Scholarlink

<p>This paper investigates nonlinear business cycle synchronization among the five largest Latin American economies using wavelet coherence and smooth transition regression models. By combining time–frequency analysis with nonlinear regime-dependent modelling, the study captures the strength of co-movement across scales and different economic conditions. Wavelet coherence results show that business cycles synchronization becomes more pronounced at medium-term frequencies. This occurs particularly when business cycles are extracted using the Christiano–Fitzgerald filter. This indicates that regional business cycle interdependence is mainly a medium-term phenomenon and is sensitive to methodological choices. In contrast, Hodrick–Prescott-based cycles tend to underestimate economically meaningful co-movement. The results strongly reject linearity and reveal that spillover effects are significantly stronger in the low-volatility regime. The COVID-19 episode generates temporary synchronization that does not reflect stable structural linkages. Brazil and Mexico display limited regional integration, while Argentina, Chile, and Colombia exhibit stronger mutual dependence, with Colombia acting primarily as a receiver of regional fluctuations. The results suggest that synchronization is partial, asymmetric, and regime dependent, implying that deep macroeconomic integration remains premature. An informed understanding of this complexity enables policymakers and researchers to construct more reliable frameworks and to promote more effective regional coordination.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Are business cycles in Latin America synchronized? – New evidence from wavelet coherence and smooth transition regression

  • 期刊论文

  • 发表日期：

    2026年7月4日

查看全文
  • Dejan Živkov,
  • Slavica Manić,
  • Marina Gajić-Glamočlija

摘要

This paper investigates nonlinear business cycle synchronization among the five largest Latin American economies using wavelet coherence and smooth transition regression models. By combining time–frequency analysis with nonlinear regime-dependent modelling, the study captures the strength of co-movement across scales and different economic conditions. Wavelet coherence results show that business cycles synchronization becomes more pronounced at medium-term frequencies. This occurs particularly when business cycles are extracted using the Christiano–Fitzgerald filter. This indicates that regional business cycle interdependence is mainly a medium-term phenomenon and is sensitive to methodological choices. In contrast, Hodrick–Prescott-based cycles tend to underestimate economically meaningful co-movement. The results strongly reject linearity and reveal that spillover effects are significantly stronger in the low-volatility regime. The COVID-19 episode generates temporary synchronization that does not reflect stable structural linkages. Brazil and Mexico display limited regional integration, while Argentina, Chile, and Colombia exhibit stronger mutual dependence, with Colombia acting primarily as a receiver of regional fluctuations. The results suggest that synchronization is partial, asymmetric, and regime dependent, implying that deep macroeconomic integration remains premature. An informed understanding of this complexity enables policymakers and researchers to construct more reliable frameworks and to promote more effective regional coordination.