Home // eKNOW 2017, The Ninth International Conference on Information, Process, and Knowledge Management // View article

Predicting Corporate Bond Prices in Japan Using a Support Vector Machine

Authors:
Hiroaki Jotaki
Yasuo Yamashita
Hiroshi Takahashi

Keywords: Corporate Bonds; Prediction; SVM

Abstract:
Predictability of returns is one of the most important concerns in bond investment. In this study, we analyze the predictability of corporate bond prices after company announcements of financial results using a support vector machine (SVM). This paper will discuss (1) the highest hit ratio found when predicting the movement of corporate bond prices using the four variables of current net earnings, management earnings forecasts, ratings, and a leading composite index, and (2) the highest hit ratio found while using a Gaussian kernel function with a parameter of 0.6 and a slack coefficient of 1.0. In addition to offering captivating insights from the results of this study regarding the mechanism by which financial reports impact prices in the bond market, our results also deepen our understanding of excess returns in asset management.

Pages: 13 to 19

Copyright: Copyright (c) IARIA, 2017

Publication date: March 19, 2017

Published in: conference

ISSN: 2308-4375

ISBN: 978-1-61208-542-5

Location: Nice, France

Dates: from March 19, 2017 to March 23, 2017