Online Program Home
  My Program

All Times EDT

Abstract Details

Activity Number: 92 - Time Series and Finance
Type: Contributed
Date/Time: Monday, August 9, 2021 : 10:00 AM to 11:50 AM
Sponsor: Business and Economic Statistics Section
Abstract #318963
Title: Granger Causality Test in Predictive Conditional Modal Regression
Author(s): Tae-Hwy Lee and Yaojue Xu*
Companies: University of California, Riverside and University of California, Riverside
Keywords: Mode; Modal regression; Granger Causality; Elicitibility
Abstract:

While a variable may not be predictable in mean using many macroeconomic and financial predictors, it may well be predictable in some quantiles especially in tails or in the mode. While the mode has its own merits relative to mean and median regressions, it has not been explored much in all disciplines. In this paper, we develop a test for Granger-Causality (GC) in the predictive regression for the conditional mode. The GC test is based on the seminal paper by Kemp and Silva (2012) and a recent paper on by Dimitriadis et al. (2019). It is in line with the recommendation of Ashley et al. (1980) to test for GC in out-of-sample prediction. We show that ENC statistic is asymptotically standard normal with zero mean under the null hypothesis of no GC in mode. Monte Carlo simulation shows ENC has a good size and power in finite samples.


Authors who are presenting talks have a * after their name.

Back to the full JSM 2021 program