## _wp1046 ## Source details **Canonical URL:** [_wp1046](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1046.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1046.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1046.pdf.json) --- ### Introduction and research question - Trade balances are strongly countercyclical in emerging economies (EM) but weakly countercyclical or even acyclical in developed markets. - Literature context and motivation: - Mendoza (1991), Neumeyer and Perri (2005), and Aguiar and Gopinath (2007) are seminal; Engel and Wang (2007) study OECD countries. - Neumeyer and Perri (2005): different international credit market environments and countercyclical interest rate shocks explain a larger fraction of net exports in EM. - Aguiar and Gopinath (2007): EM experience shocks to trend growth rather than transitory fluctuations; strong income effects generate strongly countercyclical trade balances. - Paper’s angle: analyze composition and cyclical properties of trade-balance components using UN-NBER data for 1980-2000. ### Empirical findings on composition and cyclicality (1980–2000) - Imports - Composition and cyclicality of imports in EM and developed markets are similar. - Share of capital goods in total imports ≈ one-third (group medians reported below). - Capital goods and total imports are procyclical in both EM and developed economies. - Exports - Major divergence: most EM export few capital goods; developed economies export a much larger share of capital goods. - Capital-good exports are acyclical in EM. - Without procyclical durables, overall exports tend to be acyclical in EM. - Some EM countries show countercyclical real exports (examples in text: Argentina, Mexico). ### Composition and business-cycle features (sample and key statistics) - Sample: 17 emerging economies; also reports for 6 developed economies and 3 G-7 countries. Data period 1980-2000. - Capital-good import and export shares (Table 2, group medians) - Emerging: - X_K,$/X,$ median = 8.77 [5.99,25.65] - M_K,$/M,$ median = 37.24 [36.21,40.08] - X_K,$ S/Y median = 1.02 [0.52,5.64] - M_K,$ S/Y median = 6.61 [5.57,8.28] - Developed: - X_K,$/X,$ median = 31.66 [25.24,36.30] - M_K,$/M,$ median = 36.59 [30.94,37.63] - X_K,$ S/Y median = 8.52 [6.16,9.65] - M_K,$ S/Y median = 9.57 [7.66,11.79] - Business-cycle correlations with output Y (Table 3, group medians) - Emerging: - Correlation TB/Y with output = -0.66 [-0.71,-0.51] - Correlation TB_G/Y = -0.65 [-0.70,-0.49] - Correlation G/Y = 0.73 - Correlation CI = 0.80 - Developed: - Correlation TB/Y = -0.16 [-0.35,-0.12] - Correlation TB_G/Y = -0.09 [-0.35,-0.07] - Correlation G/Y = 0.39 - Correlation CI = 0.89 - Cyclicality of trade flows (Table 4, group medians) - Emerging: - Corr XX_K with Y = -0.03 [-0.24,0.10] - Corr X_K with Y = 0.01 [-0.11,0.06] - Corr X_K/X with Y = 0.22 [0.04,0.29] - Corr M with Y = 0.65 [0.54,0.80] - Corr M_K with Y = 0.59 [0.47,0.72] - Corr M_K/M with Y = 0.37 [0.11,0.51] - Developed: - Corr XX_K with Y = 0.38 [0.36,0.58] - Corr X_K with Y = 0.43 [0.29,0.55] - Corr X_K/X with Y = 0.00 [-0.07,0.17] - Corr M with Y = 0.51 [0.48,0.63] - Corr M_K with Y = 0.62 [0.16,0.73] - Corr M_K/M with Y = 0.29 [-0.07,0.32] - Volatility at business-cycle frequencies (Table 5, selected group medians) - Emerging group medians: - % standard deviation of Y = 2.67 - % standard deviation of TB/Y = 1.92 - % standard deviation of C/Y = 1.48 - Relative volatility measures for imports/exports summarized in table (see Table 5 for full cell-level values). - Developed group medians: - % standard deviation of Y = 1.17 - % standard deviation of TB/Y = 0.68 - % standard deviation of C/Y = 0.89 ### Mechanisms and model implications — intuition - Intertemporal equilibrium perspective - Consumer optimization determines correlation between external accounts and output endogenously. - For trade balance to be countercyclical, the pro-borrowing effect induced by an expansionary productivity shock must dominate the pro-saving effect. - Standard small open economy models with complete or incomplete markets and standard preferences fail to generate the observed strongly countercyclical trade balance in EM. - Two-sector innovation - Incorporates empirical facts: EM import equipment (capital goods) in large shares and exports are largely uncorrelated with domestic business cycle. - Compared to one-sector model, agents borrow more to take advantage of expansionary home productivity shocks because imported investment becomes relatively more attractive; trade balance becomes strongly countercyclical. ### The two-sector small open economy model — structure and key equations - Production (Cobb-Douglas) - Home sector: Y_Ht = A_t K_Ht^β N_Ht^(1-β) - Export/Tradable sector: Y_Et = B_t K_Et^β N_Et^(1-β) - Total hours: N_t = N_Ht + N_Et - Resource constraints and trade balance - Home: C_t + I_Ht ≤ Y_Ht - Traded sector: I_Ft + [(1+R̄)D_{t-1} + 𝓓(D_t)] ≤ D_t + Y_Et - Portfolio adjustment cost: 𝓓(D_t) = (ϑ/2)(D_t - D̄)^2 - Debt first-order condition (log-linearized): μ̂_t - D̂_t = μ̂_{t+1} + (R̄/(1+R̄)) R̂̄_t - Current account: ca_t ≡ D_{t-1} - D_t ≡ nx_t - R̄ D_{t-1} - Investment aggregator (CES, specialized to Cobb-Douglas with ζ = 0) - G(I_Ht, I_Ft) = I_Ht^{ω_H} I_Ft^{1-ω_H}; set ω_H = 0.50 in calibration - Investment price index P_It defined in equation (11) - Preferences - Expected lifetime utility U = E_t Σ_{t=0}^∞ β^t u(C_t, N_t) - Two specifications: - Cobb-Douglas (CD) utility: consumption tends to be too smooth; standard one-sector model generates procyclical trade balance. - Greenwood–Hercowitz–Huffman (GHH) preferences: remove income effect on labor supply; generate larger consumption and labor volatility and stronger borrowing response after productivity shocks. - Budget constraint (per-period) - P_Ht C_t + P_It G(I_Ht, I_Ft) + 𝓓(D_t) + ϕ(…) ≤ r_Kt K_t + w_t N_t + [D_t - (1+R̄) D_{t-1}] - Adjustment frictions - Capital adjustment costs: quadratic ϕ(…) = ζ/2 (K_{t+1} - K_t)^2 + ζ/2 (K_{E,t+1} - K_Et)^2 - Sectoral labor adjustment costs: quadratic to avoid implausible negative comovement of sectoral labor ### Parameterization (Section IV, Argentina quarterly calibration — selected values) - Calibration targets and parameter choices - Discount rate β = 0.96 (Table 6) - Capital exponent β (production) = 0.4 - Set to match average investment to GDP ratio of 20 percent (Argentina: 1980-2000) - GHH labor exponent parameter ζ = 0.60 → Frisch elasticity 1/ζ = 1.66 - Steady-state labor supply calibrated to 0.30 - Utility curvature γ = 2 - Investment aggregator elasticity ψ_I = 1 (ζ = 0 → Cobb-Douglas); share ω_H = 0.50 - Asset-market specification: small coefficient on interest rate premium; steady-state debt D̄ chosen so average trade balance to output ratio ≈ 1 percent (Argentina averages: trade balance/output = 0.77 percent; goods trade balance/output = 1.68 percent over 1980-2000) - Productivity shocks - Home: log(A_t) = ρ_a log(A_{t-1}) + ε_{A,t}; ε_{A,t} ~ N(0, σ_A^2); ρ_a chosen so output persistence ≈ 0.70 - Export: log(B_t) = ρ_B log(B_{t-1}) + ε_{B,t}; baseline experiment uses only A_t shocks; B_t shocks added to match export volatility ### Implications of the model (Section V) — simulations and moments - One-sector small open economy (selected results) - CD preferences: trade balance procyclical; Corr(output, TB/Y) = 0.58 (Table 7) - GHH preferences: trade balance approximately acyclical; Corr(output, TB/Y) = 0.01 (Table 7) - Two-sector model with GHH preferences - Impulse responses to a one standard deviation home productivity shock: - Home production (C_t and I_Ht) increase; price of home goods P_Ht falls - Export sector and imported investment good I_Ft unaffected by home productivity (small open economy price-taker) - Investment price index P_It falls less than P_Ht; relative price P_It / P_Ht increases - Economy borrows to import investment I_Ft → strongly countercyclical trade balance - Quantitative outcomes (Table 7 and Table 8 summaries) - Table 7 (selected correlations): - Data: Corr(y, tb/y) = -0.65 [-0.70,-0.49] - Two-Sector Model: -0.47 - One-Sector Model (Cobb-Douglas): 0.58 - One-Sector Model (GHH): 0.01 - Table 8 (actual and simulated business cycle moments, %): - Data: σ_y = 2.67; σ_c/σ_y = 1.92; σ_i/σ_y = 3.91; σ_i/σ_y (F?) = 6.40; σ_y^F = 2.99; σ_n = 1.92 - Two-Sector Model: σ_y = 2.70; σ_c/σ_y = 0.71; σ_i/σ_y = 3.52; σ_i/σ_y (F?) = 3.50; σ_y^F = 2.94; σ_n = 1.23 - One-Sector Model (Cobb-Douglas): σ_y = 2.50; σ_c/σ_y = 0.20; σ_i/σ_y = 2.90; σ_i/σ_y (F?) = 1.26; σ_y^F = 1.20 - One-Sector Model (GHH): σ_y = 2.65; σ_c/σ_y = 0.73; σ_i/σ_y = 2.62; σ_i/σ_y (F?) = 2.23; σ_y^F = 0.35 - Note: Model statistics are averages over 100 simulations of 200 periods; series detrended with the Hodrick-Prescott filter. ### Key conclusions and suggested extensions - Main findings - Typical EM: capital good imports sizable (median M_K/GDP = 6.98 percent reported in text); capital good exports small (< 2 percent median X_K/GDP). - EM trade balances strongly countercyclical (median correlation TB/Y with Y = -0.66); contrast with developed economies. - EM exports tend to be acyclical while imports (including capital goods imports) are procyclical; larger net capital good importers tend to have more countercyclical