{"title":"Retail Credit","authors":"Raymond A. Anderson","doi":"10.1093/oso/9780192844194.003.0003","DOIUrl":"https://doi.org/10.1093/oso/9780192844194.003.0003","url":null,"abstract":"This chapter covers retail credit, which has different data and modelling needs than wholesale. (1) Scorecard terminology—presented is a points-based model (other forms are acknowledged). The goal is to identify rare events, e.g. loan defaults, liquidations, bankruptcies or other undesirable outcomes. (2) Retail models—types across the credit cycle {solicitation, origination, collection, recovery, fraud}, what is being measured {risk, response, retention, revenue}, whose data is used {bespoke, generic, pooled, borrowed} and how it is done {empirical, hybrid, expert judgment}. (3) Data sources—focus is on credit bureaux and credit registries, their spread across various countries, ownership types of credit bureaux and some behind their establishment and spread. (4) Risk indicators—presentation of scores to end-users or downstream processes, as distinct from risk grades. (5) FICO scores—provided by major credit bureaux, with details of different versions and types, plus an imperfect formula for converting their scores into probabilities.","PeriodicalId":286194,"journal":{"name":"Credit Intelligence & Modelling","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125345054","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}
{"title":"Segmentation","authors":"Raymond A. Anderson","doi":"10.1093/oso/9780192844194.003.0022","DOIUrl":"https://doi.org/10.1093/oso/9780192844194.003.0022","url":null,"abstract":"Segmentation identifies subgroups better served if treated separately, especially for risk-heterogeneous populations. Trade-offs occur between the resulting extra lift and the extra costs and complexities. It provides little where risk-homogeneity is enforced by strong filtering mechanisms. (1) Overview—i) drivers—operational, strategic, feedstock or interactional; ii) inhibitors—limits on the number of segments {insufficient data, costs of development, implementation, monitoring}; iii) mitigators—steps to reduce model count {interaction characteristics, alternative transformation and development methodologies}. (2) Analysis—i) learning types—supervised and unsupervised; ii) finding interactions—how to measure interactions for binary targets; iii) segment mining—comparing multiple options; iv) boundary analysis—assessing the impact for cases that switch segments. (3) Presentation—tabular and graphic means of presenting comparisons of different options, especially against having a single model. It includes performance within and across segments, drill-downs into segments and strategy curves showing differences in Accept and Bad rates.","PeriodicalId":286194,"journal":{"name":"Credit Intelligence & Modelling","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134400982","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}
{"title":"Stats & Maths & Unicorns","authors":"Raymond A. Anderson","doi":"10.1093/oso/9780192844194.003.0011","DOIUrl":"https://doi.org/10.1093/oso/9780192844194.003.0011","url":null,"abstract":"This chapter covers basic statistical concepts. Most statistics relate to hypothesis testing, and others to variable selection and model fitting. The name is because an exact match between a theoretical and empirical distribution is as rare as a unicorn. (1) Dispersion—measures of random variations—variance and its inflation factor, covariance and correlations {Pearson’s product-moment, Spearman’s rank order}, and the Mahalanobis distance. (2) Goodness-of-fit—do observations match expectations? This applies to both continuous dependent variables {R-squared and adjusted R2} and categorical {Pearson’s chi-square, Hosmer–Lemeshow statistic}. (3) Likelihood—assesses estimates’ goodness-of-fit to binary dependent variables {log-likelihood, deviance}, plus the Akaike and Bayesian information criteria used to penalize complexity. (4) The Holy Trinity of Statistics—i) Neyman–Pearson’s ‘likelihood ratio’—the basis for model comparisons; ii) Wald’s chi-square—for potential variable removal; iii) Rao’s score chi-square—for potential variable inclusion. These are all used in Logistic Regression.","PeriodicalId":286194,"journal":{"name":"Credit Intelligence & Modelling","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116729730","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}
