EBA guidelines on Interest Rate Risk in the Banking Book (IRRBB) are designed to help EU banks effectively manage their interest rate risk and maintain a stable earnings stream. EBA also requires the credit spread risk from the banking book (CSRBB). Banks can effectively manage their exposure to interest rates and spread risks by implementing a comprehensive IRRBB and CSRBB management framework that includes: regular stress testing, sensitivity analysis, effective hedging strategies, and appropriate governance and risk management structures. Under those frameworks, banks must regularly monitor and report interest rates and credit spread risk metrics.
{"title":"Approaching IRRBB and CSRBB: a case study in line with the EBA approach","authors":"Michail Michoulas, D. I. Akkizidis","doi":"10.47473/2020rmm0122","DOIUrl":"https://doi.org/10.47473/2020rmm0122","url":null,"abstract":"EBA guidelines on Interest Rate Risk in the Banking Book (IRRBB) are designed to help EU banks effectively manage their interest rate risk and maintain a stable earnings stream. EBA also requires the credit spread risk from the banking book (CSRBB). Banks can effectively manage their exposure to interest rates and spread risks by implementing a comprehensive IRRBB and CSRBB management framework that includes: regular stress testing, sensitivity analysis, effective hedging strategies, and appropriate governance and risk management structures. Under those frameworks, banks must regularly monitor and report interest rates and credit spread risk metrics.","PeriodicalId":296057,"journal":{"name":"Risk Management Magazine","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123748959","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Supply Chain Finance is as a portfolio of financing and risk mitigation practices and techniques to optimize the management of the working capital and liquidity invested in supply chain processes and transactions. SCF techniques existing on the market can be divided into three categories: receivable purchase, advanced payable, and loans. These financing solutions are significantly ‘eventdriven’, since they aim at satisfying the financial requirements of buyers and sellers, that are triggered by purchase orders, invoices, receivables, other claims, and related pre-shipment and post-shipment processes along the increasingly complex supply chains in which they are involved. Along the way from raw material procurement to production, sales and end-users, several source of risks can threaten the possibility of completing the transactions and the regular functioning of supply chain finance. Digitization can help in managing these risks, facilitating the control of the factors underlying them.
{"title":"Supply Chain Finance techniques and risks","authors":"F. Querci","doi":"10.47473/2020rmm0120","DOIUrl":"https://doi.org/10.47473/2020rmm0120","url":null,"abstract":"Supply Chain Finance is as a portfolio of financing and risk mitigation practices and techniques to optimize the management of the working capital and liquidity invested in supply chain processes and transactions. SCF techniques existing on the market can be divided into three categories: receivable purchase, advanced payable, and loans. These financing solutions are significantly ‘eventdriven’, since they aim at satisfying the financial requirements of buyers and sellers, that are triggered by purchase orders, invoices, receivables, other claims, and related pre-shipment and post-shipment processes along the increasingly complex supply chains in which they are involved. Along the way from raw material procurement to production, sales and end-users, several source of risks can threaten the possibility of completing the transactions and the regular functioning of supply chain finance. Digitization can help in managing these risks, facilitating the control of the factors underlying them.","PeriodicalId":296057,"journal":{"name":"Risk Management Magazine","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133740935","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
D. Curcio, I. Gianfrancesco, Annalisa Pansini, Alina Preger
This paper contributes to prior literature and to the current debate concerning the prudential supervisory framework to measure interest rate risk in the banking book (IRRBB), which has been significantly changed on April 2016, when the Basel Committee on Banking Supervision (BCBS) published the latest update of its measurement standards. The consultation launched by the European Banking Authority (EBA) on December 2021, aiming at introducing the supervisory outlier test (SOT) on net interest income (NII), presents several issues and policy implications which could influence in the next future banks' asset and liability management strategies, their internal control systems, risk policies and procedures. By analyzing a sample of 28 Italian commercial banks at the end of 2021, representing more than 70% of Italian baking system’s total assets , we observe that the thresholds proposed by the EBA appear very strict and significantly depend on: i) the sample considered, ii) the lower bound applied to interest rates in the downward scenarios and iii) the current level of interest rates term structure. Our results suggest that the proposed values should be considered with caution as it seems that their potential impacts have not been thoroughly assessed. Further analyses are therefore necessary to guarantee greater robustness of the methodology used for the calibration of the thresholds, taking also into account a wider sample of banks and longer time series, as well as the correlation between the two approaches.
