Negative Binomial Model with an Application to Special Treatment Count Data

Author

Ni, Zhongxin ; Li, Ting

Author_Institution

Dept. of Finance, Shanghai Univ., Shanghai, China

fYear

2011

fDate

12-14 Aug. 2011

Firstpage

1

Lastpage

4

Abstract

Special Treatment (ST) is a particular regulation for the exchange to prompt the risk to investors. Since ST count data are discrete, traditional Orthogonal Least Square method cannot get better goodness of fitting. However, Poisson regression is appropriate to be used to analyze discrete data. In addition, once it is over-dispersed, Negative Binomial (NB) regression model is more suitable. In this paper, we study the influence of non-financial factors on ST occurrence event count with NB model. Result shows that China Interbank Offered Rate, Import, Export, Deposits of Enterprises and one-period lagged ST count are evidently associated with ST risk.

Keywords

Poisson distribution; banking; binomial distribution; international trade; investment; regression analysis; risk management; China interbank offered rate; Poisson regression; ST risk management; discrete data analysis; export; import; investor risk; negative binomial regression model; nonfinancial factor; special treatment count data; Companies; Data models; Dispersion; Economics; Fitting; Niobium; Predictive models;

fLanguage

English

Publisher

ieee

Conference_Titel

Management and Service Science (MASS), 2011 International Conference on

Conference_Location

Wuhan

Print_ISBN

978-1-4244-6579-8

Type

conf

DOI

10.1109/ICMSS.2011.5998339

Filename

5998339