{"id":1879,"date":"2021-10-10T17:23:23","date_gmt":"2021-10-10T17:23:23","guid":{"rendered":"http:\/\/beeeye.com\/?p=1879"},"modified":"2022-01-31T10:21:31","modified_gmt":"2022-01-31T10:21:31","slug":"agile-credit-risk-modeling-platform","status":"publish","type":"post","link":"http:\/\/beeeye.com\/agile-credit-risk-modeling-platform\/","title":{"rendered":"Five must-haves for an agile credit risk modeling platform"},"content":{"rendered":"<div class=\"fusion-fullwidth fullwidth-box fusion-builder-row-1 fusion-flex-container nonhundred-percent-fullwidth non-hundred-percent-height-scrolling\" style=\"background-color: rgba(255,255,255,0);background-position: center center;background-repeat: no-repeat;border-width: 0px 0px 0px 0px;border-color:#eae9e9;border-style:solid;\" ><div class=\"fusion-builder-row fusion-row fusion-flex-align-items-flex-start\" style=\"max-width:105%;margin-left: calc(-5% \/ 2 );margin-right: calc(-5% \/ 2 );\"><div class=\"fusion-layout-column fusion_builder_column fusion-builder-column-0 fusion_builder_column_1_1 1_1 fusion-flex-column\"><div class=\"fusion-column-wrapper fusion-flex-justify-content-flex-start fusion-content-layout-column\" style=\"background-position:left top;background-repeat:no-repeat;-webkit-background-size:cover;-moz-background-size:cover;-o-background-size:cover;background-size:cover;padding: 0px 0px 0px 0px;\"><div class=\"fusion-text fusion-text-1\"><p>When a strong wind gusts, only the trees that can bend won\u2019t break. Your credit risk model is the same &#8211; if it won\u2019t be able to rapidly change, you might as well be looking at a crystal ball. Chances are, your model is not updated yet with the most current data. Why? Because updating most credit risk platforms is a laborious task that takes a very long time. Lots of red tape, regulations, forms and people are involved. This is never optimal &#8211; a model starts to deteriorate the moment you begin using it &#8211; but in this whirlwind of a market this is a real problem for lenders. The percent of errors increase, leading to an increase in default rate and clean rejects. <\/p>\n<h3 data-fontsize=\"31\" data-lineheight=\"35.65px\" class=\"fusion-responsive-typography-calculated\" style=\"--fontSize:31; line-height: 1.15; --minFontSize:31;\">Is your credit modeling platform agile?<\/h3>\n<p>Right now, <a href=\"https:\/\/www.wsj.com\/articles\/u-s-consumer-prices-fell-in-march-11586523372\">the winds are strong<\/a>, so take note: Here are <b>five things you should look for in a <\/b><a href=\"http:\/\/beeeye.com\/solution\/#_eyeonrisk_platform\"><b>credit risk platform<\/b><\/a> that will help you stay agile, and profitable.<\/p>\n<h5 data-fontsize=\"21\" data-lineheight=\"31.5px\" class=\"fusion-responsive-typography-calculated\" style=\"--fontSize:21; line-height: 1.5; --minFontSize:21;\"><b>A (mostly) no-code approach<\/b><\/h5>\n<ul>\n<li style=\"font-weight: 400;\">Models can be developed in many ways &#8211; you can use anything from pen and paper to coding in <a href=\"https:\/\/www.python.org\/about\/gettingstarted\/\">Python<\/a>. A platform that requires coding means you (or someone else) need to be a skilled developer. Especially if the coder and the analysts are not the same person, any update takes time. The no code approach enables the domain experts to develop the model easily, quickly and accurately. Of course, in some instances, you&#8217;ll still need to code. You want a platform that lets you easily customize and add functionality. Look for a platform that lets you enjoy the advantages of a no-code approach but still lets you add your custom code. This way you&#8217;ll achieve the necessary agility you should be looking for.   <\/li>\n<\/ul>\n<h5 data-fontsize=\"21\" data-lineheight=\"31.5px\" class=\"fusion-responsive-typography-calculated\" style=\"--fontSize:21; line-height: 1.5; --minFontSize:21;\"><b>Continuous model retraining<\/b><\/h5>\n<ul>\n<li style=\"font-weight: 400;\">Models are constantly comparing historical data to production data. As soon as the model is deployed to production there\u2019s already some discrepancy between the training data (old) and the production data (new). This hurts the model\u2019s ability to deliver accurate predictions. What you want is a platform that can leverage the latest production data and retrain the model based on this data. A platform that does this automatically ensures you model delivers the best results.  <\/li>\n<\/ul>\n<h5 data-fontsize=\"21\" data-lineheight=\"31.5px\" class=\"fusion-responsive-typography-calculated\" style=\"--fontSize:21; line-height: 1.5; --minFontSize:21;\"><b>Risk-free experimentation<\/b><\/h5>\n<p>Got a new hypothesis you\u2019d like to test? A platform that lets you experiment on existing models without breaking production will help you ensure model agility. If your platform is code-only based and the developer is not the analysts, the model building process often moves in one direction. This means that taking back a production model to the \u201clab\u201d for more research is cumbersome. A platform that includes both research and deployment activities lets you easily copy your production model and experiment with different hypotheses.  In addition &#8211; working on a web based platform allows for clear and quick communication with the model development team. This is not the case with desktop applications or single-user applications like SAS.<\/p>\n<h5 data-fontsize=\"21\" data-lineheight=\"31.5px\" class=\"fusion-responsive-typography-calculated\" style=\"--fontSize:21; line-height: 1.5; --minFontSize:21;\"><b>External data updates<\/b><\/h5>\n<p>For credit risk, external data like <a href=\"https:\/\/www.investopedia.com\/articles\/mortgages-real-estate\/11\/valuing-real-estate.asp\">real estate analysis of property <\/a>can be as important as historical, internal results, especially when this data reflects a new reality. A platform that lets you easily experiment with new external data will keep your models up to date. Indeed, working with external data is not trivial: There are many providers to choose from, attaining the data can be costly and might require some development work to acquire and incorporate. A no-code approach eliminates the hassle of developing when you only want to experiment with the external data. Once you decide to incorporate it, integration should be a click away.<\/p>\n<h5 data-fontsize=\"21\" data-lineheight=\"31.5px\" class=\"fusion-responsive-typography-calculated\" style=\"--fontSize:21; line-height: 1.5; --minFontSize:21;\"><b>Automation of mundane tasks<\/b><\/h5>\n<ul>\n<li style=\"font-weight: 400;\">We LOVE credit risk modeling. We really do. But even modeling geeks like us have to admit that while things like finding the most significant data point in a model are super exciting, other things are, well, a bit less so (we\u2019re looking at you, model updates paper-trailing). When all the mundane tasks are automated, you have more time for the fun &#8211; and business critical &#8211; stuff. A platform that\u2019s focused on your industry matches your compliance needs and internal processes, and can automate any repetitive tasks. This means you can rapidly respond to change. <\/li>\n<\/ul>\n<p>If your platform is keeping you fixed in place, find one that will lend you more flexibility so you can withstand any winds. If you\u2019re looking to increase your agility,  BeeEye\u2019s EyeOnRisk platform might be the right platform for you. <a href=\"http:\/\/beeeye.com\/contact\/\">Contact us<\/a> to learn more.<\/p>\n<\/div><\/div><style type=\"text\/css\">.fusion-body .fusion-builder-column-0{width:100% !important;margin-top : 0px;margin-bottom : 0px;}.fusion-builder-column-0 > .fusion-column-wrapper {padding-top : 0px !important;padding-right : 0px !important;margin-right : 2.375%;padding-bottom : 0px !important;padding-left : 0px !important;margin-left : 2.375%;}@media only screen and (max-width:1024px) {.fusion-body .fusion-builder-column-0{width:100% !important;}.fusion-builder-column-0 > .fusion-column-wrapper {margin-right : 2.375%;margin-left : 2.375%;}}@media only screen and (max-width:640px) {.fusion-body .fusion-builder-column-0{width:100% !important;}.fusion-builder-column-0 > .fusion-column-wrapper {margin-right : 2.375%;margin-left : 2.375%;}}<\/style><\/div><\/div><style type=\"text\/css\">.fusion-body .fusion-flex-container.fusion-builder-row-1{ padding-top : 0px;margin-top : 0px;padding-right : 30px;padding-bottom : 0px;margin-bottom : 0px;padding-left : 30px;}<\/style><\/div>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":3,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_mi_skip_tracking":false,"footnotes":""},"categories":[12],"tags":[],"_links":{"self":[{"href":"http:\/\/beeeye.com\/wp-json\/wp\/v2\/posts\/1879"}],"collection":[{"href":"http:\/\/beeeye.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/beeeye.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/beeeye.com\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"http:\/\/beeeye.com\/wp-json\/wp\/v2\/comments?post=1879"}],"version-history":[{"count":1,"href":"http:\/\/beeeye.com\/wp-json\/wp\/v2\/posts\/1879\/revisions"}],"predecessor-version":[{"id":3399,"href":"http:\/\/beeeye.com\/wp-json\/wp\/v2\/posts\/1879\/revisions\/3399"}],"wp:attachment":[{"href":"http:\/\/beeeye.com\/wp-json\/wp\/v2\/media?parent=1879"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/beeeye.com\/wp-json\/wp\/v2\/categories?post=1879"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/beeeye.com\/wp-json\/wp\/v2\/tags?post=1879"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}