IPO mechanism selection by using Classification and Regression Trees

作     者：Kucukkocaoglu, Guray Alp, Ozge Sezgin 

作者机构：Baskent Univ Dept Management Fac Econ & Adm Sci TR-06490 Ankara Turkey Baskent Univ Fac Commercial Sci Dept Accounting & Financial Management TR-06490 Ankara Turkey 

出 版 物：《QUALITY & QUANTITY》 (质与量)

年 卷 期：2012年第46卷第3期

页      面：873-888页

核心收录：

学科分类：0303[法学-社会学] 03[法学] 0714[理学-统计学（可授理学、经济学学位）] 

主　　题：IPO selling mechanisms Classification and Regression Trees 

摘      要：The Turkish IPO market gives issuers and underwriters a choice of three different IPO selling mechanisms. The current paper sheds new light on the determinants of these issue procedures within the context of the following methods (i) book building mechanism, (ii) fixed price offer, and (iii) sale through the stock exchange. Most of the empirical models in the IPO literature use binary probit and logit models to determine the factors behind the choice of one method over another and try to answer the question of why is such a mechanism chosen. To understand the reasons on issuers selection of IPO mechanism, we have conducted a Classification and Regression Trees (CART) methodology to represent decision rules in a form of binary trees. Our results indicate that, CART methodology predicts a firms IPO selling mechanism with 77.42% accuracy. The most important variable that determines the IPO selling mechanism is the Arrangement Type between the issuer and the underwriter as in the form of best effort and firm-commitment.