Modeling and Analyzing the Investment Criteria of Venture Capitalist via Multivariate Statistical Analysis

• Main Authors: Kuo, Kuan-Chang, 郭冠樟
• Other Authors: Fang, Wenchang
• Format: Others
• Language: zh-TW
• Published: 2005
• Online Access: http://ndltd.ncl.edu.tw/handle/17686571244113916488

博士 === 國立臺北大學 === 企業管理學系 === 93 === Past researches have cited the evaluation criteria used by the venture capitalists (Tyebjee & Bruno, 1984; Fried & Hisrich, 1994; Hisrich & Jankowicz, 1990). A few studies use financial and non-financial variables to discuss the investment criteria. Our study adopts the concept of Intellectual Capital to build the investment model. Intellectual Capital includes both financial and non-financial evaluation criteria. The objectives of this study described here are (1) Using Intellectual Capital to develop the investment criteria of venture capitalist, (2) Define the investment criteria by intellectual capital and financial ratio, (3) Apply Prospectus to select the measurable financial indicators, and (4) Comparing the accuracy of prediction models. Our research is divided into two sections. The first section is to develop an intellectual capital framework for investment criteria of venture capitalist. We use In-depth Interviews and Analytic Hierarchy Process. In this study, we use eight indicators to define Intellectual Capital: Human Capital, Relational Capital, Structural Capital, Innovational Capital, Financial Structure, Liquidity, Management Efficiency, and Profitability. The second section is to build the investment models by intellectual capital framework. We use Multivariate Statistical Analysis (Logit regression, Back-propagation neural network, and Support vector machine) to build and compare the accuracy of prediction models. Our study finds that VC managers valued more un-financial dimensions (0.5603) than financial dimensions (0.4397). This result is similar with Macmillan (1985) and Liu (1997) findings for interviews with VCFs. By comparing the prediction accuracy between SVM, Logit, and BPN, we find out that SVM models have the best prediction accuracy for IPOs to issue stock. Our recommendations for future research are to extend the prediction model in other periods (Seed period, Start up period, and Turnaround period) and other industries.