主  題:Adjusted Principal Component Analysis of Binary Factor Models

內(nèi)容簡(jiǎn)介:This paper considers binary factor models. In binary data, one additional factor arises and dominates even if there is no time-invariant factor or intercept in the true model. To this end, we introduce an adjusted Principle Component Analysis method. We provide theoretical foundations for this method. The proposed method has advantages of easy computation and robustness to many data generating processes. A large amount of experiments are carried out to numerically verify the good performance of the adjusted PCA in binary factor models and examine our findings. Finally, we apply the proposed method to the determinants of dividend initiation study using S&P500 firms' data from 1998 to 2016.

報(bào)告人:王曦      博士

時(shí)  間:2019-03-25    15:00

地  點(diǎn):位育樓117

舉辦單位:金融學(xué)院  科研部  經(jīng)濟(jì)與金融研究院