trade balances. - A two-sector small open economy model with GHH preferences, portfolio and capital/labor adjustment costs, and imported investment goods can generate a strongly countercyclical trade balance consistent with EM facts. - Suggested research extensions - Empirical: explain why export cyclicality differs across EM and developed countries; study common shocks and trade as a transmission channel. - Model: develop frameworks documenting how EM transition to developed status and analyze implications for trade balance behavior, consumption, and export variety at product level. *Source: _wp1046 - Section IV. calibrates the model. Section V. studies the implications of the two-sector model (IMF working paper content provided).* ### References.......................................................................................................... ... ### _wp1046 - References.......................................................................................................... ... ### Introduction and research question - Trade balances are strongly countercyclical in emerging economies (EM) but weakly countercyclical or even acyclical in developed markets. - Literature context: - Mendoza (1991), Neumeyer and Perri (2005), and Aguiar and Gopinath (2007) are seminal; Engel and Wang (2007) study OECD countries. - Neumeyer and Perri (2005): different international credit market environments and countercyclical interest rate shocks explain a larger fraction of net exports in EM. - Aguiar and Gopinath (2007): EM experience shocks to trend growth rather than transitory fluctuations; strong income effects generate strongly countercyclical trade balances. - Oviedo (2005) discusses conditions under which interest-rate shocks cause business cycles. - Eaton and Kortum (2001) and Duttagupta and Spilimbergo (2004) are noted for related evidence on export composition. - This paper’s angle: analyze composition and cyclical properties of trade-balance components using UN-NBER data for 1980-2000. ### Empirical findings on composition and cyclicality (1980–2000) - Imports: - The composition and cyclicality of imports in EM and developed markets are rather similar. - The share of capital goods in total imports is roughly a third. - Capital goods and total imports are procyclical in both EM and developed economies. - Exports: - Major divergence between EM and developed economies. - Most emerging economies export few capital goods; when they export more capital goods, it is only a selective set. - Capital-good exports are acyclical in EM. - Without procyclical durables, overall exports tend to be acyclical in EM. - Some EM countries (example: Argentina, Mexico) have countercyclical real exports. - In developed economies, exports mirror imports: capital goods constitute a third of overall exports and their strong procyclicality drives much of exports’ procyclicality. ### Mechanisms and model implications - Intertemporal equilibrium perspective: - Consumer optimization determines the correlation between external accounts and output endogenously. - For the trade balance to be countercyclical, the pro-borrowing effect induced by an expansionary productivity shock must dominate the pro-saving effect—standard small open economy models fail on this dimension. - General equilibrium models with incomplete markets and optimizing agents that replicate G-10 dynamics do not generate a sufficiently countercyclical trade balance. - Even the small open economy model with Greenwood et al. (1988) preferences fails to produce the observed countercyclicality in EM. - Two-sector model innovation: - Incorporates the empirical facts that EM need to import equipment and exports are uncorrelated with the domestic business cycle. - Otherwise similar to the standard small open economy model with preferences and adjustment costs in line with Schmitt-Grohe and Uribe (2003). - Compared to the one-sector model, agents need to borrow more to benefit from an expansionary productivity shock, and the trade balance becomes strongly countercyclical. ### Data, figures, and tables referenced (contents list) - Data coverage: UN-NBER trade data, 1980-2000. - Figures listed (titles only): - Business Cycles in Argentina, 1980–2000 - Business Cycles in Thailand, 1980–2000 - Median Capital Goods Trade Balance and Cyclicality of the Trade Balance, 1980–2000 - Impulse Response Function to a Productivity Shock in a Standard Small Open Economy Model - Impulse Response Function to Productivity Shock in the Home Sector - Tables listed (titles only): - Data Source - Capital Goods Import and Export Shares in EM and Developed