Predicting Sino-Foreign Joint Venture Equity Control with Transaction Cost Factors: A Neural Network Analysis

摘要

Equity control is one of the key determinants of performance in international joint venture. This study employs the transaction cost framework to predict equity control via artificial neural networks (ANNs). Comparisons are made with the traditional statistical modeling approaches. ANNs produce a more parsimonious set of independent variables that yield a higher classification rates than with logistic regression. It can be concluded that ANNs with their complex, nonlinear structure are able to model the relationship between transaction cost factors and majority/minority ownership, and percent equity ownership more accurately than traditional methods.