Working Paper: CEPR ID: DP15114
Authors: Claudia Foroni; Massimiliano Marcellino; Dalibor Stevanovic
Abstract: We consider simple methods to improve the growth nowcasts and forecasts obtainedby mixed frequency MIDAS and UMIDAS models with a variety of indicators during theCovid-19 crisis and recovery period, such as combining forecasts across various specificationsfor the same model and/or across different models, extending the model specificationby adding MA terms, enhancing the estimation method by taking a similarity approach,and adjusting the forecasts to put them back on track by a specific form of intercept correction.Among all these methods, adjusting the original nowcasts and forecasts by an amountsimilar to the nowcast and forecast errors made during the financial crisis and followingrecovery seems to produce the best results for the US, notwithstanding the different sourceand characteristics of the financial crisis. In particular, the adjusted growth nowcasts for2020Q1 get closer to the actual value, and the adjusted forecasts based on alternative indicatorsbecome much more similar, all unfortunately indicating a much slower recovery thanwithout adjustment and very persistent negative effects on trend growth. Similar findingsemerge also for the other G7 countries.
Keywords: COVID-19; forecasting; GDP; mixed-frequency
JEL Codes: No JEL codes provided
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
Cause | Effect |
---|---|
Adjustment of original nowcasts and forecasts (C53) | More accurate nowcasts for 2020 Q1 (C53) |
Adjustment of original nowcasts and forecasts (C53) | Closer to actual observed values (C51) |
Adjustment of original nowcasts and forecasts (C53) | Slower recovery than without adjustments (E19) |
Adjustment of original nowcasts and forecasts (C53) | Convergence of predictions across various models (C52) |
Adjustment of original nowcasts and forecasts (C53) | Improved reliability (L15) |
Adjustment of original nowcasts and forecasts (C53) | Persistent negative effects on trend growth (F62) |
Adjustment of original nowcasts and forecasts (C53) | Effective across different contexts (G7 countries) (F68) |