{"title":"Reject-Inference","authors":"Raymond A. Anderson","doi":"10.1093/oso/9780192844194.003.0023","DOIUrl":"https://doi.org/10.1093/oso/9780192844194.003.0023","url":null,"abstract":"Rejects had not the opportunity to perform. Marginal Rejects are often cherry-picked based upon other data, or cheapened through down-sells, which distorts an Accepts-only model. Reject inference addresses resultant distortions but is contentious. (1) The basics—i) pointers—basic considerations; ii) missing at random, or not; iii) terminology—data manipulation, allocation, methodology; iv) characteristic analysis—for reject inference; v) swap-set analysis—proposed versus past; v) population flow diagram. (2) Intermediate models—especially ‘known Good/Bad’, which may use bureaux’s performance data. Others are Accept/Reject and Cashed/Uncashed. Possible formulae are provided for extrapolated performance assignments. (3) Inference smorgasbord—i) supplementation; ii) performance surrogates; iii) reject is Bad; iv) augmentation; v) weight of evidence (WoE) adjustments; vi) iterative reclassification; vii) extrapolation of accept performance. (4) Favoured technique—involving i) fuzzy-parcelling—record cloning and weight adjustments; ii) extrapolation—graphical setting of performance-adjustment parameters; iii) attribute-level adjustments—where needed; v) practicalities—variable names and coding, with an example.","PeriodicalId":286194,"journal":{"name":"Credit Intelligence & Modelling","volume":"99 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125046179","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}
{"title":"The Birth of Modern Credit Intelligence","authors":"Raymond A. Anderson","doi":"10.1093/oso/9780192844194.003.0007","DOIUrl":"https://doi.org/10.1093/oso/9780192844194.003.0007","url":null,"abstract":"The history of external credit-intelligence players, with many accused of spying. (1)Pre-revolution—collaborative efforts of close-knit communities or business groups. (2)United Kingdom—guardian/trade protection societies, established as member co-operatives. (3)United States—for-profit ‘mercantile agencies, most founded by wholesalers capitalizing on their credit dossiers. Consumer credit bureaux developed separately, as did ‘credit men’ employed by companies. (4)The Big Three credit bureaux—Equifax, Experian and TransUnion, but also Centrale Rischi Finanziari (CRIF) and CreditInfo. Further details are provided regarding their geographic representation, and economic factors. (5)Rating agencies—Moody’s, Standard and Poors (S&P) and Fitch Ratings, which focus on wholesale credit, but some services overlap with small and medium enterprise (SME) credit. (6)High-level observations—i) segments served (consumer, micro, small- and medium-sized enterprises (MSME) and wholesale); ii) economies of scale—drove establishment, expansions, and mergers and acquisitions (M&A) activity; iii) communications—especially print publishing; iv) geographical spread—limitations; v) product mix—banks with deep data and experience are less reliant on external data.","PeriodicalId":286194,"journal":{"name":"Credit Intelligence & Modelling","volume":"46 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126310250","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}
{"title":"Scaling and Banding","authors":"Raymond A. Anderson","doi":"10.1093/oso/9780192844194.003.0025","DOIUrl":"https://doi.org/10.1093/oso/9780192844194.003.0025","url":null,"abstract":"Next is to convert proxies and coefficients into points that can be tallied into a score, which is needed for communicating model outputs, and implementation. There are two parts. (1) Scaling—i) what it means, why it’s done, and desired qualities; ii) percentages—why not; iii) fixed ranges—specify highest and lowest values; iv) scaling parameters—use of a benchmark score and odds, and fixed points-to-double odds; v) other considerations—presentation to a non-technical audience, and adverse reason codes. (2) Banding—i) zero constraints—neither the number of groups or risk of each; ii) fitted distributions—to match a specified frequency distribution; iii) benchmarked—to given average risk levels {AAA, AA+, AA, … C}; iv) fixed-band boundaries—upper and lower limits for each grade. The last two options are appropriate for Master Rating Scales, which are recommended to ensure consistency of meaning and communication within an organization.","PeriodicalId":286194,"journal":{"name":"Credit Intelligence & Modelling","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130182467","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}