Economies - Business Cycles Characteristics - Cyclicality of Trade Flows - Business Cycle Volatility in Emerging and Developed Economies - Parameter Values - Correlation - Actual and Simulated Business Cycle Moments - Appendices: - Appendix A: Data Sources - Appendix B: Small-Open Economy Model ### Key conclusions - The countercyclicality of current accounts in EM is driven importantly by cyclical demand for capital-good imports during expansions. - A two-sector small-open economy model that reflects EM trade composition (equipment imports and export acyclicality) can generate a strongly countercyclical trade balance, by increasing the borrowing response to productivity shocks relative to a one-sector model. *Source: _wp1046 - References.......................................................................................................... ...* ### Section IV. calibrates the model. Section V. studies the implications of the two-sector model ### _wp1046 - Section IV. calibrates the model. Section V. studies the implications of the two-sector model ### Composition and Business Cycle Features of Trade Flows - Sample: 17 emerging economies (EM); also reports for 6 developed economies and 3 G-7 countries. Data period referenced: 1980-2000 (table summaries). - Observation 1: Capital good import and export shares - Median share of capital goods imports M_K in total goods imports M_$: EM = 37.24; Developed = 36.59. - Median share of capital goods exports X_K in total goods exports X_$: EM = 8.77 percent; Developed = 31.66. - Wilcoxon rank-sum test: rejects equality of median capital good export shares at the 5 percent level; does not reject equality for capital good import shares. - Empirical note: Some EM (Brazil, the Philippines, South Korea) experienced capital goods export shares rising up to a third of total exports in the 1990s, concentrated in selective products (road vehicles and semiconductors; SITC Rev.2 subcategories 78 and 776). - Observation 2: Capital goods and GDP - Median share of capital goods exports X_K in GDP (EM) < 2 percent. - Median share of capital goods imports M_K in GDP (EM) = 6.98 percent. - In both EM and developed economies the share of capital goods imports in GDP ≈ share of equipment investment in GDP. - Observation 3: Cyclicality of trade balance - Correlation median of real GDP with trade balance over output TB/Y: EM = -0.66; Developed = -0.16. - Wilcoxon rank-sum test: rejects equality of correlations at the 1 percent level. - Results robust when excluding services (goods trade balance TB_G/Y). - Observation 4: Cyclicality of exports and imports - Median correlation of output with exports (X) in EM = -0.03 (exports acyclical on median); some EM show strongly countercyclical exports (Argentina: correlation X with Y = -0.60; Mexico = -0.70). - Capital good exports X_K similarly acyclical in EM; total imports M and equipment imports M_K are procyclical in EM. Group median correlation of equipment imports with output = 0.59. - Cross-country relation (figure 3 linear regression): a = -0.37 + 0.036 b, where a = cyclicality of goods trade balance, b = median of capital goods trade balance; t-statistic in parentheses for coefficients reported as (1.80) for intercept and (1) for slope (as printed). - Observation 5: Volatility at business cycle frequencies - Capital good imports M_K (trade data) are more volatile relative to output Y than investment from national accounts. For several countries the relative standard deviation of equipment imports is double that of investment from national accounts (table 5). - Relative standard deviations of total exports and imports are in the range of gross fixed capital formation from national accounts. ### The Two-Sector Small Open Economy Model (structure and key equations) - Agents and timing - Representative identical agents, price takers; per capita variables. - Production (Cobb-Douglas) - Home (non-tradable) sector: Y_Ht = A_t K_Ht^β N_Ht^(1-β). Total hours: N_t = N_Ht + N_Et. - Export (tradable) sector: Y_Et = B_t K_Et^β N_Et^(1-β). - Resource constraints and trade balance - Home: C_t + I_Ht ≤ Y_Ht. - Traded sector: I_Ft + [(1+R̄)D_{t-1} + 𝓓(D_t)] ≤ D_t + Y_Et. - Portfolio adjustment cost function: 𝓓(D_t) = (ϑ/2)(D_t - D̄)^2. - First-order condition for debt: μ_t [1 - ϑ(D_t - D̄)] = E_t[μ_{t+1} (1+R̄)]. Log-linearized: μ̂_t - D̂_t = μ̂_{t+1} + (R̄/(1+R̄)) R̂̄_t. - Current account: ca_t ≡ D_{t-1} - D_t ≡ nx_t - R̄ D_{t-1}, where nx_t is net exports (trade balance). - Investment aggregator (CES; specialized to Cobb-Douglas) - General: G(I_Ht, I_Ft) = [ω_H^{1-ζ} I_Ht^ζ + ω_F^{1-ζ} I_Ft^ζ]^{1/ζ}, elasticity ψ_I = 1/(1-ζ). - With ζ = 0 → Cobb-Douglas: G = I_Ht^{ω_H} I_Ft^{1-ω_H}; set ω_H = 0.50 in calibration. - Investment price index P_It defined in equation (11); equilibrium relation P_Ht = [((1-ω_H)/ω_H)(I_Ht/I_Ft)]^{-1/ψ_I} (equation 10). - Preferences - Expected lifetime utility U = E_t Σ_{t=0}^∞ β^t u(C_t, N_t), 0 < β < 1. - Two specifications: Cobb-Douglas (CD) utility (equation 13) and Greenwood–Hercowitz–Huffman (GHH) preferences (equation 14). - Properties: - CD: consumption tends to be too smooth; generates procyclical trade balance in standard one-sector small open economy. - GHH: removes income effect on labor supply; labor supply determined by current real wage w_t = -U_N/U_C = ϖ N_t^ζ; Frisch elasticity = 1/ζ; generates larger consumption and labor volatility, can produce borrowing response after productivity shocks. - Budget constraint (per-period) - P_Ht C_t + P_It G(I_Ht, I_Ft) + 𝓓(D_t) + ϕ(…) ≤ r_Kt K_t + w_t N_t + [D_t - (1+R̄) D_{t-1}], where ϕ(K_{t+1},K_E,t+1, K_t, K_Et) are capital adjustment costs. - Factor-market frictions - Capital adjustment costs (quadratic): ϕ(…) = ζ/2 (K_{t+1} - K_t)^2 + ζ/2 (K_{E,t+1} - K_Et)^2. - Sectoral labor adjustment costs (quadratic) to avoid implausible negative comovement of sectoral labor (parameter ϖ in first-order condition shown). ### Parameterization (Section IV) - Calibration targets and parameter choices (Argentina, quarterly frequency) - Discount factor β: matches average real interest rate on Argentine foreign debt (value in Table 6; referenced but not numerically restated in text excerpt). - Depreciation rate δ: set to match average investment to GDP ratio of 20 percent (average 1980-2000). - Capital exponent β (production functions) = 0.4. - GHH labor exponent 1 + ζ: baseline set with ζ = 0.60 → Frisch elasticity 1/ζ = 1.66. - Steady-state labor supply calibrated to 0.30 (via labor weight parameter ω_L). - Utility curvature γ (denoted ϵ in text) = 2. - Investment aggregator elasticity ψ_I = 1 (ζ = 0 → Cobb-Douglas); share ω_H = 0.50. - Adjustment cost parameters for aggregate and sector-specific capital: chosen to match observed volatility of aggregate investment and capital stock. - Asset-market: small coefficient on interest rate premium (as in Schmitt-Grohe and Uribe (2003) and Neumeyer and Perri (2005)); steady-state debt D̄ chosen so steady-state average trade balance to output ratio ≈ 1 percent (Argentina averages: trade balance/output = 0.77 percent; goods trade balance/output = 1.68 percent over 1980-2000). - Productivity shocks - Home productivity A_t: log(A_t) = ρ_a log(A_{t-1}) + ε_{A,t}; ε_{A,t} ~ N(0, σ_A^2). - Persistence ρ_a chosen so model generates output persistence ≈ 0.70 (typical EM serial correlation); σ_a chosen to match output volatility. - Export productivity B_t follows: log(B_t) = ρ_B log(B_{t-1}) + ε_{B,t}. In baseline experiment only A_t shocks are the exogenous driving forces; shocks to B_t are added to generate export volatility consistent with data. ### Implications of the Model (Section V) - Comparison one-sector vs two-sector; role of preferences - One-sector small open economy: - With CD preferences: trade balance is procyclical; correlation between output and TB/Y = 0.58 (table 7). - With GHH preferences: trade balance approximately acyclical; correlation between output and TB/Y = 0.01. - Mechanism: CD → smooth consumption → savings effect dominates; GHH → stronger consumption response → borrowing effect can dominate. - Two-sector model with GHH preferences (results) - Impulse responses to a one standard deviation home productivity shock: - Home production (C_t and I_Ht) increase; price of domestically produced goods P_Ht falls. - Export sector and imported investment good I_Ft unaffected (small open economy price-taker). - Investment price index P_It falls less than P_Ht; therefore real price of investment P_It / P_Ht increases following the shock. - Result: economy borrows more to import investment I_Ft, producing a strongly countercyclical trade balance. - Quantitative outcome: model produces a strongly countercyclical trade balance (first column of table 7; specific correlation value not numerically restated in text excerpt but described as “strongly countercyclical”). - Second moments and volatility (table 8 summary) - EM data median second moments reported in table 8 (not numerically restated in excerpt). - Model performance: - CD one-sector: consumption volatility relative to output is low; cannot match hours volatility. - GHH one-sector: matches higher consumption and hours volatility better. - Two-sector model: consistent with key observed features — investment