Showing papers on "Credit risk published in 2022"

Quantile Connectedness: Modeling Tail Behavior in the Topology of Financial Networks

Abstract: We develop a new technique to estimate vector autoregressions with a common factor error structure by quantile regression. We apply our technique to study credit risk spillovers among a group of 17 sovereigns and their respective financial sectors between January 2006 and December 2017. We show that idiosyncratic credit risk shocks propagate much more strongly in both tails than at the conditional mean or median. Furthermore, we develop a measure of the relative spillover intensity in the right and left tails of the conditional distribution that provides a timely aggregate measure of systemic financial fragility and that can be used for risk management and monitoring purposes. This paper was accepted by Gustavo Manso, finance.

The impact of green lending on credit risk: evidence from UAE’s banks

Abstract: This study investigates the impact of UAE's Green Credit Policy on the non-performing loan. One of the main pillars in the UAE green agenda 2015-2030 is the green finance that has been growing in high acceleration in the Gulf Cooperation Council (GCC) countries and the whole world. Consequently, the main objective of this study is to investigate in the financial risks that associated with green lending and whether an increasing in green lending will decrease the non-performing loans ratio (NPLR) of UAE banks, based on the period 2015-2020 dataset of 23 UAE's banks. To achieve this objective, we have used a regression technique that includes a two-stage least square regression analysis and random-effect regression analysis to test if the increase in green credit ratio can reduce the NPL ratio in a sample of UAE's banks. The current study can be considered the first empirical attempt that conducted on the banking sector in UAE, to discover the variables that might have a direct impact on the NPL ratio. The results reveal that the ratio of green loans has a negative impact on the NPL ratio, as much as the return of equity, while the quality of credit, inefficiency, and the bank size have a positive impact on NPL ratio. But as was not as expected, we found that the impact of solvency ratio has a negative significant on the NPL ratio. Finally, the current study introduces a new value to the current literature about the impact of green lending policies and provides a new perspective which supports the financial sustainability in UAE.

Bank green lending and credit risk: an empirical analysis of China's Green Credit Policy

Abstract: This study empirically investigates the relationship between Chinese banks’ green lending and their credit risk, and how China’s green finance regulations contribute to the solvency of individual banks and the resilience of the financial system as a whole. Using a sample of 41 Chinese banks for the period 2007-2018, we find that the association between a bank’s (relative) green lending as a proportion of its overall loan portfolio and its credit risk depends critically on the size and structure of state ownership. While the implementation of China’s Green Credit Policy reduces credit risk for the major state-controlled banks, it increases credit risk for city and regional commercial banks. This performance difference is largely due to information and expertise asymmetries, with city and regional commercial banks having less access to information and expertise necessary to evaluate the credit risk of green lending. Understanding this phenomenon can help policymakers tailor green finance policies according to banks’ characteristics. It also suggests that mechanisms and platforms for city/regional commercial banks to learn from major state-controlled banks could be beneficial.

A two-stage hybrid credit risk prediction model based on XGBoost and graph-based deep neural network

Abstract: The credit risk prediction technique is an indispensable financial tool for measuring the default probability of credit applicants. With the rapid development of machine learning and the application of big data, increasingly sophisticated models have been designed to construct effective credit risk prediction models. In this study, we propose a two-stage hybrid model to enhance the prediction performance of credit risk. First, to make full use of the classified information hidden in credit data, we employ XGBoost to linearize and transform the original features into a high-dimensional sparse feature matrix. Second, to effectively process the transformed high-dimensional data and to discover the relationships between the features, a recently proposed graph-based neural network (forgeNet) model, which is good at addressing high-dimensional data, is deployed to predict the credit risk. The real-world credit data of the Lending Club for the period from 2007 to 2016 were collected and partitioned based on the economic cycle to validate the robustness of the proposed model. The experimental results show that feature transformation and feature graph mining are two pragmatic processes for credit risk prediction when analyzing credit data. Furthermore, the proposed model is robust against different economic cycles and achieves the best average prediction results of 87.52%, 93.13% and 85.59% in terms of accuracy, F1-score, and G-mean compared with other benchmarks, including individuals, hybrid models, and ensembles. The average performance of the proposed model rose by 6.14, 7.59 and 6.18 percentage points, respectively, which demonstrates the outperformance of the proposed two-stage model in credit risk prediction applications

OUP accepted manuscript

Abstract: Abstract Accurate credit ratings are important for both investors and regulators. We demonstrate that the market for credit risk provides an important source of discipline for credit rating agencies (CRAs). We examine a model in which a CRA’s rating is followed by a market for credit risk that provides a public signal – the price. More informative trading increases the CRA’s incentives to be accurate by making rating errors more transparent. We show that this source of discipline is (a) robust to moral hazard, multiple CRAs, and connected primary and secondary markets and (b) specific to the market for credit risk.

