻ, is slightly decreasing with respect to recapitalization during NPL resolution (Figure 1–3). - Because bank decreases lending to L sector but does not increase lending to H sector; aggregate productivity improves but total credit falls unless NPLs are fully resolved. - NIM increases as recapitalization proceeds in this stage (Figure 1–4). - Deposit rates constant; asset returns increase as bank reallocates credit from L to H sector; profitability rises, potentially attracting outside equity including foreign investors. ### The case of full NPL resolution (ൌ1ݔ, the first best) - Complete shutdown of credit to the low-productivity sector occurs once NPLs fully recognized and resolved with full recapitalization. Two subcases: - NPLs fully resolved but capital constraint still binding (ൌ1ݔ and ൐0ߣ). - NPLs fully resolved and capital constraint slack (ൌ1ݔ and ൌ0ߣ). - After full NPL resolution, credit to H sector increases with recapitalization (Figure 1–2). - Lagrange multiplier becomes ൌߣ ߠ ଶ ݖ ௛ ߛ ఏ ݁൫ ீ ݁൅ ஻ ݇߶െ ത ௟ ൯ ଵିఏ െ ܴ െ ߛ which is decreasing in ݁ ீ . - Equilibrium lending rate ܴ ௅,௛ ∗ decreases with ߣ, so lending rate and credit to H sector ݇ ௛ increase with ݁ ீ . - Bank uses capital buffers to extend new lending rather than absorb NPL losses; NIM declines along with recapitalization (Figure 1–4). - Aggregate output increases with recapitalization after completing NPL resolution (Figure 1–3). - ܻ response to ݁ ீ is V-shaped, kinked at ൌ1ݔ: ܻ slightly decreases when ൏1ݔ, then increases once ൌ1ݔ achieved. Credit misallocation suppresses growth by lowering aggregate productivity and restricting high-productivity firms’ access to credit. - Further increases in recapitalization become ineffective once capital constraint becomes slack: ߣ = 0 implies lending rates, credit amounts, output, and profitability measures all constant with respect to ݁ ீ . Government has no incentive to inject more capital once constraint slack. ### Government policy on optimal amount of recapitalization (political economy equilibrium) - Government budget constraint: ݁ൌܶ ீ ,ܩ where ܶ is tax revenue, ݁ ீ is recapitalization, and ܩ is other government expenditures. - Government maximizes social welfare: ݕߙ൅ܻൌܷ ௟ ݂ߚ൅ ሺ ܩ ሻ - ܻ is aggregate output; ݕ ௟ is output in low-productivity sector; ݂ܩ(⋅) is social welfare from government expenditures satisfying ݂ ᇱ (⋅)൐0 and ݂ ᇱᇱ (⋅)൏0. - Parameters ߙ and ߚ represent relative importance of ݕ ௟ and ݂ܩ(⋅): high ߙ means government favors low-productivity sector (SOEs); high ߚ means government values other expenditures more. - Optimal recapitalization cases: - (i) zero capitalization (ࢋ ࡳ ൌ૙), - (ii) partial capitalization (ࢋ ࡳ ൐૙ and ࣅ൐૙), and - (iii) full capitalization (ࢋ ࡳ ൐૙ and ࣅൌ૙). - Cases (i) and (iii) are corner solutions; (ii) is an interior solution satisfying first order condition (FoC) ߲݁ ܷ߲ ீ ⁄ ൌ0. - Sequential procedure to find government optimum: compute ݁ ீ∗ satisfying FoC ܷ߲ ߲݁ ீ ൌ ܻ߲ ߲݁ ீ ߙ൅ ݕ߲ ௟ ߲݁ ீ ݂ߚെ ᇱ ሺ ݁ੇܶ ீ ሻ ൌ0. - If no interior ݁ ீ satisfies FoC and ߲݁ ܷ߲ ீ ⁄ ൏0 for all ݁ ீ, government chooses corner ݁ ீ ൌ0. This corner more likely when ߙ and/or ߚ large. - If even when ߣ reaches zero ߲݁ ܷ߲ ீ ⁄ ൐0, define ݁ ீ∗ as minimum ݁ ீ bringing ߣ = 0; large ߚ makes this corner solution more likely. - Compare social welfare under ݁ ீ ൌ0 (ܷ ଴ ) and under ݁ ீ ݁ൌ ீ∗ (ܷ ∗ ). Government chooses ݁ ீ ݁ൌ ీ∗ if ܷ ଴ ܷె ∗ ൏0. - Political economy implications: - Higher ߙ and ߚ make NPL resolution by recapitalization less likely. NPL resolution via public recapitalization will not take place if government has special interest to help SOEs (high ߙ) or values other expenditures more (high ߚ). - Because ݂(⋅) is concave (݂ ᇱᇱ (⋅)൏0), when tax revenues ܶ are high, government tends to choose capitalization by public money. ### Empirical evidence on credit misallocation