Research · Econometrics
Inflation forecasting
I collected macroeconomic data and implemented ARIMA/SARIMA models for research comparing inflation forecasts across 20 countries in the Americas.
Building the empirical foundation
Under the supervision of Pamela Córdova, PhD, I assisted with research comparing traditional econometric and neural-network approaches to inflation forecasting.
I led the collection and organization of a historical database spanning 11 macroeconomic indicators and aggregates for 36 countries across the Americas. I also guided fellow research interns in using Refinitiv Eikon, Excel, Stata, and EViews to collect and standardize the data.
From a broader database to a focused study
The modeling study focused on the 20 countries with the most complete datasets. Within that work, I directed the implementation of ARIMA and SARIMA models for 12-month CPI-derived inflation, supporting model estimation and forecasting in EViews and Stata.
My contribution was to the data and traditional econometric side of the comparison. The experience made the relationship between a modeling choice and the underlying data particularly concrete.
Result
A 36-country data collection effort supporting a 20-country forecasting study, with contributions to ARIMA/SARIMA estimation.
Next project