{"title":"The new supervisory outlier test (SOT) on net interest income (NII): empirical evidence from a sample of Italian banks","authors":"D. Curcio, I. Gianfrancesco, Annalisa Pansini, Alina Preger","doi":"10.47473/2020rmm0117","DOIUrl":"https://doi.org/10.47473/2020rmm0117","url":null,"abstract":"This paper contributes to prior literature and to the current debate concerning the prudential supervisory framework to measure interest rate risk in the banking book (IRRBB), which has been significantly changed on April 2016, when the Basel Committee on Banking Supervision (BCBS) published the latest update of its measurement standards. The consultation launched by the European Banking Authority (EBA) on December 2021, aiming at introducing the supervisory outlier test (SOT) on net interest income (NII), presents several issues and policy implications which could influence in the next future banks' asset and liability management strategies, their internal control systems, risk policies and procedures. By analyzing a sample of 28 Italian commercial banks at the end of 2021, representing more than 70% of Italian baking system’s total assets , we observe that the thresholds proposed by the EBA appear very strict and significantly depend on: i) the sample considered, ii) the lower bound applied to interest rates in the downward scenarios and iii) the current level of interest rates term structure. Our results suggest that the proposed values should be considered with caution as it seems that their potential impacts have not been thoroughly assessed. Further analyses are therefore necessary to guarantee greater robustness of the methodology used for the calibration of the thresholds, taking also into account a wider sample of banks and longer time series, as well as the correlation between the two approaches.","PeriodicalId":296057,"journal":{"name":"Risk Management Magazine","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132328477","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This work aims to investigate the main problems that impact the pricing models and the sensitivity measures of American options written on shares without a pay-out, in the presence of negative interest rates with a specific focus on the Monte Carlo method. The first paragraph carries out a review of the anomalies caused by such an odd condition and focuses thereafter on the core topic of the research by treating a wide range of numerical models suitable for unbiased evaluation of the early exercise, thus expanding the existing literature. The two following paragraphs are dedicated to describing the models used for the correct estimation of fair value: binomial lattice models (Cox-Ross-Rubinstein - CRR Tree, Leisen Reimer - LR Tree, Jarrow-Rudd - JR Tree and Tian Tree), trinomial stochastic trees, Finite Difference Method (FDM) scheme and the Longstaff-Schwartz Monte Carlo. Particular attention is paid to this last approach which allows to combine the flexibility of traditional numerical integration schemes for stochastic processes on equity with the estimation of the convenience of exercising the American option ahead of time. After conducting quantitative tests both on pricing and on the estimation of sensitivity measures, the LR Tree was selected as the most performing deterministic algorithm to be compared with the Monte Carlo stochastic technique. The final part of the work focuses on quantifying the valuation gap introduced by negative interest rates in the valuation of American options written on an unprofitable underlying comparing the traditional valuation approach and the deterministic Leisen Reimer model and the Longstaff-Schwartz stochastic model.
