Implementing an Approximate Dynamic Factor Model to Nowcast GDP Using Sensitivity Analysis

Abstract

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may CESifoAbstract Dynamic factor models based on Kalman Filter techniques are frequently used to nowcast GDP. This study deals with the selection of indicators for this practice. We propose a two-tiered mechanism which is shown in a case study to produce more accurate nowcasts than a benchmark stochastic process and a standard model including extreme bounds fragile indicators.Nowcasting accuracy nearly measures up to the one of real-time forecasts by an institution with an interest in high-quality nowcasts.JEL classification: C38, C53

The mutual predictability of Bitcoin and web search dynamics

“…The methods of the present study comprise a battery of multivariate time series techniques in the time and frequency domain such as testing for Granger‐causality in the frequency domain (Breitung & Candelon, 2006) and forecast accuracy testing between models in the time domain (Diebold & Mariano, 1995; Harvey et al, 1997). They have been successfully applied in diverse contexts; see, among many others, Tastan (2015), Aluko and Adeyeye (2020), and Fromentin and Tadjeddine (2020) for spectral Granger‐causality applications and Carstensen et al (2011) and Duarte and Süssmuth (2018) for applications of forecasting ability tests across models. To assess the cause–effect reaction time for frequency bands with significant Granger‐causality, a phase delay measure recently developed by Breitung and Schreiber (2018) is computed.…”

Composite global indicators from survey data: the Global Economic Barometers

“…E.g Abberger et al (2018). for Switzerland;Duarte and Süssmuth (2018) for Spain.23 For an overview of variable selection methods for partial least squares, seeMehmood et al (2012).24 This is the result of judgement based on economic reasoning. In some instances, the expectations are ambiguous, and we leave the outcome undetermined.…”