Risk factor investing has grown in popularity in recent years and has become a cornerstone of investment portfolios. The goal of factor investing is to generate more returns in the long run. This paper aims to studies the term structure of equity factor. The authors consider the point of view of an American investor and use risk, diversification and performance measures.
The authors combine two methodologies as follows: wavelets and copulas. The authors use daily, weekly and monthly equity factor returns to calibrate wavelets and copulas. Copula functions are useful instruments to describe “joint” fluctuations in different markets, especially to capture nonlinearities, providing a reasonable alternative to the assumption of joint normality. To check robustness, the authors propose three different wavelet mother functions to decompose the data and three different time horizons and the authors add a complementary exercise based on performance and diversification measurement without wavelet transform.
The authors identify temporal horizons for which diversification benefits would be optimized with the decrease in the level of dependence or even the inversion of the dependence structure. Thus, investors seeking diversification with factor investing have to care about the considered horizon: size and book to market factors seem to provide better diversification in the short term. Momentum strategies seem to deliver better diversification in the long run. All the results are very consistent.
Very few papers have documented the diversification properties of the equity risk factors. While factors are built to capture systematic risk premia, their diversification properties are still poorly understood. It is necessary to take into account non-normality of risk factors and to study the diversification over different time horizons. The solution is to use wavelet methodology to decompose returns into temporal series of different maturities.
CitationDownload as .RIS
Emerald Publishing Limited
Copyright © 2020, Emerald Publishing Limited