{"title":"The impact of negative interest rates on the pricing of options written on equity: a technical study for a suitable estimate of early termination","authors":"A. Bottasso, P. Giribone, Lorenzo Bruno","doi":"10.47473/2020rmm0116","DOIUrl":"https://doi.org/10.47473/2020rmm0116","url":null,"abstract":"This work aims to investigate the main problems that impact the pricing models and the sensitivity measures of American options written on shares without a pay-out, in the presence of negative interest rates with a specific focus on the Monte Carlo method. The first paragraph carries out a review of the anomalies caused by such an odd condition and focuses thereafter on the core topic of the research by treating a wide range of numerical models suitable for unbiased evaluation of the early exercise, thus expanding the existing literature. The two following paragraphs are dedicated to describing the models used for the correct estimation of fair value: binomial lattice models (Cox-Ross-Rubinstein - CRR Tree, Leisen Reimer - LR Tree, Jarrow-Rudd - JR Tree and Tian Tree), trinomial stochastic trees, Finite Difference Method (FDM) scheme and the Longstaff-Schwartz Monte Carlo. Particular attention is paid to this last approach which allows to combine the flexibility of traditional numerical integration schemes for stochastic processes on equity with the estimation of the convenience of exercising the American option ahead of time. After conducting quantitative tests both on pricing and on the estimation of sensitivity measures, the LR Tree was selected as the most performing deterministic algorithm to be compared with the Monte Carlo stochastic technique. The final part of the work focuses on quantifying the valuation gap introduced by negative interest rates in the valuation of American options written on an unprofitable underlying comparing the traditional valuation approach and the deterministic Leisen Reimer model and the Longstaff-Schwartz stochastic model.","PeriodicalId":296057,"journal":{"name":"Risk Management Magazine","volume":"108 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133982001","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Franco Fiordelisi, Carlo Palego, Annalisa Richetto, Giulia Scardozzi
We analyze what are the main pricing components for performing loans. By exploiting a survey conducted by the authors in AIFIRM (2021), we provide empirical evidence about whether and to what extent various pricing criteria are related to interest income within the internal model framework. Our main findings are that banks’ interest income is positively related to the adoption of advanced internal risk-based models, the calculation of the break-even rate, and the implementation of the risk-adjusted profitability measures in the pricing, while it is negatively linked to higher market competition, a decentralized pricing function (allowing more customeroriented loans prices). The results make urgent to monitor and develop improve current risk models to support both central offices and the sales network in the process of formulating loan prices and monitoring the value consequently created.
{"title":"Risk-Adjusted Loan Pricing","authors":"Franco Fiordelisi, Carlo Palego, Annalisa Richetto, Giulia Scardozzi","doi":"10.47473/2020rmm0115","DOIUrl":"https://doi.org/10.47473/2020rmm0115","url":null,"abstract":"We analyze what are the main pricing components for performing loans. By exploiting a survey conducted by the authors in AIFIRM (2021), we provide empirical evidence about whether and to what extent various pricing criteria are related to interest income within the internal model framework. Our main findings are that banks’ interest income is positively related to the adoption of advanced internal risk-based models, the calculation of the break-even rate, and the implementation of the risk-adjusted profitability measures in the pricing, while it is negatively linked to higher market competition, a decentralized pricing function (allowing more customeroriented loans prices). The results make urgent to monitor and develop improve current risk models to support both central offices and the sales network in the process of formulating loan prices and monitoring the value consequently created.","PeriodicalId":296057,"journal":{"name":"Risk Management Magazine","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128306990","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this work, we propose a hierarchical model to introduce Forward-Looking effects on the Loss Given Default Rate (LGDR) estimate, as required by IFRS9. The Framework consists of two modules: a SURTS satellite model (Seemingly Unrelated Regressions Model Time Series), which analyses the dynamics of the systemic LGSR (bad loans LGDR) and a set of selected macroeconomic factors, and a Beta Inflated-(0,1) model which estimates the LGSR for the single entity. The basic hypotheses for the construction of the hierarchical model will also be illustrated, underlining how this approach is particularly relevant for LSIs (Less Significant Institutions). The theoretical aspects are followed by an application on a series released by the Bank of Italy, presenting the LGDR estimation process on an archive of closed bad loans by a set of banks belonging to the CABEL (ICT Service Provide) network. By way of example, we illustrate the forecast results for the three-year period 2022-2024 for the systemic LGDR. Other aspects related to the construction of LGDR models are addressed, such as the segmentation of the portfolios and the selection of individual attributes. In particular, we introduce the NPL vintage as an explanatory variable in the LGDR model, outlining the interconnections with the effects of macroeconomic projections.