Credit scoring based on tree-enhanced gradient boosting decision trees

Abstract: Credit scoring is an important tool for banks and lending companies to realize credit risk exposure management and gain profits. GBDTs, a group of boosting-type ensemble algorithms, have shown promising improvement for credit scoring. However, GBDT improves the credit scoring performance by iteratively modifying only the fitting target for each base classifier and invariably works on the same features, which limits the diversity of individual classifiers in GBDT; Moreover, the performance-interpretability dilemma motivated a large number of works to focus on the pursuit of high-performance ensemble strategies, which leads to the lack of explorations on the interpretability of the credit scoring models. Based on the above limitations, two tree-based augmented GBDTs (AugBoost-RFS and AugBoost-RFU) are proposed in this work for credit scoring. In the proposed methods, a step-wise feature augmentation mechanism is introduced for GBDT to enrich the diversity of individual base classifiers; Tree-based embedding technologies simplify the process of feature augmentation and inherit interpretability of GBDT. Results on 4 large-scale credit scoring datasets show AugBoost-RFS/AugBoost-RFU outperforms GBDT; Besides, supervised tree-based step-wise feature augmentation for GBDT achieves comparable results to neural network-based step-wise feature augmentation while significantly improve the augmentation efficiency. Moreover, the intrinsic global interpreted results and decision rules of tree-enhanced GBDTs, as well as the marginal contributions of features that are visualized by TreeSHAP demonstrate AugBoost-RFS/AugBoost-RFU can be good candidates for interpretable credit scoring.

Assessing credit risk of commercial customers using hybrid machine learning algorithms

Abstract: Given the large amount of customer data available to financial companies, the use of traditional statistical approaches (e.g., regressions) to predict customers’ credit scores may not provide the best predictive performance. Machine learning (ML) algorithms have been explored in the credit scoring literature to increase predictive power. In this paper, we predict commercial customers’ credit scores using hybrid ML algorithms that combine unsupervised and supervised ML methods. We implement different approaches and compare the performance of the hybrid models to that of individual supervised ML models. We find that hybrid models outperform their individual counterparts in predicting commercial customers’ credit scores. Further, while the existing literature ignores past credit scores, we find that the hybrid models’ predictive performance is higher when these features are included.

Be good to be wise: Environmental, Social, and Governance awareness as a potential credit risk mitigation factor

Abstract: Integrating Environmental, Social, and Governance (ESG) factors into credit risk assessment is the new frontier for credit risk management as regulators and investors increasingly require banks to channel loans to “sustainable” borrowers and ultimately foster sustainable growth. Our findings show that higher ESG awareness is strongly associated with better creditworthiness (proxied by the Altman Z-score). We apply a two-step methodology to 3331 companies from various industries and geographies in the 2000–2016 period which reveals that high ESG awareness scores are strongly and very significantly associated with a reduction in firm credit risk. We check the robustness by using the Probability of Default as a dependent variable and an instrumental variable constructed with a factor analysis. Our results support the appropriateness of the introduction of ESG awareness parameters in the creditworthiness assessment of borrowers.

How Banks Can Leverage Credit Risk Evaluation to Improve Financial Performance

Abstract: The research aims to examine the impact of credit risk evaluation on the financial performance of American and European commercial banks during the period 2017-2021. A set of 37 commercial banks were selected to represent the entire banking industry of those two continents. To measure this relationship, two mathematical models were created. Research has revealed that credit risk evaluation influences the financial performance of the American and European commercial banks as represented by ROE and ROA. The study also concludes that the credit risk evaluation indicators analyzed in this study have a substantial effect on the financial performance of American and European commercial banks. The study suggests banks enhance their credit risk evaluation to generate more profits. It also cites the indicators of non-performing loans or gross loans, provision for facilities loss/net facilities, as well as the leverage ratio as significant in determining credit risk evaluation. Banks must put together strategies that will not only limit the banks’ exposure to credit risk, but also enhance the banks’ performance, as well as competitiveness. Further research can be conducted in developing nations to understand the impact of credit risk evaluation in such economies.

Machine learning-driven credit risk: a systemic review

Abstract: Abstract Credit risk assessment is at the core of modern economies. Traditionally, it is measured by statistical methods and manual auditing. Recent advances in financial artificial intelligence stemmed from a new wave of machine learning (ML)-driven credit risk models that gained tremendous attention from both industry and academia. In this paper, we systematically review a series of major research contributions (76 papers) over the past eight years using statistical, machine learning and deep learning techniques to address the problems of credit risk. Specifically, we propose a novel classification methodology for ML-driven credit risk algorithms and their performance ranking using public datasets. We further discuss the challenges including data imbalance, dataset inconsistency, model transparency, and inadequate utilization of deep learning models. The results of our review show that: 1) most deep learning models outperform classic machine learning and statistical algorithms in credit risk estimation, and 2) ensemble methods provide higher accuracy compared with single models. Finally, we present summary tables in terms of datasets and proposed models.

