Modeling with Mixed Frequency Variables: A Review of Recently Extended Methods in Time Series Econometrics

Abstract

Recent theoretical econometric studies have focused on mixed frequency data. These studies are of great importance since they emphasize the role of information in economic modeling. In the current Time Series approach, temporal aggregation is often turned to a period of identical alternation; however, such aggregation leads to information loss in higher-frequency data. The mixed frequency studies provide a way to avoid the need for such temporal aggregation. In particular, the main result of this branch of econometric studies is to improve explanatory power, prediction, and efficiency in time series modeling with mixed frequency data. Accordingly, this paper attempts to specify the developments and shortcomings of this new branch of econometrics by reviewing the extant literature.

Keywords

• mixed frequency data
• temporal aggregation
• time series
• efficiency
• midas regression.