{"title":"Modello LGSR forward looking","authors":"D. Cavallini, Francesco Letizia","doi":"10.47473/2020rmm0118","DOIUrl":"https://doi.org/10.47473/2020rmm0118","url":null,"abstract":"In this work, we propose a hierarchical model to introduce Forward-Looking effects on the Loss Given Default Rate (LGDR) estimate, as required by IFRS9. The Framework consists of two modules: a SURTS satellite model (Seemingly Unrelated Regressions Model Time Series), which analyses the dynamics of the systemic LGSR (bad loans LGDR) and a set of selected macroeconomic factors, and a Beta Inflated-(0,1) model which estimates the LGSR for the single entity. The basic hypotheses for the construction of the hierarchical model will also be illustrated, underlining how this approach is particularly relevant for LSIs (Less Significant Institutions). The theoretical aspects are followed by an application on a series released by the Bank of Italy, presenting the LGDR estimation process on an archive of closed bad loans by a set of banks belonging to the CABEL (ICT Service Provide) network. By way of example, we illustrate the forecast results for the three-year period 2022-2024 for the systemic LGDR. Other aspects related to the construction of LGDR models are addressed, such as the segmentation of the portfolios and the selection of individual attributes. In particular, we introduce the NPL vintage as an explanatory variable in the LGDR model, outlining the interconnections with the effects of macroeconomic projections.","PeriodicalId":296057,"journal":{"name":"Risk Management Magazine","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129767255","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The purpose of this document is to provide an overview of IFRS 17 and its possible practices according to a context analysis conducted by EIOPA and ESMA. Furthermore, the implications of the standard will be evaluated with respect to a traditional life and pension products. This requires a deeper insight into the standard, the construction of fictive insurance product and the determination of measurements techniques. With the recent release of the standard, IFRS 17 is unexplored to many within the insurance industry. Thus, the effects of the standard on the financial statements of insurance companies and strategies for reaching objectives such as profit smoothing are yet unknown. Furthermore, the standard is principle based and does hence not specify a practice. Therefore, to determine a practice that complies with the standard and enables achievements of desired objectives is vital. In conclusion, IFRS 17 is expected to bring substantial benefits to financial stability in the EU, mainly through the transparency channel; the requirements in IFRS 17 may push insurance corporations to improve internal processes, including enhancing their internal risk management frameworks.
{"title":"Implications of IFRS 17 in European financial stability: accounting methodology and evaluation modelling","authors":"Stefano de Nichilo","doi":"10.47473/2020rmm0114","DOIUrl":"https://doi.org/10.47473/2020rmm0114","url":null,"abstract":"The purpose of this document is to provide an overview of IFRS 17 and its possible practices according to a context analysis conducted by EIOPA and ESMA. Furthermore, the implications of the standard will be evaluated with respect to a traditional life and pension products. This requires a deeper insight into the standard, the construction of fictive insurance product and the determination of measurements techniques. With the recent release of the standard, IFRS 17 is unexplored to many within the insurance industry. Thus, the effects of the standard on the financial statements of insurance companies and strategies for reaching objectives such as profit smoothing are yet unknown. Furthermore, the standard is principle based and does hence not specify a practice. Therefore, to determine a practice that complies with the standard and enables achievements of desired objectives is vital. In conclusion, IFRS 17 is expected to bring substantial benefits to financial stability in the EU, mainly through the transparency channel; the requirements in IFRS 17 may push insurance corporations to improve internal processes, including enhancing their internal risk management frameworks.","PeriodicalId":296057,"journal":{"name":"Risk Management Magazine","volume":"90 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116705578","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The estimate of the probability of default plays a central role for any financial entity that wants to have an overview of the risks of insolvency it may incur by having economic relations with counterparties. This study aims to analyze the calculation of such measure in the context of counterparty risk from a current and prospective standpoint, by using dynamic neural networks. The forecasting aspect in the calculation of such risk measure is becoming more and more important over time as current regulation is increasingly based on a "Through the Cycle" and not a "Point in Time" assessment, consequently giving fundamental importance to such estimate. To this end, three different models aimed at calculating the Probability of Default have been investigated: the CDS method, the Z-Spread method, and the KMV method (Kealhofer, Merton and Vasicek). First, the different techniques have been applied to one of the main suppliers of gas and energy in Italy as a reference company. Then, they have been applied to calculate the same risk measure on the 50 companies included in one of the most important European indices, the Euro Stoxx 50.
