Vendor

Loan Application Analytics with CUFX

What is CUFX

CUFX is an open, vendor agnostic and broad integration standard designed by leading Credit Unions (CU) and vendors to reduce the time and cost of systems integration.

CUFX Objectives

  • Improve the member experience and help CUs rapidly innovate
  • Design a standard to simplify application to core integration
  • Reduce time, cost and redundant system integration efforts

Use Case for CUFX

Credit unions maintain applications for different loan products in multiple sources and spend lot of engineering time to create reports and analysis for executives to answer various business questions related to loan applications. CUFX provides a standardized model that enables a common nomenclature among multiple business units and helps in unified reporting and analytics process once multiple sources are normalized.

  • Number of applications received per month across different loan product types?
  • What percent of applications were denied?
  • How are the application disposition on a loan type level, branch level, geography, credit scores, LTV, DTI, race, gender, & age?
  • How has the volume of loan applicants trended over time?
  • Is the credit quality of applicants decreasing or increasing over time?
  • Has application decisioning been consistent over time based on credit score, LTV, DTI, etc
  • Do certain product types have higher approval ratios than others?
  • Is there evidence of disparate treatment or fair lending at the institution?
  • Perform necessary predictive analytics to understand the nature of approval process and perform what-if analysis to approve loans for risky members at higher interest rates
  • Perform analytics to understand the reason behind withdrawns and how to win them back
  • Create segmentation for targeted marketing purpose

CUFX for Loan Application Model

CUFX has standard schema model for application, applicant, loan product applied for, status of the application at different time period, etc. This will enable a standard nomenclature across different business units and also reduce time for reporting and analytics across multiple loan product types. cufx_application_applicant

  • Application: An application contains the details of a consumer who is becoming a customer of a financial institution (if they are not already). It also includes the details of the products requested.
  • Applicant: Collection of fields to describe an applicant for the product.
  • Party: The party object defines all the fields necessary to define a person, business or vendor related to a financial institution. The ID uniquely identifies the party within the financial institution.
  • finalCreditBureauScore: Credit score that was used to approve the application. This is typically calculated using a variety of methods at each financial institution based on the credit report data for the primary, the joint or a combination of both for primary and joint credit statistics.
  • productAppliedFor: Product applied for and its requested amount, quoted rate, etc.
  • productDetail: Loan or Deposit

Loan Application Data Pipeline Process

The below big data lake pipeline process showcases the flow to consume raw application data from the source, normalize and standardize it as per CUFX model in the transient phase, and enhance or correlate with external data on Application & Applicant entities in the refined phase and make it available for reporting and other analytics such as diagnostic, descriptive, predictive, and targeted marketing purposes. applications_detailed_data_pipeline

Loan Application Analytics with Tableau Public

To keep things simple for this blog, we have used Tableau public and the csv files that were extracted from Hive, and created different visualizations. These data files are based on CUFX normalized model obtained from the refined phase and mostly uses Applicant and Application entities. 

Email me when people comment –

You need to be a member of CULytics to add comments!

Join CULytics

Comments

  • CU Employee Community President

    Thanks for sharing this. very timely for what we are doing.

    We are looking at loan analytics along following member flow steps.

    1. Marketing Analytics - Campaign (Social, Search, Banner, Email, etc) analysis, Website analysis, Conversions, Click thru rate, Cost per lead/opportunity, etc.

    2. Application Analytics - Application funnel analysis, Started vs Completed

    3. Loan Origination Analytics - Approved vs Funded, Applicant Analysis, Region Analysis, Product Type Analysis, Profitability/Expected Loss, etc.

    4. Funding and Usage Analytics

    5. Member Experience - Overall member experience, NPS, etc.

This reply was deleted.