Skip to main content
Debunking the Top-3 Pooled Model Myths

In my previous blog post, we explored how using pooled models can help you say “yes” to more good credit applicants. In this final blog in my series, let’s explore how can lenders take advantage of advanced modeling technology to cost-effectively originate profitable, compliant decisions that protect your bottom line AND deliver an optimized customer experience to the best customers.

Does that sound like a tall order? You bet. Is it doable? Absolutely.

First, let’s do some myth debunking about pooled models. We hear these ones time and time again.

MYTH #1: We already use a credit bureau score to assess application risk. We don’t need another score to sort out the good from the bad.

DEBUNK: Lenders need powerful, focused risk prediction, and 80% of measurable risk is decided at origination. Individual scores and pooled models are complementary and provide substantial benefit together. Many scores are based on credit and non-credit seekers, meaning that repayment behavior is weighted for non-credit seekers. Pooled models can be solely developed for credit applicants, making them much more sensitive to repayment risks of a new obligation.

MYTH #2: We don’t have the time, money, or resources to deploy and support complex models. It’s all too complicated and hard to manage.

DEBUNK: Lenders we talk to are often concerned about model development, deployment, and maintenance costs. However, many pooled model suites offer “models to go” that are off-the-shelf and cost-effective to deploy and maintain. The leading pooled models are flexible, agile, streamlined, and can assess risk quickly without complex data gathering or costly development and deployment. Armed with industry-leading pooled models, lenders can apply application scoring to new or niche portfolios (like geographic expansion) that lack the extensive history required for building a custom model. This helps lenders grow portfolios responsibly.

MYTH #3: We will always struggle to balance compliance and customer experience. It’s an unwinnable war, and something has to give.

DEBUNK: It’s the never-ending battle—staying up-to-date with compliance, offering transparency, and delivering the best customer experience. They sound like they are always going to be at odds, but with pooled models, that equation changes. Pooled models should offer world-class predictive analytics with intuitive reason codes that help lenders stay compliant. With highly predictive pooled models, intuitive reason codes are calculated with the score to explain the most important influencing factors. They are often used to assist “adverse action” notification when an applicant is denied credit, proving helpful when responding to consumer and legal inquiries. This “enriched documentation” improves CEX, provides transparency, and supports compliance.

Best Practices: Must-Have Checklist for Winning Pooled Models

To improve decision making and reduce costs with pooled models, ask these questions:

  • Does the model suite solely target credit seekers? Is the model sample representative of the application population, or is it weighted for the general public (which includes non-credit seekers)? This can affect the predictive sensitivity for repayment risk on a new obligation.
  • Does the model suite streamline operation and cut decision time?
  • Does it offer consistent scaling across products to allow credit grantors to compare risks between different products?
  • Does the model suite use scorecard technology augmented with machine learning techniques?
  • Does the model suite offer provide access to a bureau characteristic library to capture more aspects of information?

To support regulatory compliance with pooled models, investigate these questions:

  • Does the model suite comply with fair lending laws – such as the Fair Credit Reporting Act (FCRA), the Equal Credit Opportunity Act (ECOA), and the Fair Housing Act (FHA)?
  • Is the model suite tailored to achieve precision, accuracy, robustness, stability, and reliability? Can those results be measured?
  • Does the model suite provide enriched documentation support, including a regulatory package to address compliance questions and additional guidance for new and existing clients?
  • Does it generate intuitive reason codes to explain what items on the customer’s credit report most negatively influenced the score computation? Do the reason codes help you quickly respond to regulatory concerns and customer questions?
  • Is the model suite designed to be Empirically Derived, Demonstrably and Statistically Sound (EDDSS)? Using offers certain legal protections for creditors from fair lending challenges to their underwriting practices based on age discrimination and other types of discrimination, including disparate treatment and disparate impact.

To drive profitability and get value from your pooled models, demand proof and measurable results:

  • How many model suites and models are available? Are there consistent scaling and performance definitions? The best solutions provide diverse model suites and multiple tailored scorecards; for example: Installment Suite (8 scorecard models), Home Equity Suite (1 model), Revolving Suite (2 models).
  • How well do the model suites perform? The Bad Capture Rate for the worst 20% of the scorecard population should give big gains. Industry-leading pooled models can deliver impressive results:
    • Home Equity: Identified 76% of the potential bads over the next year.
    • Installment: Identified 62% of the potential bads over the next year.
    • Revolving: Identified 77% of the potential bads over the next year.
  • What are the proven results and from how many records? The best solutions have 5M+ records assessed from a representative data sample. They also include 250+ predictors from characteristic libraries and have a 2-year performance window for credit bureau data.
  • What about advanced modelling technology? This can be a game-changer for many lenders, and it is distinct competitive advantage to operationalize analytics in today’s data-driven world.

The Bottom Line

Today’s lenders must be able to make fast, informed, and precise decisions to reduce delinquency and charge-off losses, approve more applicants and increase profitability, streamline operations, cut decision time, and ensure regulatory compliance. And, one score just isn’t enough.

To learn more about how industry-leading predictive pooled models (like FICO® Application Risk Models (ARM) 4.0) can help your organization build more profitable portfolios while intelligently assessing and managing customer-level risk, visit: 

related posts