TY - JOUR
T1 - Are macroeconomic factors adequate proxies for systematic influences in stock returns? A South African perspective
AU - Szczygielski, Jan J.
AU - Brümmer, Leon M.
AU - Wolmarans, Hendrik P.
AU - Zaremba, Adam
PY - 2020
Y1 - 2020
N2 - We investigate whether macroeconomic factors adequately proxy for systematic influences in stock returns within the South African context. We also investigate whether a commonly used solution to factor omission in macroeconomic factor models, the residual market factor, adequately reflects systematic influences not reflected by a set of macroeconomic factors. Our contribution lies in precisely quantifying the ability of macroeconomic and residual market factors to proxy for systematic drivers of returns. Systematic influences are represented by statistically derived factor scores which are then related to a set of carefully selected macroeconomic factors. We find that the identification of macroeconomic factors that proxy for systematic influences is a challenge in itself. Once identified, macroeconomic factors are poor and unstable proxies for systematic influences and the use of a residual market factor does not significantly improve the approximation of factor scores. Our conclusion is that macroeconomic linear factor models are likely to be underspecified, even if a residual market factor is included. This has implications for researchers, investors, econometricians and economists that rely on macroeconomic factor models to study financial markets.
AB - We investigate whether macroeconomic factors adequately proxy for systematic influences in stock returns within the South African context. We also investigate whether a commonly used solution to factor omission in macroeconomic factor models, the residual market factor, adequately reflects systematic influences not reflected by a set of macroeconomic factors. Our contribution lies in precisely quantifying the ability of macroeconomic and residual market factors to proxy for systematic drivers of returns. Systematic influences are represented by statistically derived factor scores which are then related to a set of carefully selected macroeconomic factors. We find that the identification of macroeconomic factors that proxy for systematic influences is a challenge in itself. Once identified, macroeconomic factors are poor and unstable proxies for systematic influences and the use of a residual market factor does not significantly improve the approximation of factor scores. Our conclusion is that macroeconomic linear factor models are likely to be underspecified, even if a residual market factor is included. This has implications for researchers, investors, econometricians and economists that rely on macroeconomic factor models to study financial markets.
KW - factor scores
KW - linear factor model
KW - macroeconomic factors
KW - residual market factor
KW - systematic influences
KW - underspecification
UR - http://www.scopus.com/inward/record.url?scp=85080873698&partnerID=8YFLogxK
U2 - 10.1080/10293523.2020.1723854
DO - 10.1080/10293523.2020.1723854
M3 - Article
AN - SCOPUS:85080873698
SN - 1029-3523
VL - 49
SP - 34
EP - 52
JO - Investment Analysts Journal
JF - Investment Analysts Journal
IS - 1
ER -