and SOEs in Vietnam (2005–2015) - Dataset: annual Worldscope listed firms’ financial data, 2005–2015. Data cleaning: drop Telecommunications, Utilities, Financials (ICB: 6000-9000); remove outliers: roa outside [−0.17, 0.38], intr negative or >40%, d_debt >100%, d_ta >50%. - Variables defined for regressions: - dum_soe = 1 if state share > 10 percent, 0 otherwise. - roa = EBIT/total asset. - roa_soe = dum_soe * roa. - intr = interest expenses/total debt. - d_debt = Year-on-year growth rate of total debt. - Findings on loan rates between SOEs and non-SOEs: - Regression: ܜܖܑ ܜ ൌβ ଴ ൅β ଵ ൅controls܍ܗܛ_ܕܝ܌ - Control variables include ܉ܗܚ ି૚ܜ , ܉ܗܚ ܜ , and log of total assets. - Estimation result: β ଵ ൌె0.0044 ሺݐ ݁ݑ݈ܽݒ: େ2.30ሻ. - Interpretation: SOEs obtain credit at lower rates on average even after controlling for profitability and firm size. - Loan rates and profitability (ex-ante and ex-post): - Regression specification includes lagged and contemporaneous roa, roa_soe, controls, firm fixed effects, year dummies; sub-sample analysis for “good firms” (ROA > median) and “bad firms” (ROA < median). - Observation: unprofitable SOEs obtain credit at lower rates in the ex-ante stage. For unprofitable SOEs (bad firm sample), relationship between loan rates and profitability is biased positive (i.e., β ଷ ൐0), implying underperforming SOEs obtain additional loans at lower loan rates. - Loan growth and profitability (forbearance lending test): - Regression: ܜ܊܍܌_܌ ܜ ൌβ ଴ ൅β ଵ ܉ܗܚ ି૚ܜ ൅β ଶ ܍ܗܛ_܉ܗܚ ି૚ܜ ൅control. Controls: total asset, year dummies, firm fixed effects. If forbearance lending to SOEs occurs, expect β ଶ ൏0. - Estimation results (Table 2 summary): relationship between loan growth and profitability is clearly positive for all firms on average (β ଵ ൐0) in whole sample and subsamples — profitable firms increase debt, unprofitable firms decrease debt. - For SOEs, particularly SOEs in bad firm sample, relationship is biased negative compared with private firms (β ଶ ൏0), and no clear relationship between loan growth and profitability is observed for SOEs (β ଵ ൅β ଶ close to zero). - Interpretation: some SOEs obtain forbearance lending despite underperforming profitability. - Selected reported coefficients from tables: - Table 1 (Loan Rates and Profitability) — Dependent Variable: inter t - Columns: all firms | good firms (above median) | bad firms (below median) - roa t-1: -0.064*** | -0.154*** | -0.111* - roa t: 0.148*** | 0.216*** | 0.105** - roa_soe t-1: 0.015 | 0.030 | 0.183** - roa_soe t: 0.050 | -0.043 | 0.069 - Table 2 (Debt Growth and Profitability) — Dependent Variable: d_debt t - Columns: all firms | good firms (above median) | bad firms (below median) - roa t-1: 0.731*** | 0.693 | 1.768*** - roa_soe t-1: -0.800** | -1.421* | -1.252** - Note: *, **, *** indicate significance at 10 percent, 5 percent, and 1 percent levels respectively. ### Conclusion and policy implications - Extent of bank recapitalization is a political economy equilibrium reflecting government’s special interest in low-productivity sector (SOEs) and its budget tightness. - Partial or full recapitalization may be chosen depending on parameters (ߙ, ߚ) and fiscal space (ܶ). - First-best (fast NPL resolution and full recapitalization) improves credit allocation and growth relative to second-best (partial recapitalization). - First-best requires government fiscal space (marginal cost of funds below a threshold); fiscal capacity is key for financial stability. - Empirical analysis of Vietnam supports the importance of policy regarding credit misallocation between SOEs and non-SOEs, showing evidence of preferential loan rates and forbearance lending to SOEs. *Source: IMF staff estimates and analysis contained in the provided chapter.* --- _Source: https://www.imf.org/-/media/files/publications/cr/2017/cr17191.pdf_