Robust econometric inference for stock return predictability

Alexandros Kostakis, Tassos Magdalinos, Michalis P. Stamatogiannis

Research output: Contribution to conferencePaper

Abstract

Conclusions of empirical analyses on the existence of stock return predictability vary according to the time series properties of the economic and financial variables considered as potential predictors. Given the uncertainty regarding the degree of persistence of these variables, this study utilizes a very general modelling framework robustifying inference with respect to misspecification of regressor persistence. We conduct a battery of predictability tests for the period 1927-2007 extending the novel instrumental variable IVX estimation, developed by Phillips and Magdalinos (2009), to multivariate systems of predictive regressions with an intercept in the model. The resulting modified IVX approach is easy to implement and yields chi-squared inference for general linear restrictions on the regression coefficients that is robust to the degree of persistence of the predictor variables. In addition to extending the class of generating mechanisms for predictive regressions, this approach extends the range of testable predictability hypotheses by allowing joint inference on combinations of both explanatory and dependent variables.

Conference

Conference65th European Meeting of the Econometric Society
CountryNorway
CityOslo
Period25/08/1129/08/11

Fingerprint

Inference
Stock return predictability
Persistence
Econometrics
Predictability
Predictors
Predictive regressions
Economic variables
Misspecification
Linear restrictions
Instrumental variable estimation
Financial variables
Predictability of stock returns
Coefficients
Uncertainty
Modeling

Cite this

Kostakis, A., Magdalinos, T., & Stamatogiannis, M. P. (2011). Robust econometric inference for stock return predictability. Paper presented at 65th European Meeting of the Econometric Society, Oslo, Norway.

Robust econometric inference for stock return predictability. / Kostakis, Alexandros; Magdalinos, Tassos; Stamatogiannis, Michalis P.

2011. Paper presented at 65th European Meeting of the Econometric Society, Oslo, Norway.

Research output: Contribution to conferencePaper

Kostakis, A, Magdalinos, T & Stamatogiannis, MP 2011, 'Robust econometric inference for stock return predictability' Paper presented at 65th European Meeting of the Econometric Society, Oslo, Norway, 25/08/11 - 29/08/11, .
Kostakis A, Magdalinos T, Stamatogiannis MP. Robust econometric inference for stock return predictability. 2011. Paper presented at 65th European Meeting of the Econometric Society, Oslo, Norway.
Kostakis, Alexandros ; Magdalinos, Tassos ; Stamatogiannis, Michalis P./ Robust econometric inference for stock return predictability. Paper presented at 65th European Meeting of the Econometric Society, Oslo, Norway.
@conference{094c6478ff6e4510a33679eef267a4f9,
title = "Robust econometric inference for stock return predictability",
abstract = "Conclusions of empirical analyses on the existence of stock return predictability vary according to the time series properties of the economic and financial variables considered as potential predictors. Given the uncertainty regarding the degree of persistence of these variables, this study utilizes a very general modelling framework robustifying inference with respect to misspecification of regressor persistence. We conduct a battery of predictability tests for the period 1927-2007 extending the novel instrumental variable IVX estimation, developed by Phillips and Magdalinos (2009), to multivariate systems of predictive regressions with an intercept in the model. The resulting modified IVX approach is easy to implement and yields chi-squared inference for general linear restrictions on the regression coefficients that is robust to the degree of persistence of the predictor variables. In addition to extending the class of generating mechanisms for predictive regressions, this approach extends the range of testable predictability hypotheses by allowing joint inference on combinations of both explanatory and dependent variables.",
author = "Alexandros Kostakis and Tassos Magdalinos and Stamatogiannis, {Michalis P.}",
year = "2011",
month = "9",
language = "English",
note = "65th European Meeting of the Econometric Society ; Conference date: 25-08-2011 Through 29-08-2011",

}

TY - CONF

T1 - Robust econometric inference for stock return predictability

AU - Kostakis,Alexandros

AU - Magdalinos,Tassos

AU - Stamatogiannis,Michalis P.

PY - 2011/9

Y1 - 2011/9

N2 - Conclusions of empirical analyses on the existence of stock return predictability vary according to the time series properties of the economic and financial variables considered as potential predictors. Given the uncertainty regarding the degree of persistence of these variables, this study utilizes a very general modelling framework robustifying inference with respect to misspecification of regressor persistence. We conduct a battery of predictability tests for the period 1927-2007 extending the novel instrumental variable IVX estimation, developed by Phillips and Magdalinos (2009), to multivariate systems of predictive regressions with an intercept in the model. The resulting modified IVX approach is easy to implement and yields chi-squared inference for general linear restrictions on the regression coefficients that is robust to the degree of persistence of the predictor variables. In addition to extending the class of generating mechanisms for predictive regressions, this approach extends the range of testable predictability hypotheses by allowing joint inference on combinations of both explanatory and dependent variables.

AB - Conclusions of empirical analyses on the existence of stock return predictability vary according to the time series properties of the economic and financial variables considered as potential predictors. Given the uncertainty regarding the degree of persistence of these variables, this study utilizes a very general modelling framework robustifying inference with respect to misspecification of regressor persistence. We conduct a battery of predictability tests for the period 1927-2007 extending the novel instrumental variable IVX estimation, developed by Phillips and Magdalinos (2009), to multivariate systems of predictive regressions with an intercept in the model. The resulting modified IVX approach is easy to implement and yields chi-squared inference for general linear restrictions on the regression coefficients that is robust to the degree of persistence of the predictor variables. In addition to extending the class of generating mechanisms for predictive regressions, this approach extends the range of testable predictability hypotheses by allowing joint inference on combinations of both explanatory and dependent variables.

UR - http://eea-esem2011oslo.org/

M3 - Paper

ER -