more volatile than output; output more volatile than consumption; reproduces strongly countercyclical trade balance while matching second-moment patterns. ### Conclusions and Extensions - Main findings - Typical emerging economy: capital good imports are a sizable share of imports and of GDP (median M_K/GDP = 6.98 percent) while capital good exports are a small share of exports and of GDP (< 2 percent median X_K/GDP). - EM trade balances are strongly countercyclical (median correlation TB/Y with Y = -0.66); this contrasts with developed economies. - EM exports tend to be acyclical while imports (including capital goods imports) are procyclical; larger net capital good importers tend to have more countercyclical trade balances. - A two-sector small open economy model with GHH preferences, portfolio, capital and labor adjustment costs, and imported investment goods can generate a strongly countercyclical trade balance consistent with EM facts. - Suggested research extensions - Empirical: explain why export cyclicality differs across EM and developed countries; study common shocks and trade as a transmission channel. - Model: develop frameworks documenting how EM mature (transition to developed status) and analyze implications for trade balance behavior, consumption, and export variety at product level. *Source: _wp1046 - Section IV. calibrates the model. Section V. studies the implications of the two-sector model (IMF working paper content provided).* ### References ### _wp1046 - References ### References (Selected bibliography) - Aguiar, Mark, and Gita Gopinath, 2007, “Emerging market business cycles: the cycle is the Trend,” Journal of Political Economy, 115 (1), pp. 69-102. - Backus, David, Patrick J Kehoe, and Finn E Kydland, 1992, “International Real Business Cycles,” Journal of Political Economy, 100 (4), pp. 745-75. - Bekaert, Geert, and Campbell R. Harvey, 2000, “Foreign Speculators and Emerging Equity Markets,” Journal of Finance, 55 (2), pp. 565-613. - Bems, Rudolfs, 2008, “Aggregate Investment Expenditures on Tradable and Nontradable Goods,” Review of Economic Dynamics, 11 (4), pp. 852-83. - Boldrin, Michele, Lawrence J. Christiano, and Jonas D. M. Fisher, 2001, “Habit Persistence, Asset Returns, and the Business Cycle,” American Economic Review, 91 (1), pp. 149-66. - Christiano, Lawrence J., and Terry J. Fitzgerald, 1998, “The business cycle: it’s still a Puzzle,” Economic Perspectives, Q IV, pp. 56-83. - Correia, Isabel, Joao C. Neves, and Sergio Rebelo, 1995, “Business cycles in a small open Economy,” European Economic Review, 39 (6), pp. 1089-113. - De Long, Bradford, and Lawrence Summers, 1993, “How strongly do developing economies benefit from equipment investment?” Journal of Monetary Economics, 32(3), pp. 395-415. - Duttagupta, Rupa, and Antonio Spilimbergo, 2004, “What Happened to Asian Exports During the Crisis?” IMF Staff Papers, 51 (1):4. - Eaton, Jonathan, and Samuel Kortum, 2001, “Trade in capital goods,” European Economic Review, 45 (7), pp.1195-235. - Engel, Charles and Jian Wang, 2007, “International trade in durable goods: understanding volatility, cyclicality, and elasticities,” mimeo, University of Wisconsin. - Feenstra, Robert C., Robert E. Lipsey, Haiyan Deng, Alyson C. Ma, and Hengyong Mo, 2005, “World Trade Flows: 1962-2000,” NBER Working Papers, 11040, NBER. - Greenwood, Jeremy, Zvi Hercowitz, and Gregory W Huffman, 1988, “Investment, Capacity Utilization, and the Real Business Cycle,” American Economic Review, 78 (3), pp. 402-17. - Hsieh, Chang-Tai, and Peter J. Klenow, 2007, “Relative Prices and Relative Prosperity,” American Economic Review, 97 (3), pp. 562-85. - Kydland, Finn E, and Edward C Prescott, 1982, “Time to Build and Aggregate Fluctuations,” Econometrica 50 (6), pp. 1345-70. - Mendoza, Enrique G., 1991, “Real Business Cycles in a Small Open Economy,” American Economic Review, 81(4), pp. 797-818. - Navaretti, Giorgio Barba, Isidro Soloaga, and Wendy Takacs, 2000, “Vintage Technologies and Skill Constraints: Evidence from U.S. Exports of New and Used Machines”, World Bank Economic Review, 14 (1), pp. 91-109. - Neumeyer, Pablo A, and Fabrizio Perri, 2005, “Business cycles in emerging economies: the role of interest rates,” Journal of Monetary Economics, 52 (2), pp. 345-80. - Oviedo, Marcelo, 2005, “World Interest Rate, Business Cycles, and Financial Intermediation in Small Open Economies,” Working papers, Iowa State. - Prasad, Eswar, Kenneth S. Rogoff, Shang-Jin Wei, and M. Ayhan Kose, 2004, “Financial Globalization, Growth and Volatility in Developing Countries” NBER Working Papers, 10942. - Raffo, Andrea, 2008, “Net