ESG and corporate credit spreads

Abstract: Purpose The authors investigate the implications of environmental, social and governance (ESG) practices of firms for the pricing of their credit default swaps (CDS). In doing so, the authors compare European and US firms and consider nonlinear and indirect effects. This complements the previous literature focusing on linear and direct effects using bond yields and credit ratings of US firms. Design/methodology/approach For this purpose, the authors apply fixed effects regressions on a comprehensive panel data set of US and European firms. Further, nonlinear and indirect effects are investigated utilizing quantile regressions and a path analysis. Findings The evidence indicates that higher ESG ratings mitigate credit risks of US and European firms from 2007 to 2019. The risk mitigation effect is U-shaped across ESG quantiles, which is consistent with opposing effects of growing stakeholder influence capacity and diminishing marginal returns on ESG investments. The authors further reveal a mediating indirect volatility channel that substantially amplifies the direct effect of ESG on credit risk. A one-standard-deviation improvement in ESG ratings is estimated to reduce CDS spreads of low, medium and high ESG firms by approximately 4%, 8% and 3%, respectively. Originality/value This is the first study to examine whether credit markets reflect regional differences between Europe and the US with regard to the ESG-CDS-relationship. In addition, this paper contributes to the existing literature by investigating differences in the response of CDS spreads across ESG quantiles and to study potential indirect channels connecting ESG and CDS spreads using structural credit risk variables.

OUP accepted manuscript

Abstract: Abstract What drives short-term credit spreads: credit risk, liquidity risk, or both? We investigate this issue using the structural approach to credit risk modeling and a novel data set of secondary market transaction prices for Chinese commercial papers (CPs). In particular, we propose and test a structural model with jump risk and exogenous market illiquidity under which the predicted yield spreads can be decomposed into a credit component and a liquidity component. We find that credit risk and, especially liquidity risk, are important determinants of short-term yield spreads. Our model-based decomposition results show that, on average, credit risk and market liquidity account for about 25% and 52% of CP yield spreads, respectively. For comparison, we also examine the drivers of the US CP yield spreads using security-level data. We find that credit risk accounts for a small fraction of the observed yield spreads but liquidity contributes a much greater proportion.

Consumer lending efficiency: commercial banks versus a fintech lender

Abstract: Abstract Fintechs are believed to help expand credit access to underserved consumers without taking on additional risk. We compare the performance efficiency of LendingClub’s unsecured personal loans with similar loans originated by banks. Using stochastic frontier estimation, we decompose the observed nonperforming loan (NPL) ratio into three components: the best-practice minimum NPL ratio, the excess NPL ratio, and a statistical noise, the former two of which reflect the lender’s inherent credit risk and lending inefficiency , respectively. As of 2013 and 2016, we find that the higher NPL ratios at the largest banks are driven by inherent credit risk, rather than lending inefficiency. Smaller banks are less efficient. In addition, as of 2013, LendingClub’s observed NPL ratio and lending efficiency were in line with banks with similar lending volume. However, its lending efficiency improved significantly from 2013 to 2016. As of 2016, LendingClub’s performance resembled the largest banks – consistent with an argument that its increased use of alternative data and AI/ML may have improved its credit risk assessment capacity above and beyond its peers using traditional approaches. Furthermore, we also investigate capital market incentives for lenders to take credit risk. Market value regression using the NPL ratio suggests that market discipline provides incentives to make less risky consumer loans. However, the regression using two decomposed components (inherent credit risk and lending inefficiency) tells a deeper underlying story: market value is significantly positively related to inherent credit risk at most banks, whereas it is significantly negatively related to lending inefficiency at most banks. Market discipline appears to reward exposure to inherent credit risk and punish inefficient lending.

Complex network propagation effect based on SIRS model and research on the necessity of smart city credit system construction

Abstract: In this paper, we first consider one of the interconnected enterprises in the economy as nodes in a complex network. Based on the SIRS risk contagion model, we describe the transmission process of credit risk between enterprises after the bursting of the stock market with the contagion process, and clarify their diffusion mechanism and effect. The propagation effect of SIRS model in complex network is simulated and analyzed. The results show that when the risk contagion intensity exceeds a certain threshold and the proportion of infected enterprises in the economy exceeds a certain level, it will inevitably lead to the insecurity of the whole economy. In most cases, the policy intervention of the regulatory authorities is necessary. If the crisis is allowed to infect, it is likely to induce the financial risk of the whole system; The construction of smart city credit system can use big data to solve the problem of information island, promote the co construction and sharing of data resources, assist the regulators to effectively prevent and block the transmission of credit risk, and nip the risk in the bud, so as to maintain social stability and economic security.

Multi-classification assessment of bank personal credit risk based on multi-source information fusion

Abstract: There have been many studies on machine learning and data mining algorithms to improve the effect of credit risk assessment. However, there are few methods that can meet its universal and efficient characteristics. This paper proposes a new multi-classification assessment model of personal credit risk based on the theory of information fusion (MIFCA) by using six machine learning algorithms. The MIFCA model can simultaneously integrate the advantages of multiple classifiers and reduce the interference of uncertain information. In order to verify the MIFCA model, dataset collected from a real data set of commercial bank in China. Experimental results show that MIFCA model has two outstanding points in various assessment criteria. One is that it has higher accuracy for multi-classification assessment, and the other is that it is suitable for various risk assessments and has universal applicability. In addition, the results of this research can also provide references for banks and other financial institutions to strengthen their risk prevention and control capabilities, improve their credit risk identification capabilities, and avoid financial losses.