Equity index returns predictability and fama-french factors

: a frequency domain analysis

Abstract

We extend the analysis of Faria and Verona (2020) tothe Fama French 5 factor model to predict the Equity Risk Premium in theS&P500 index. The Fama French 5 factor model is decomposed using themaximal overlap discrete wavelet transform so that we canstudy the forecasting performance of each factor’s different frequencies andtest them out-of-sample. The main findings of this study are that the factorsthemselves are not good predictors of the equity risk premiumout-of-sample, but the 16-128 month frequencyof the HML (high minus low) and, especially the RMW (robust minus weak),are good predictors of the Equity Risk Premium especially in post-2008 crisis. Ourresults support recent findings in the asset pricing literature that the business-cyclefrequency components of financial variables play a crucial role in forecastingthe equity premium. Thus, for both investors and academics, these findingsare of great relevance.

Date of Award	13 Jul 2022
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Gonçalo Faria (Supervisor) & Fabio Verona (Co-Supervisor)