exports, consumption volatility and international business cycle Models,” Journal of International Economics, 75 (1), pp. 14-29. - Ravn, Morten and Harald Uhlig, 2002, “On Adjusting the Hodick-Prescott Filter for the Frequency of Observations,” The Review of Economics and Statistics, 84 (2), pp. 371-75. - Schmitt-Grohe, Stephanie, and Martin Uribe, 2003, “Closing small open economy models,” Journal of International Economics, 61 (1), pp. 163-85. ### Appendix A: Data sources and construction - National accounts: - National accounts data are from the IMF International Financial Statistics (IFS). - Emerging market classification follows Standard and Poor’s and the International Finance Corporation (IFC). - Real variables for GDP, consumption, gross fixed capital formation, imports and exports are obtained by dividing nominal components of GDP by the GDP deflator (as in Neumeyer and Perri (2005)). - Trade data: - Feenstra et al. (2005) provide bilateral trade data by commodity for 1962-2000. - Dataset constructed from United Nations data over two periods: (i) 1962-83 (SITC Rev.1, all trading partners) and (ii) 1984-2000 (UN comtrade, flows above $100,000 per year from 72 reporter countries, SITC Rev.2). - Feenstra et al. (2005) prioritize importer-reported flows; if importer report unavailable, exporter report is used. - Capital goods classification and measurement: - Eaton and Kortum (2001) approximate trade in capital equipment by goods associated with major equipment-producing industries: (i) electrical machinery, (ii) nonelectrical machinery, and (iii) instruments. - For each country the study constructs total capital good imports (M_K,$) and exports (X_K,$) of machinery and transport equipment corresponding in SITC Rev.2 to category 7: power-generating machinery and equipment (71), machinery specialized for particular industries (72), metalworking machinery (73), general industrial machinery and equipment (74), office machines and automatic data processing (75), telecommunications (76), electrical machinery, apparatus and appliance (77), road vehicles (78), and transport equipment (79). - Deflator series for imported and exported equipment are not available for most countries; an equipment deflator for a given country may differ from the price of equipment produced and exported from that country. - Purchasing power parity and price considerations: - Hsieh and Klenow (2007) find capital goods tend to be no more expensive in poor countries than rich countries; high relative price of capital in poor countries arises because consumption goods are much cheaper. - The U.N. International Comparison Program (ICP) collects price data on between 500 and 1500 individual goods and services, but data exist only for selected countries and years (benchmarks listed for 1970, 1975, 1980, 1985, 1990, 1996). - Penn World Tables PPP prices effectively reflect prices prevailing in rich countries per Hsieh and Klenow (2007). - Conversion to domestic currency and real quantities: - Dollar value of capital good imports M_K,$ and exports X_K,$ are converted to domestic currency M_K,DOM using spot exchange rate S_t. - Under freely traded capital goods, absolute purchasing power holds: P_Kt = P_K,$t * S_DOM=$t, where P_K is imported equipment price (foreign equipment price). - Real capital good imports M_Kt are calculated as: M_Kt = (M_K,$t * S_DOM=$t) / Defl_GDPt. ### Appendix B: Small-open economy model (baseline and equilibrium) - Preferences and utility: - The utility function is time-separable: E0 Σ_{t=1} U(C_t; L_t), with L_t leisure and employment N_t = 1 - L_t. - Popular specifications for U(.) used in the literature are Cobb-Douglas or Greenwood-Hercowitz-Huffman (GHH). - Budget constraint (equation form provided): - C_t + I_t + (1 + R^*_t-1) D_t-1 + φ(D_t) + ψ(K_t+1 - K_t) = Y_t + D_t; - R^*_t-1 denotes the world interest rate; D_t is the foreign (dollar denominated) non-indexed bond (net foreign asset position). - Timing for the bond follows Schmitt-Grohe and Uribe (2003). - Functional forms and adjustment costs: - Portfolio adjustment cost for bonds: φ(D_t) = (ζ/2) (D_t - d̄)^2 (implied by the text presenting quadratic adjustment). - First-order condition for bond holdings (text expression): ρ_t [1 - (D_t - d̄)] = β E_t [ρ_{t+1} (1 + R^*_{t+1})]; presented as Marginal Benefit = Marginal Cost of unit debt increase. - Capital adjustment cost specified quadratic: ψ(K_{t+1} - K_t) = η/2 (K_{t+1} - K_t)^2. - Equilibrium definition: - An equilibrium is a set of allocations and prices such that: - Households maximize utility subject to the budget constraint and capital accumulation technology