Country Spreads and Emerging Countries: Who Drives Whom? Martin Uribe and Vivian Yue (JIE, 26)
Country Interest Rates and Output in Seven Emerging Countries Argentina Brazil.5.5...5.5.5. 94 95 96 97 98 99 Ecuador.4.3.2. 94 95 96 97 98 99.5..5 94 95 96 97 98 99 Mexico.5 94 95 96 97 98 99. Peru. Philippines.5.5.5 94 95 96 97 98 99 94 95 96 97 98 99 South Africa.6.4.2 94 95 96 97 98 99 Output Country Interest Rate 2
The Empirical Model A ŷ t î t tby t ˆR t us ˆR t = B ŷ t î t tby t ˆR t us ˆR t + ɛ y t ɛ i t ɛ tby t ɛ rus t ɛ r t Identification Assumptions: A is lower triangular R US t follows a univariate process Countries: Argentina, Brazil, Ecuador, Mexico, Peru, the Phillipines, South Africa. Sample Period: 994: to 2:4 3
Impulse Response To A Country- Spread Shock Output Investment.5.2..4.5.6.2.8.25.3 5 5 2.2 5 5 2 Trade Balance to GDP Ratio World Interest Rate.4.3.5.2..5 5 5 2 5 5 2 Country Interest Rate Country Spread.8.8.6.6.4.4.2.2 5 5 2 5 5 2 Point Estimate Error Band 4
Impulse Response To A World-Interest- Rate Shock Output Investment.5 2 3 4 5.5 6 5 5 2 5 5 2 2 Trade Balance to GDP Ratio World Interest Rate.5.8.5.6.4.2 5 5 2 5 5 2 Country Interest Rate Country Spread 3 2.5 2.5 2 2.5.5.5.5.5.5 5 5 2 5 5 2 Point Estimate Error Band 5
Impulse Response To An Output Shock Output 3 Investment.8 2.5.6 2.4.5.2.5 5 5 2 5 5 2 Trade Balance to GDP Ratio World Interest Rate..5.2.3.4.5.5 5 5 2 5 5 2 Country Interest Rate Country Spread.2.2.4.4.6.6.8.8 5 5 2 5 5 2 Point Estimate Error Band 6
Variance Decomposition Output Investment.25.2.3.25.2.5.5...5.5 5 5 2 quarters 5 5 2 quarters.4 Trade Balances to GDP Ratio 2 World Interest Rate.3.5.2..5 5 5 2 quarters 5 5 2 quarters Country Interest Rate Country Spread.8.8.7.7.6.6.5.5.4.4.3.3.2.2.. 5 5 2 quarters 5 5 2 quarters ɛ rus + ɛ r ɛ rus 7
Alternative Identification Scheme: Place Country Spreads first in the VAR system Implication: Output and investment expand in response to an increase in the world interest rate. Problem: It s difficult to rationalize this implication on theoretical grounds. 8
Aggregate Volatility With and Without Feedback of Spreads from Domestic Variables Variable Feedback No Feedback Std. Dev. Std. Dev. ŷ 3.65 3.7 î 4..93 tby 4.38 3.52 R 6.5 4.77 Result: Eliminating feedback of spreads from domestic variables reduces aggregate volatility by about 2 percent. Caution: The Lucas critique applies. We will redo this exercise using a theoretical optimizing model. 9
Summary of Empirical Findings. An increase in the world interest rate or in the country spread causes output and investment to fall and the trade balance to improve. 2. An increase in the world interest rate causes a delayed overshooting in the country spread 3. The effects of world-interest-rate shocks on domestic variables is measured with significant uncertainty. 4. US-interest-rate shocks account for 2 percent of aggregate fluctuations in Emerging Markets. 5. Country-spread shocks explain about 2 percent of aggregate fluctuations in EM 6. About 6 percent of movements in country spreads are explained by country-spread shocks.