{"title":"Side Histories","authors":"Raymond A. Anderson","doi":"10.1093/oso/9780192844194.003.0005","DOIUrl":"https://doi.org/10.1093/oso/9780192844194.003.0005","url":null,"abstract":"Histories of side-topics providing context, especially economic history and registration/identification. (1) Industrial revolutions—labels assigned by various academics, which are not fixed. These drove demand for credit. (2) Booms and busts, bubbles and bursts—major economic upsets since the 17th century, many of which have followed upon technological changes and irrational exuberance. Many busts led to innovations in credit intelligence. (3) Registration—why and how. Starting with social hierarchies and management of obligations, whether upwards of citizens to government {taxation, military}, or downwards with provision of support and services. It is evidenced in various forms {passports, references, certificates, tokens}. (4) Identification—means used for individuals, desired qualities of which may not be possible simultaneously {specific, immutable, assessable, communicable, interrogable and utility}. Different types can be or are used {visual, oral, disclosed, authenticators, invasive}. Many present ethical and practical issues, varying by the period and technologies employed.","PeriodicalId":286194,"journal":{"name":"Credit Intelligence & Modelling","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122816962","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}
{"title":"Model Training","authors":"Raymond A. Anderson","doi":"10.1093/oso/9780192844194.003.0024","DOIUrl":"https://doi.org/10.1093/oso/9780192844194.003.0024","url":null,"abstract":"The chapter provides an approach and issues for model-training using Logistic Regression. (1) Regression—key model qualities plus i) options and settings, and ii) outputs to be expected/demanded. (2) Variable selection—i) criteria; ii) automation; iii) stepwise review; iv) constraining betas, where coefficients do not make sense; v) stepping by Gini, model pruning. (3) Correlation checks—i) multicollinearity—checks of variance inflation factors; ii) correlations—further checks to guard against the inclusion of highly correlated variables. (4) Blockwise variable selection—treatment in groups: i) variable reduction; ii) staged, or hierarchical regression; iii) embedded, model outputs as predictors; iv) ensemble, using outputs of other models. (5) Multi-model comparisons—Lorenz curves and strategy curves, should choices not be clear. (6) Calibration—i) simple adjustment by a constant; ii) piecewise, varying adjustments by the prediction; iii) score and points—adjusting the final score or constituent points; iv) MAPA, for more complex situations","PeriodicalId":286194,"journal":{"name":"Credit Intelligence & Modelling","volume":"111 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129351327","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}
{"title":"Target Definition","authors":"Raymond A. Anderson","doi":"10.1093/oso/9780192844194.003.0018","DOIUrl":"https://doi.org/10.1093/oso/9780192844194.003.0018","url":null,"abstract":"Predictive models need something to predict. Extracted data is used to set (or confirm) the target definition, if not cast in stone. This is one of the most crucial steps of the process, which may not be straightforward. (1) Overview—i) continuous versus binary targets, and distinctions between probability of default (PD), exposure-at-default (EAD) and loss given default (LGD); ii) definition requirements—relevance, focus, transparency, adequacy and data quality; iii) performance components—automated counters/statuses and manual statuses; iv) code cross-checks—to determine whether statuses are properly understood. (2) Definition strictness—i) status nodes—define treatment {out-of-scope, exclusion, trivial balance, Bad/Indeterminate/Good}; ii) roll-rates—used to define delinquency levels; iii) trivial balances—avoid penalization of minor infractions; iv) closed accounts—possible treatments. (3) Integrity checks—i) consistency—period-on-period changes in distribution; ii) characteristics—ensuring they have the intended influence within the definition; iii) swap-set—assessing new versus old OR alternative definitions.","PeriodicalId":286194,"journal":{"name":"Credit Intelligence & Modelling","volume":"72 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126846174","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
求助内容：
应助结果提醒方式：