Abstract
Latin America is a region marked by a constant and endogenous pattern of volatility that halts its development process. This article consists of empirically testing its volatility characteristics in terms of their regularities, using cycle theory, comparing it to other developing and developed regions. This paper (1) uses an asymmetric band pass filter decomposition to isolate economic cycles of distinct natures on the GDP growth time series for 136 countries in the Maddison Project Database, covering the period 1950-2018. (2) We calculate each country?s decomposed cycle amplitude and average duration, and (3) apply K-means clustering methods to classify the results into volatility groups, studying and understanding its features and characteristics. The main conclusions are that the majority of Latin American countries are subject to the relative dominance of the long-run economic cycles explaining the overall volatility, which could be linked to the high dependency in commodity exports, as changes in inputs caused by technology drive changes in specialization. Data shows that LAC is not the most volatile region of the world, as argued in the Structuralist literature. However, it has some common characteristics as a region in terms of the origin of its volatility.
Original language | English |
---|---|
Pages (from-to) | 7 |
Number of pages | 27 |
Journal | Cepal Review |
Volume | 139 |
Publication status | Published (VoR) - 30 Apr 2023 |
Keywords
- Macroeconomic Volatility
- Economic Cycles
- Time Series
- Filter Decomposition
- Cluster Analysis