Extracting Common Factors to Classify Companies Listed in the Stock Exchange of Thailand by Using an Accounting Based Model

Abstract

The objective of this study is to classify the company listed in the Stock Exchange of Thailand by applying the Principle Component Analysis (PCA) to overcome the multicollinearity problem in the purpose of extracting the common factors used in the Logistic regression model. The data set are gathered from the companies’ annual financial statements registered in the Stock Exchange of Thailand since 2002 to 2016. The data set consists of 101 companies which there are only 29 distressed companies that have been defaulted on corporate bonds and there are 72 non-distressed companies that have not been defaulted on corporate bonds. The log-odds score are calculated based on the predicted default probability generated from the Logistic regression whose variables are integrated with the 4 statistically significant common factors. These scores are used to classify the company’s risk level. The overall accuracy of the predicted model is 92.3% which we can group the companies’ default probability into 5 different categories as very good, good, rather good, rather not good, and not good with a credit score below 475, 475-500, 500-535, 535-570, and above 570 respectively in each of which the percentages of companies contained in each categories are 58.42, 12.87, 4.95, 8.91, and 14.85 respectively.

Authors and Affiliations

Krisada Khruachalee

Keywords

Related Articles

M-Learning Transformation Framework and Strategy in Mobile Economy

Mobile technology is a powerful tool for driving organizational performance. Mobile creates learning opportunities including innovation, collaboration and research. For the Education sector, mobile connectivity provides...

Enhancing E-Service Quality of Airline Industry: an Empirical Study in Customers of Low Cost Airline

The purpose of this study examined the influence of e-service quality on loyalty intention of customer’s low cost airline. The questionnaire was tool to collected information from 295 customers of low cost airline in Tha...

Effectiveness of Building Inspection Policy Implementation in Bangkok, Thailand According to Building Control Act B.E. 2522 (A.D. 1979)

A quantitative research is applied as a main research methodology together with a qualitative research. In a quantitative research, questionnaires are used as a data gathering method from study population which is 140 au...

Hybrid Genetic Algorithms Feature Selection and Decision Trees Classifier for Student’s Computer Self-Efficacy

Data mining techniques such as Decision Tree have been applied to the field of retail sale, e-commerce, banking etc. Data mining, the extraction of hidden relationship information from large dataset, is a powerful new te...

IT Capability and Business Data Analytics Affect Firm Performance through Market Orientation

Information technology (IT) becomes more important for business operation. Data management and marketing capability also support the better firm performance. This research aims to study the utilization of IT capability a...

Download PDF file
  • EP ID EP598732
  • DOI -
  • Views 174
  • Downloads 0

How To Cite

Krisada Khruachalee (2018). Extracting Common Factors to Classify Companies Listed in the Stock Exchange of Thailand by Using an Accounting Based Model. International Journal of the Computer, the Internet and Management, 26(1), 58-63. https://europub.co.uk/articles/-A-598732