TY - JOUR
T1 - Random coefficient and errors-in-variables models for beta estimates
T2 - Methods and applications
AU - Lee, Cheng F.
N1 - Funding Information:
This researchh as been partially supported by the Research Board of the University of Illinois at Urbana-Champaign.T he author is grateful to two refereesfo r their helpful commentsI.n addition, he is grateful to ProfessorR . P. Beattya nd V. J. Defeo for their useful suggestions.
PY - 1984/12
Y1 - 1984/12
N2 - Based upon both theoretical and empirical arguments, the market model has been specified as a random coefficient and errors-in-variables (RCEV) rates of return generating process. The impacts of measurement errors associated with market rates of return and the risk-free on the estimated beta coefficient and estimated random coefficient parameters are analyzed in detail. It is shown that the beta coefficient obtained from RCEV can be decomposed into a) true component, b) bias due to measurement errors, c) bias due to specification error, and d) interaction bias.
AB - Based upon both theoretical and empirical arguments, the market model has been specified as a random coefficient and errors-in-variables (RCEV) rates of return generating process. The impacts of measurement errors associated with market rates of return and the risk-free on the estimated beta coefficient and estimated random coefficient parameters are analyzed in detail. It is shown that the beta coefficient obtained from RCEV can be decomposed into a) true component, b) bias due to measurement errors, c) bias due to specification error, and d) interaction bias.
UR - http://www.scopus.com/inward/record.url?scp=34548809330&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=34548809330&partnerID=8YFLogxK
U2 - 10.1016/0148-2963(84)90028-6
DO - 10.1016/0148-2963(84)90028-6
M3 - Article
AN - SCOPUS:34548809330
SN - 0148-2963
VL - 12
SP - 505
EP - 516
JO - Journal of Business Research
JF - Journal of Business Research
IS - 4
ER -