{"id":12397,"date":"2023-02-09T09:29:49","date_gmt":"2023-02-09T09:29:49","guid":{"rendered":"https:\/\/www.rubyx.io\/?p=12397"},"modified":"2025-03-10T10:52:29","modified_gmt":"2025-03-10T10:52:29","slug":"is-this-the-end-of-expert-scorecards","status":"publish","type":"post","link":"https:\/\/www.rubyx.xyz\/fr\/inspiration\/is-this-the-end-of-expert-scorecards\/","title":{"rendered":"Est-ce la fin des fiches d'\u00e9valuation des experts ?"},"content":{"rendered":"

Confiance, facilit\u00e9, transparence... et probl\u00e8mes<\/strong><\/h3>\n\n\n\n

Les fiches de notation aident traditionnellement les institutions financi\u00e8res \u00e0 \u00e9valuer la solvabilit\u00e9 d'un demandeur de pr\u00eat et \u00e0 orienter la d\u00e9cision de pr\u00eat. Pour en cr\u00e9er une, des experts connaissant bien les activit\u00e9s d'une institution aident \u00e0 s\u00e9lectionner et \u00e0 pond\u00e9rer une s\u00e9rie de variables afin d'\u00e9valuer le risque de cr\u00e9dit d'une personne. Par exemple, une institution de microfinance proposant des pr\u00eats aux petites entreprises peut inclure le profil sociod\u00e9mographique, l'historique des paiements, l'activit\u00e9 commerciale, les garanties et d'autres activit\u00e9s financi\u00e8res telles que l'\u00e9pargne et l'assurance.<\/p>\n\n\n\n

Le r\u00e9sultat est un syst\u00e8me de confiance<\/strong> Il s'agit d'une \"carte de pointage expert\" dont la mise en \u0153uvre ne n\u00e9cessite que tr\u00e8s peu d'expertise technique. La plus simple peut \u00eatre calcul\u00e9e \u00e0 la main, ce qui permet \u00e0 l'institution de d\u00e9terminer rapidement si le demandeur atteint le seuil n\u00e9cessaire pour obtenir un pr\u00eat. En raison de cette simplicit\u00e9, les cartes de score expertes peuvent \u00eatre faciles \u00e0 comprendre<\/strong> pour les clients et les r\u00e9gulateurs, et m\u00eame - si l'institution le souhaite - \u00eatre totalement transparentes<\/strong>.<\/p>\n\n\n\n

Aussi simples et directs qu'ils soient, les tableaux de bord des experts sont de plus en plus consid\u00e9r\u00e9s comme probl\u00e9matiques<\/strong>Elles souffrent de la partialit\u00e9 des personnes qui les cr\u00e9ent, peuvent exposer les institutions financi\u00e8res \u00e0 davantage de risques et m\u00eame compromettre leur mission d'inclusion financi\u00e8re. Voyons pourquoi.<\/p>\n\n\n\n

Partialit\u00e9s dans l'\u00e9laboration des cartes de pointage des experts<\/strong><\/h3>\n\n\n\n

Les d\u00e9fis pos\u00e9s par les tableaux de bord d'experts commencent par leur construction. Qu'il s'agisse du choix de l'\u00e9chantillon pour cr\u00e9er l'algorithme, ou des variables s\u00e9lectionn\u00e9es et du poids qui leur est accord\u00e9, il existe un certain nombre de fa\u00e7ons dont les les pr\u00e9jug\u00e9s de l'homme peuvent compromettre leur efficacit\u00e9 en tant qu'outil d'\u00e9valuation.<\/p>\n\n\n\n

s\u00e9lection de variables biais\u00e9es<\/strong>
Pour qu'une carte de cr\u00e9dit soit efficace, le score de cr\u00e9dit final doit \u00eatre objectivement corr\u00e9l\u00e9 avec les comportements de remboursement futurs. Cependant, les variables s\u00e9lectionn\u00e9es par des experts individuels sur la base de leur exp\u00e9rience peuvent ne pas \u00eatre celles qui ont le plus grand pouvoir pr\u00e9dictif pour l'ensemble de la client\u00e8le. Leur inclusion dans la carte de score peut aboutir \u00e0 un syst\u00e8me fortement biais\u00e9 par l'exp\u00e9rience subjective d'un ou de quelques experts. D'apr\u00e8s notre exp\u00e9rience, il n'est pas rare que les variables s\u00e9lectionn\u00e9es dans les cartes de pointage ne soient pas corr\u00e9l\u00e9es - ou pire, qu'elles soient inversement corr\u00e9l\u00e9es - avec les autres variables de la carte de pointage.<\/p>\n\n\n\n

Partialit\u00e9 de la pond\u00e9ration<\/strong>
Ce ne sont pas seulement les variables s\u00e9lectionn\u00e9es qui peuvent poser probl\u00e8me, mais aussi la mani\u00e8re dont elles sont pond\u00e9r\u00e9es. Par exemple, lorsqu'un \u00e9tablissement de cr\u00e9dit choisit de donner plus de poids \u00e0 des variables sociod\u00e9mographiques pr\u00e9sentant une forte colin\u00e9arit\u00e9 (c'est-\u00e0-dire une corr\u00e9lation), la carte de score peut produire des r\u00e9sultats aberrants. Nous avons parfois observ\u00e9 cette situation : Le client A, avec un excellent historique de remboursement et d'\u00e9pargne, est not\u00e9 comme non \u00e9ligible ; le client B, avec un profil similaire, une couverture de garantie moindre et plusieurs arri\u00e9r\u00e9s de cr\u00e9dit, est not\u00e9 comme \u00e9ligible - simplement en raison du poids des variables sociod\u00e9mographiques.<\/p>\n\n\n\n

