Abstract
This paper analyzes the influence of text information on credit markets in Japan, focusing on headline news, a source of information that has immediate influence on the money market and also which is regarded as an important source of information when making investment decisions. In this research, we employ an automatic text classification algorithm in order to classify the headline news into several categories. As a result of intensive analysis, we made the following findings (Antweiler W, Frank MZ J Financ 59:1259–1294, 2004): it is possible to build a headline news algorithm to an accuracy of 80% (Ben-Saud T Adopting a liability-led strategy. Pension management, April, pp 34–35, 2005); headline news has an influence on corporate bond spreads in Japan after items of news become public (Black F, Cox J J Financ, 31:351–367, 1976). The reaction of CDS spread is different from that of corporate bond spread even though both spreads relate to credit risk. These results are suggestive from both academic and practical viewpoints.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
For example, liability-driven investments (LDI) have drawn attention in the practical pension fund business (Ben-Saud 2005).
- 2.
Default risk is the key factor in their values and variations.
- 3.
In this analysis, first, we estimate JGB yield curve and then apply it to establish estimation of spreads.
- 4.
We put these conditions for the following reasons (Antweiler and Frank 2004): corporate bonds with maturity from 3 years to 5 years have the largest in number (Ben-Saud 2005), and it is necessary to analyze bonds with similar maturity in order to get rid of the influence of maturity in analyses, because changes in spreads are different according to maturity.
- 5.
Itraxx Japan produced the spread data; we use it for analyses as it is.
- 6.
Similar tendencies are observed in analyses of the stock market. The reaction of the CDS market is similar to that of the stock market.
- 7.
Further research on companies with high credit rating is planned for the future.
- 8.
As for stock, we calculate excess return by subtracting average return from stock return of the target company in order to remove the influence of market movement as a whole.
- 9.
As for corporate bond and CDS, an excess return is calculated as follows: duration × spread change. We set the duration of corporate bonds at 4 years and the duration of CDS at 5 years which reflects the actual mean duration of each market.
References
Antweiler W, Frank MZ (2004) Is all that talk just noise? The information content of internet stock message boards. J Financ 59:1259–1294
Ben-Saud T (2005) Adopting a liability-led strategy. Pension management, April, pp 34–35
Black F, Cox J (1976) Valuing corporate securities: some effects of bond indenture provision. J Financ 31:351–367
Crosbie PJ, Galai D, Mark R (2000) A comparative analysis of current credit risk models. J Bank Financ 24:59–117
Lewis DD (1998) Naïve (Bayes) at forty: the independence assumption in information retrieval. In: Proceedings of the 10th European Conference on Machine Learning (ECML). Springer-Verlag, New York, pp 4–15
Longstaff FE, Schwartz ES (1995) Valuing risky debt: a new approach. J Financ 50:789–821
Merton RC (1974) On the pricing of corporate debt: the structure of interest rates. J Financ 29:449–470
Mitchell TM (1997) Machine learning (ECML). McGraw-Hill, New York
Nigam K, McCallum AK, Thrum S, Mitchell T (2000) Text classification from labeled and unlabeled documents using EM. Mach Learn 39:103–134
Salton G, McGill MJ (1983) Introduction to modern information retrieval. McGraw-Hill, New York
Salton G, Yang CC (1973) On the specification of term values in automatic indexing. J Doc 29:351–372
Sparck JK (1972) Statistical interpretation of term specificity and its application retrieval. J Doc 28:11–21
Takahashi S, Takahashi H, Tsuda K (2004) An efficient learning system for knowledge of asset management. In: Negoita M, Howlett R, Jain L (eds) Lecture notes in computer science 3213(knowledge-based intelligent information and engineering systems). Springer, Berlin/New York, pp 509–515
Takahashi S, Takahashi M, Takahashi H, Tsuda K (2006) Analysis of the validity of textual data in stock market through text mining. WSEAS Trans Bus Econ 3:310–315
Takahashi S, Takahashi H, Tsuda K (2009) Analysis of the effect of headline news in financial market through text categorization. Int J Comput Appl Technol 35:204–209
Tetlock PC (2007) Giving content to investor sentiment: the role of Media in the Stock Market. J Financ 68:1139–1168
Tumarkin R, Tobert FW (2001) News or noise? Internet message board activity and stock prices. Financ Anal J 57:41–51
Wuthrich B, Cho V, Leung S, Perrmunetilleke D, Sankaran K, Zhang J, Lam W (1998) Daily prediction of major stock indices from textual www data. KDDM’98 conference. AAAI Press, New York, pp 364–368
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer Nature Singapore Pte Ltd.
About this chapter
Cite this chapter
Jotaki, H., Yamashita, Y., Takahashi, S., Takahashi, H. (2018). Analyzing the Influence of Headline News on Credit Markets in Japan. In: Kurahashi, S., Takahashi, H. (eds) Innovative Approaches in Agent-Based Modelling and Business Intelligence. Agent-Based Social Systems, vol 12. Springer, Singapore. https://doi.org/10.1007/978-981-13-1849-8_7
Download citation
DOI: https://doi.org/10.1007/978-981-13-1849-8_7
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-13-1848-1
Online ISBN: 978-981-13-1849-8
eBook Packages: Economics and FinanceEconomics and Finance (R0)