{"title":"Current and prospective estimate of counterparty risk through dynamic neural networks","authors":"Alessio Agnese, P. Giribone, F. Querci","doi":"10.47473/2020rmm0112","DOIUrl":"https://doi.org/10.47473/2020rmm0112","url":null,"abstract":"The estimate of the probability of default plays a central role for any financial entity that wants to have an overview of the risks of insolvency it may incur by having economic relations with counterparties. This study aims to analyze the calculation of such measure in the context of counterparty risk from a current and prospective standpoint, by using dynamic neural networks. The forecasting aspect in the calculation of such risk measure is becoming more and more important over time as current regulation is increasingly based on a \"Through the Cycle\" and not a \"Point in Time\" assessment, consequently giving fundamental importance to such estimate. To this end, three different models aimed at calculating the Probability of Default have been investigated: the CDS method, the Z-Spread method, and the KMV method (Kealhofer, Merton and Vasicek). First, the different techniques have been applied to one of the main suppliers of gas and energy in Italy as a reference company. Then, they have been applied to calculate the same risk measure on the 50 companies included in one of the most important European indices, the Euro Stoxx 50.","PeriodicalId":296057,"journal":{"name":"Risk Management Magazine","volume":"54 48","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114006036","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The current EU capital regulation requires that banks comply with two main frameworks at the same time: one for prudential purposes, the other for resolution purposes. The first one includes both a risk-weighted requirement (RW) and a leverage ratio requirement (LR). Similarly, the resolution framework, which ensures that banks have enough loss-absorbing and recapitalization capacity through a Minimum Requirement of Eligible Liabilities (MREL), is based on two ratios that are to be met in parallel: the MREL as a percentage of risk weighted assets (MREL-RW) and the MREL as a percentage of the total exposure measure used for the purpose of the leverage ratio (MREL-LR). According to the EU regulation, the CBR is only required on top of the two risk-weighted requirements (RW and MREL-RW).
{"title":"Overlaps between minimum requirements and capital buffers: the usability of the combined buffer requirement for Italian banks","authors":"Wanda Cornacchia, G. Guerra","doi":"10.47473/2020rmm0107","DOIUrl":"https://doi.org/10.47473/2020rmm0107","url":null,"abstract":"The current EU capital regulation requires that banks comply with two main frameworks at the same time: one for prudential purposes, the other for resolution purposes. The first one includes both a risk-weighted requirement (RW) and a leverage ratio requirement (LR). Similarly, the resolution framework, which ensures that banks have enough loss-absorbing and recapitalization capacity through a Minimum Requirement of Eligible Liabilities (MREL), is based on two ratios that are to be met in parallel: the MREL as a percentage of risk weighted assets (MREL-RW) and the MREL as a percentage of the total exposure measure used for the purpose of the leverage ratio (MREL-LR). According to the EU regulation, the CBR is only required on top of the two risk-weighted requirements (RW and MREL-RW).","PeriodicalId":296057,"journal":{"name":"Risk Management Magazine","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114651078","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Paolo Di Biasi, Rita Gnutti, Andrea Resti, D. Vergari
As the financial services landscape witnesses an unprecedented change, banks can use machine learning (“ML”) to expand their databases through alternative sources providing unstructured and semi-structured information, such as transaction data and digital footprint data. However, ML algorithms also suffer from several potential shortcomings, as they may overfit sample data and prove unstable over time, they may quickly become obsolete and need re-estimation, and they may prove hard to interpret. This paper joins the debate on ML in banks by providing three case studies that highlight the benefits of machine learning, while showing how its drawbacks can be minimised: a rating model developed within the IRB framework, a challenger model used to validate a bank’s main model for retail PDs, and an early warning system based on transaction data.
{"title":"Machine Learning for Credit risk: three successful Case Histories","authors":"Paolo Di Biasi, Rita Gnutti, Andrea Resti, D. Vergari","doi":"10.47473/2020rmm0108","DOIUrl":"https://doi.org/10.47473/2020rmm0108","url":null,"abstract":"As the financial services landscape witnesses an unprecedented change, banks can use machine learning (“ML”) to expand their databases through alternative sources providing unstructured and semi-structured information, such as transaction data and digital footprint data. However, ML algorithms also suffer from several potential shortcomings, as they may overfit sample data and prove unstable over time, they may quickly become obsolete and need re-estimation, and they may prove hard to interpret. This paper joins the debate on ML in banks by providing three case studies that highlight the benefits of machine learning, while showing how its drawbacks can be minimised: a rating model developed within the IRB framework, a challenger model used to validate a bank’s main model for retail PDs, and an early warning system based on transaction data.","PeriodicalId":296057,"journal":{"name":"Risk Management Magazine","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122888241","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：