Monitoring corporate credit risk with multiple data sources
| DOI | https://doi.org/10.1108/IMDS-02-2022-0091 |
| Published date | 16 November 2022 |
| Date | 16 November 2022 |
| Pages | 434-450 |
| Subject Matter | Information & knowledge management,Information systems,Data management systems,Knowledge management,Knowledge sharing,Management science & operations,Supply chain management,Supply chain information systems,Logistics,Quality management/systems |
| Author | Du Ni,Ming K. Lim,Xingzhi Li,Yingchi Qu,Mei Yang |
Monitoring corporate credit risk
with multiple data sources
Du Ni
School of Management, Nanjing University of Posts and Telecommunications,
Nanjing, China
Ming K. Lim
Adam Smith Business School, University of Glasgow, Glasgow, UK
Xingzhi Li
Chongqing Jiaotong University, Chongqing, China, and
Yingchi Qu and Mei Yang
Chongqing University, Chongqing, China
Abstract
Purpose –Monitoring corporate credit risk (CCR) has traditionally relied on such indicators as income, debt
and inventory at a company level. These data are usually released on a quarterly or annual basis by the target
company and include, exclusively, the financial data of the target company. As a result of this exclusiveness,
the models for monitoring credit risk usually fail to account for some significant information from different
sources or channels, like the data of its supply chain partner companies and other closely relevant data yet
available from public networks, and it is these seldom used data that can help unveil the immediate CCR
changes and how the risk is being propagated along the supply chain. This study aims to discuss the a
forementioned issues.
Design/methodology/approach –Going beyond the existing CCR prediction data, this study intends to
address the impact of supply chain data and network activity data on CCR prediction, by integrating machine
learning technology into the prediction to verify whether adding new data can improve the predictability.
Findings –The results show that the predictive errors of the datasets after adding supply chain data and
network activity data to them are made the ever least. Moreover, intelligent algorithms like support vector
machine (SVM), compared to traditionally used methods, are better at processing nonlinear datasets and
mining complex relationships between multi-variable indicators for CCR evaluation.
Originality/value –This study indicates that bringing in more information of multiple data sources
combined with intelligent algorithms can help companies prevent risk spillovers in the supply chain from
causing harm to the company, and, as well, help customers evaluate the creditworthiness of the entity to lessen
the risk of their investment.
Keywords Corporate credit risk, Prediction, Multiple data sources, Machine learning, Support vector machine
Paper type Research paper
1. Introduction
One major characteristic of corporate credit risk (CCR) is that investors must bear certain
risks if chipping in financial securities or purchasing bonds of some business companies. If a
bond issuer fails to repay the principal and interest when due, the investors will suffer the
losses (Bazarbash, 2019;Zhang et al., 2022). Then CCR assessment is applied, in a way, to help
ensure the worthiness of the purchase or cushion the investors from the losses. However,
several studies suggest that the CCR prediction is greatly affected by the evaluation
standards (Fracassi et al., 2016;Wang et al., 2020), and It is not easy to include all factors
leading to corporate default by a single evaluation method (Li et al., 2020). Therefore, when
IMDS
123,2
434
Funding: This work was supported by 2022 Scientific Research Startup Fund of Chongqing Jiaotong
University [Grant No. F1210045] and the graduate research and innovation foundation of Chongqing,
China [Grant No. CYS21047]. The authors declare no competing interests.
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/0263-5577.htm
Received 20 February 2022
Revised 22 May 2022
2 August 2022
Accepted 3 September 2022
Industrial Management & Data
Systems
Vol. 123 No. 2, 2023
pp. 434-450
© Emerald Publishing Limited
0263-5577
DOI 10.1108/IMDS-02-2022-0091
predicting CCR, it is necessarily better to consider the effect of its evaluation and try a broader
coverage of predictive data.
In general, the traditional indicators for CCR prediction mainly come from the target
company’s financial data (Bonsall et al., 2017;Wu and Brynjolfsson, 2015), which are usually
released by the target company on a quarterly or annual basis. However, these traditional
indicators have certain intrinsic deficiencies; for instance, some companies’financial data are
not released regularly (Grover et al., 2018;Wu et al., 2022). Even worse, their financial reports
are easily manipulated that there are often seen overstating assets, improper disclosures and
immoral financial frauds, yet it is impossible to testify these manipulations as easily (Lev,
2018). Luckily, the deficiencies find their remedies in the development of artificial intelligence
and communication technology which has rendered a more transparent financial and
information flow between companies and institutions (AlShamsi et al., 2021;Huang and Rusk,
2021). The CCR can be revealed sooner by a careful study of some data available on public
networks before the target company suffers the financial losses, and even before the target
company realizes a credit risk may occur (Moat et al., 2016). Besides, some researchers have
found that credit risks are able to spread along the supply chain networks and adversely
affect the target companies through the supply chain (Agca et al., 2021), and therefore, in
considering the company’s own financial situation, the data expansion by including data of
network activities along the supply chains are proposed to be employed to predict the CCR in
a more timely and more precise way.
This study will first review the previous studies on CCR prediction in Section 2. Based on
the gaps explored in the literature review, Section 3 is to present the methods concerning data
acquisition, compilation and processing for the CCR model analysis deployed by this study.
Section 4 is for the results of the data analysis and an elaborate discussion about them.
Finally, the conclusions of this study are presented in Section 5.
2. Literature review
Traditional indicators for CCR prediction are mainly collected from data of a target company
like cash flow (Wu and Brynjolfsson, 2015), financial proportion heterogeneity (Niemann
et al., 2008), debt cost (Mansi et al., 2012), corporate governance (Bonsall et al., 2017), to name
some of them. Generally, these data are released by a single company either quarterly or
annually. This exclusiveness, to some extent, has tarnished the credibility of the data, and an
increase in financial frauds over the years has furthered undermined the credibility of
publicized financial data The trick is that the researchers for CCR prediction cannot verify
whether the company has exaggerated its financial data or not (Lev, 2018;Zhuang et al., 2021),
and some small and medium-sized enterprises or private companies do not release their
financial data regularly at all, as a result of which CCR institutions obtain no valuable
information for CCR prediction (Grover et al., 2018;Wu et al., 2022), and also as a result of
which the institutions have turned to the other data sources and research methods for a
trustworthy and effective CCR prediction.
In regard of the valuable data sources for CCR prediction, several studies claim that rapid
advances in technologies such as artificial intelligence have made financial and information
flows between companies and the institutions for evaluation more transparent. Non-financial
data of network activities like the negative news (Mart
ınet al., 2021), social media evaluation
(Bazarbash, 2019;Ghasemkhani et al., 2015) and Internet search (Liu, 2020) can also have an
impact on CCR prediction, as such data on the internet can reveal the corresponding
preferences of corporate managers. In other words, there should be a correlation between
online activities and economic performance. By quantifying the search on the internet with
the help of Google and Wikis for financial-related information, the researchers hold that they
can capture the real-time CCR changes ahead of the risk report in its traditional financial
Monitoring
CCR with
multiple data
sources
435
To continue reading
Request your trial