Sample selection bias<\/strong>
Another important source of error that is often overlooked is sample selection bias. This occurs when the default data on which the algorithm is based comes from, for example, customers who already have a credit history and who have therefore already been selected beforehand by a credit committee.<\/p>\n\n\n\n

This can produce a vicious circle: the algorithm continues to favour the same type of customers already selected by the credit committee, while it continues to exclude customers similar to those not selected, even though they haven\u2019t had the chance to prove themselves yet. Several studies have also shown excluding higher risk customers from the data sample results in less accurate predictions for all applicants.<\/p>\n\n\n\n

Inflexibility leads to unfairness<\/strong><\/h3>\n\n\n\n

In addition to sample bias in construction, the inflexible construction of expert scorecards and the credit scoring system can result in other kinds of unfairness for customers.<\/p>\n\n\n\n

By their very nature, scorecards are inflexible. They\u2019re a static representation of a person\u2019s financial situation at a given moment. While that situation can change \u2013 even rapidly \u2013 the person\u2019s score does not. And people are rarely given the opportunity or incentive to positively influence the variables that determine their score. For instance, a positive change of repayment behaviour on the customer\u2019s part will not necessarily be rewarded by a better score.<\/p>\n\n\n\n

Additionally, credit thresholds determined by expert scorecards may fail to take into consideration the size of the request, with applicants classified as risky or not regardless of how much they want to borrow. However, an applicant\u2019s ability to repay is never absolute, but relative to the amount requested: a customer judged a high risk for a standard small business loan may still be perfectly able to repay a nano loan<\/a>.<\/p>\n\n\n\n

These inflexibilities and oversights can make the scorecard result something of a sentence \u2013 once ineligible, always ineligible. They can unfairly reinforce exclusion, and lead to frustration on the part of the customer and ultimately affect their relationship with the bank.<\/p>\n\n\n\n

Unnecessarily complex and unscalable<\/strong><\/h3>\n\n\n\n

Finally, despite the potential for simplicity described above, the scorecard\u2019s actual simplicity is conditioned by the number and complexity of its variables. It\u2019s not uncommon to find scorecard models built with dozens of variables with multiple thresholds, resulting in extreme complexity for the scorer.<\/p>\n\n\n\n

For example, if a scorecard has 12 variables, each with a different threshold (e.g. for age: 1 point if the customer is less than 20 years old, 5 points if they\u2019re 20-30 years old, 10 points if they\u2019re 30+, and so on) it can result in thousands of possible configurations. Each one of those configurations leads to a unique decision \u2013 and the sheer number of decisions necessary increases the likelihood of error.<\/p>\n\n\n\n

To make matters worse, variables are not always clearly defined and sometimes require additional data collection, which results in wasted time<\/strong> for loan officers.<\/p>\n\n\n\n

This means that expert scorecards can be a challenge to successfully use, scale up and automate, making it difficult to handle large numbers of applicants in a timely manner.<\/p>\n\n\n\n

The end of the expert scorecard?<\/strong><\/h3>\n\n\n\n

Despite the perceived trust, ease and transparency of expert scorecards, this approach has many limitations which can cost the financial institution in wasted time, missed opportunities, and poor risk management<\/a>. Customer inclusion and satisfaction can also be severely compromised because of construction bias and a customer\u2019s inability to improve their score when their financial situation changes, or through good behaviour.<\/p>\n\n\n\n

Thankfully, however, we\u2019re now in a much better position to overcome the limitations of expert scorecards. Recent developments in behavioural and data science have seen the emergence of a range of data-intensive and AI solutions that can help financial institutions improve their credit scoring \u2013 and help them say goodbye to the limitations of the expert scorecard.<\/p>","protected":false},"excerpt":{"rendered":"

Les fiches d'experts sont l'une des approches d'\u00e9valuation du cr\u00e9dit les plus largement adopt\u00e9es, en particulier par les institutions de microfinance actives dans les march\u00e9s \u00e9mergents. Mais avec les r\u00e9centes avanc\u00e9es analytiques des algorithmes d'\u00e9valuation du cr\u00e9dit, les cartes de score expertes sont-elles toujours une approche valable pour d\u00e9terminer la solvabilit\u00e9 ?<\/p>","protected":false},"author":8,"featured_media":13345,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"om_disable_all_campaigns":false,"inline_featured_image":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":""},"categories":[5],"tags":[],"class_list":["post-12397","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-inspiration"],"aioseo_notices":[],"_links":{"self":[{"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/posts\/12397","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/comments?post=12397"}],"version-history":[{"count":1,"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/posts\/12397\/revisions"}],"predecessor-version":[{"id":13378,"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/posts\/12397\/revisions\/13378"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/media\/13345"}],"wp:attachment":[{"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/media?parent=12397"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/categories?post=12397"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.rubyx.xyz\/fr\/wp-json\/wp\/v2\/tags?post=12397"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}