The Theoretical Model Standard small open economy neoclasscial model with 3 modifications: Habit formation Gestation lags and convex adjustment costs in investment Working-capital constraint on firms
Households max E β t U(c t µ c t,h t ), t= subject to d t = R t d t w t h t u t k t + c t + i t +Ψ(d t ) i t = 4 3 i= s it. s i+t+ = s it k t+ =( δ)k t + k t Φ lim j E t d t+j+ j s= R t+s ( ) s3t k t 2
Decentralizing the Debt Adjustment Costs Domestic Banks: Borrow externally at rate R t Lend domestically at rate R d t Face operational costs Ψ(d t ) Compete atomistically for domestic deposits Domestic Banks Objective max d t Optimality Condition R d t [d t Ψ(d t )] R t d t R d t = R t Ψ (d t ) 3
Firms Evolution of the Firm s Debt Position d f t = Rd t df t F (k t,h t )+w t h t +u t k t +π t κ t +κ t Working-Capital Constraint κ t ηw t h t ; η Firm s Objective max E t= β t λ t λ π t Optimality Conditions F h (k t,h t )=w t [+η ( R d t R d t )] F k (k t,h t )=u t 4
Driving Forces ˆR t =.63 ˆR t +.5 ˆR t us +.35 ˆR t us t +.6ŷ t +.î t.2î t +.29tby t.9tby t + ɛ r t, ˆR us t =.83 ˆR us t + ɛrus t, where ɛ r t ɛrus t are mean-zero iid innovations with standard deviations equal to.3 and.7, respectively. 5
Functional Forms [ c µ c ω h ω] γ U(c µ c,h)= γ F (k, h) =k α h α Φ(x) =x φ 2 (x δ)2 ; φ> Ψ(d) = ψ 2 (d d) 2 Calibrated Parameters (Quarterly) ω =.45 γ =2 α =.32 R = β =.277 δ =.25 6
Estimating φ, ψ, η, and µ Criterion: Minimize the distance between empirical and theoretical Impulse Response Functions. Formally, φ, ψ, η, and µ are set so as to minimize [IR e IR m (ψ,φ,η,µ)] Σ IR e [IR e IR m (ψ,φ,η,µ)], Result: ψ =.2 φ = 28 η =.3 µ =.26 7
Theoretical and Estimated Impulse Response Functions Response of Output to ε rus Response of Output to ε r.5.5 5 5 2 Response of Investment to ε rus..2.3 5 5 2 Response of Investment to ε r 2 4 6 5 5 2 Response of TB/GDP to ε rus 2.5.5 5 5 2 Response of Country Interest Rate to ε rus 3 2 5 5 2.5 5 5 2 Response of TB/GDP to ε r.4.3.2. 5 5 2 Response of Country Interest Rate to ε r.8.6.4.2 5 5 2 Empirical IR Error Band -x-x Theoretical IR 8
Counterfactual Experiment : Country Spreads Don t Respond To The World Interest Rate Replace baseline Interest-Rate process with: ˆR t =.63 ˆR t + ˆR t us.63 ˆR t us t +.6ŷ t +.î t.2î t +.29tby t.9tby t + ɛ r t, Result: Aggregate volatility due to Rt us shocks falls by two thirds Most of the effects of world-interest-rate shocks on Emerging Countries are mediated through country spreads. 9
Counterfactual Experiment 2: Country Spreads Don t Respond To Domestic Fundamentals Replace baseline Interest-Rate process with: ˆR t =.63 ˆR t +.5 ˆR us t +.35 ˆR us t + ɛr t, Result: by ɛ r t Aggregate volatility explained jointly falls by one third. and ɛrus t 2
Summary. US-interest-rate shocks account for 2 percent of aggregate fluctuations in EM. 2. Country-spread shocks explain about 2 percent of aggregate fluctuations in EM 3. About 6 percent of movements in country spreads are explained by country-spread shocks. 4. US-interest-rate shocks affect domestic variables mostly through their effects on country spreads. 5. Domestic effects of world-interest-rate shocks are measured with significant uncertainty 6. The fact that country spreads respond to business conditions in EM exacerbates aggregate volatility in the region. 7. The US-interest-rate shocks and country-spread shocks identified in this paper are plausible in the sense that they imply similar business cycles in the context of an empirical VAR model as they do in the context of a theoretical dynamic general equilibrium model of the emerging economy. 2