Selecting a Regression Saturated by Indicators

TY - UNPB

T1 - Selecting a Regression Saturated by Indicators

AU - Hendry, David F.

AU - Johansen, Søren

AU - Santos, Carlos

N1 - JEL Classification: C51, C22

PY - 2007

Y1 - 2007

N2 - We consider selecting a regression model, using a variant of Gets, when there are more variables than observations, in the special case that the variables are impulse dummies (indicators) for every observation. We show that the setting is unproblematic if tackled appropriately, and obtain the finite-sample distribution of estimators of the mean and variance in a simple location-scale model under the null that no impulses matter. A Monte Carlo simulation confirms the null distribution, and shows power against an alternative of interest

AB - We consider selecting a regression model, using a variant of Gets, when there are more variables than observations, in the special case that the variables are impulse dummies (indicators) for every observation. We show that the setting is unproblematic if tackled appropriately, and obtain the finite-sample distribution of estimators of the mean and variance in a simple location-scale model under the null that no impulses matter. A Monte Carlo simulation confirms the null distribution, and shows power against an alternative of interest

KW - Faculty of Social Sciences

KW - indicators

KW - regression saturation

KW - subset selection

KW - model selection

M3 - Working paper

BT - Selecting a Regression Saturated by Indicators

PB - Department of Economics, University of Copenhagen

ER -