- Economies frequently experience large downturn beyond the scope of a 'normal' downturn.
- This project considers that the average size of a downturn can vary over time.
- I estimate this using the model
-
$y_t$ is the (observed) GDP component. -
$\sigma_t$ is unobserved and determines the standard deviation of GDP innovations. As$\sigma_t$ increases, the standard deviation of$y_t$ increases.
Argentina Industrial Production - Global Economic Monitor
Notebook | Description |
---|---|
Clean Data | Clean raw industrial production series to separate cycle and trend components. |
Main | Estimate stochastic volatility process. |
- Argentina features stochastic volatility - the standard deviation of GDP fluctuations changes over time.
- The volatility of GDP increased during recessions such as the 1998-2002 Argentine Great Depression and the 2008 Global Financial Crisis.
- Argentina experienced a 0.5% decline in GDP volatility following 1998-2002 Argentine Great Depression.
- pandas - data wrangling
- matplotlib/seaborn - visualizations
- pystan - bayesian methods