Tristiyanto, Tristiyanto (2015) Bankruptcy Prediction on Margin Trading and Applications. The science reports of the Kanazawa University = 金沢大学理科報告, 59. pp. 1-26.

Archive index.php_action=pages_view_main&active_action=repository_action_common_download&item_id=11119&item_no=1&attribute_id=26&file_no=1&page_id=13&block_id=21 - Published Version Download (3MB)

Abstract

Although bankruptcy and default are well known as critical factors leading to various financial recessions or financial bubbles, such as the Great Depression in the USA that began in 1929 and the Lost Decade of Japan in the 1990s, predicting when they will occur has not been studied sufficiently. In this paper, we propose a method that filters out risky investors and keeps good investors in a margin-trading simulation. Investors are divided into three classes (bankrupt, surviving and profitable) instead of the standard two (bankrupt/bad and surviving/good). As a result, bubble bursting can be thwarted, since maintaining credit absorption for the qualified investors can prevent the collapse of prices. We expose the problems with using the minimum margin as the de facto tool for controlling trading on the margin. We compare the results of four well-known data classification methods (multiple discriminant analysis, neural networks, decision trees and support vector machines) in order to determine the one that is most suitable for building a credit-scoring schema. Of the methods considered, the C4.5 algorithm for building decision trees was found to be the best. Our proposed strategy can successfully use credit scoring to tame the bubble phenomenon.

Item Type:	Article
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions:	Fakultas Matematika dan Ilmu Pengetahuan Alam (FMIPA) > Prodi Ilmu Komputer
Depositing User:	TRISTIYANT . .
Date Deposited:	21 Mar 2018 02:15
Last Modified:	21 Mar 2018 02:15
URI:	http://repository.lppm.unila.ac.id/id/eprint/5947

Actions (login required)

View Item