Real-time inflation forecasting with high-dimensional models: The case of Brazil

被引：43

作者：

Garcia, Marcio G. P. ^{[1
]}

Medeiros, Marcelo C. ^{[1
]}

Vasconcelos, Gabriel F. R. ^{[2
]}

机构：

[1] Pontifical Catholic Univ Rio de Janeiro, Dept Econ, Rua Marques de Sao Vicente 225, BR-22451900 Rio De Janeiro, Brazil

[2] Pontifical Catholic Univ Rio de Janeiro, Dept Elect Engn, Rua Marques de Sao Vicente 225, BR-22451900 Rio De Janeiro, Brazil

来源：

INTERNATIONAL JOURNAL OF FORECASTING | 2017年 / 33卷 / 03期

关键词：

Real-time inflation forecasting; Emerging markets; Shrinkage; Factor models; LASSO; Regression trees; Random forests; Complete subset regression; Machine learning; Model confidence set; Forecast combination; Expert forecasts; SELECTION; VARIABLES; LASSO; TESTS; SETS;

D O I：

10.1016/j.ijforecast.2017.02.002

中图分类号：

F [经济];

学科分类号：

02 ;

摘要：

We show that high-dimensional econometric models, such as shrinkage and complete subset regression, perform very well in the real-time forecasting of inflation in data-rich environments. We use Brazilian inflation as an application. It is ideal as an example because it exhibits a high short-term volatility, and several agents devote extensive resources to forecasting its short-term behavior. Thus, precise forecasts made by specialists are available both as a benchmark and as an important candidate regressor for the forecasting models. Furthermore, we combine forecasts based on model confidence sets and show that model combination can achieve superior predictive performances. (C) 2017 International Institute of Forecasters. Published by Elsevier B.V. All rights reserved.

引用

页码：679 / 693

页数：15

共 38 条

[1] Comparing density forecasts via weighted likelihood ratio tests [J].