given exogenous prices. - Factor markets clear; given import price of capital and demand for exports, firms maximize profit functions. - Markets clear. ### Tables and empirical moments (selected statistics and parameterization) - Table 2 (Capital Goods Import and Export Shares): Group Median values reported: - Emerging: X_K,$/X,$ median = 8.77 [5.99,25.65]; M_K,$/M,$ median = 37.24 [36.21,40.08]; X_K,$ S/Y median = 1.02 [0.52,5.64]; M_K,$ S/Y median = 6.61 [5.57,8.28]. - Developed: X_K,$/X,$ median = 31.66 [25.24,36.30]; M_K,$/M,$ median = 36.59 [30.94,37.63]; X_K,$ S/Y median = 8.52 [6.16,9.65]; M_K,$ S/Y median = 9.57 [7.66,11.79]. - Table 3 (Business Cycle Correlations with output Y): Group Median values: - Emerging: Correlation TB/Y with output = -0.66 [-0.71,-0.51]; Correlation TB_G/Y = -0.65 [-0.70,-0.49]; Correlation G/Y = 0.73; Correlation CI = 0.80. - Developed: Correlation TB/Y = -0.16 [-0.35,-0.12]; Correlation TB_G/Y = -0.09 [-0.35,-0.07]; Correlation G/Y = 0.39; Correlation CI = 0.89. - Table 4 (Cyclicality of Trade Flows): Group Median values: - Emerging medians: Corr XX_K with Y = -0.03 [-0.24,0.10]; Corr X_K with Y = 0.01 [-0.11,0.06]; Corr X_K/X with Y = 0.22 [0.04,0.29]; Corr M with Y = 0.65 [0.54,0.80]; Corr M_K with Y = 0.59 [0.47,0.72]; Corr M_K/M with Y = 0.37 [0.11,0.51]. - Developed medians: Corr XX_K with Y = 0.38 [0.36,0.58]; Corr X_K with Y = 0.43 [0.29,0.55]; Corr X_K/X with Y = 0.00 [-0.07,0.17]; Corr M with Y = 0.51 [0.48,0.63]; Corr M_K with Y = 0.62 [0.16,0.73]; Corr M_K/M with Y = 0.29 [-0.07,0.32]. - Table 5 (Business Cycle Volatility): Group medians and brackets reported: - Emerging group median for % standard deviation of Y = 2.67; % standard deviation of TB/Y = 1.92; % standard deviation of C/Y = 1.48; % standard deviation of M/X_K and others summarized as: 3.91 [3.32,4.70], 3.31 [4.58,7.60], 3.10 5.58 [4.58,7.60] (table contains cell-level values for many countries; preserve these group medians exact as reported). - Developed group median: % standard deviation of Y = 1.17; % standard deviation of TB/Y = 0.68; % standard deviation of C/Y = 0.89; other medians reported as 3.76 [3.04,3.97], 3.75 3.81 6.09 [5.93,7.20], 7.48 [7.32,8.12]. - Table 6 (Parameter values - selected): - Discount rate β = 0.96. - Curvature utility function γ = 2 (implied by notation 2 in table but preserved as shown). - Set to match average investment to GDP ratio of 20 percent (Argentina: 1980-2000) corresponds to parameter ϕ0.02 as reported. - Capital share α = 0.40. - Labor curvature θ = 0.60. - Elasticity of substitution for investment ϵ_I = 1 (Bems (2008)). - Share of home goods in investment H = 0.50. - Average trade-balance to output tb/y = 0.01 (Argentina: 1980-2000). - Table 7 (Correlation model vs data): - Data: Corr(y, tb/y) = -0.65 [-0.70,-0.49]; Corr F_ci = 0.59 [0.47,0.72]; Corr = 0.73, 0.80. - Two-Sector Model: -0.47, 0.87, 0.87. - One-Sector Model (Cobb-Douglas Preferences): 0.58, 0.96, 0.80. - One-Sector Model (GHH Preferences): 0.01, 0.98, 0.81. - Table 8 (Actual and simulated business cycle moments, Reported as %): - Data: σ_y = 2.67; σ_c/σ_y = 1.92; σ_i/σ_y = 3.91; σ_i/σ_y (F?) = 6.40; σ_y^F = 2.99; σ_n = 1.92. - Two-Sector Model: σ_y = 2.70; σ_c/σ_y = 0.71; σ_i/σ_y = 3.52; σ_i/σ_y (F?) = 3.50; σ_y^F = 2.94; σ_n = 1.23. - One-Sector Model (Cobb-Douglas): σ_y = 2.50; σ_c/σ_y = 0.20; σ_i/σ_y = 2.90; σ_i/σ_y (F?) = 1.26; σ_y^F = 1.20. - One-Sector Model (GHH): σ_y = 2.65; σ_c/σ_y = 0.73; σ_i/σ_y = 2.62; σ_i/σ_y (F?) = 2.23; σ_y^F = 0.35. - Note: Model statistics are averages over 100 simulations of 200 periods; series detrended with the Hodrick-Prescott filter. ### Figures and model impulse responses (qualitative summary) - Figure 1 and Figure 2: Business cycle visualizations for Argentina and Thailand, 1980-2000, showing series such as TB/Y, IMPEXPCapGood, and IMP. - Figure 3: Median Capital Goods Trade Balance and Cyclicality of the Trade Balance, 1980-2000 — scatter of country-level medians and cyclicality measures. - Figure 4: Impulse Response Function to a Productivity Shock in a standard small open economy model: - Panels display responses of Output (%), Labor (%), Consumption (%), Investment (%), and Trade Balance/Output for Cobb-Douglas and GHH preferences over periods 1 to 7. - Figure 5: Impulse Response Function to a Productivity Shock in the Home Sector: - Panels display responses over periods 1 to 7 for Output (%), Consumption (%), Investment (%), Relative Price of Investment (%), and Trade Balance/Output. *Source: _wp1046 - References (PDF chapter content provided).* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1046.pdf_