Page tree

 

 

 

 

 

 

 

 

FIBO Loan Content Team:

Use Case for HMDA Data Integration

 

Name:  Loan Content Team –HMDA Integration

Date: April 6, 2015

 

 

 


Table of Contents

Use Case List

1 HMDA Data Integration for Reporting

1.1 Feature Process Flow

1.2 Use Case(s)

2 Integration and Analysis of Published HMDA Data

2.1 Feature Process Flow

2.2 Use Case(s)


Revision History

Version

Date

Revision Description

.01

04/06/2015

Initial draft of HMDA integration use case for review with FIBO Loan Content Team

 

1.0

 

 

 

 

 

 

 

 

 


Use Case List

Use Case ID

Primary Actor

Use Cases

HMDA1

Lender

Rationalize Line-of-Business specific data for HMDA Reporting to create an integrated LAR report for submittal to applicable Regulators

HMDA2

Public Data User

Integration and Analysis of HMDA Data from CFPB published data

 

 

 

 

1        HMDA Data Integration for Reporting

1.1    Feature Process Flow

 

 

 

 

 

 

 

 

 

 

 

 

1.2    Use Case(s)

Use Case ID:

HMDA1

Use Case Name:

Rationalize Line-of-Business specific data for HMDA Reporting to create an integrated LAR report for submittal to applicable Regulators

Created By:

L. Calahan

Last Updated By:

 

Date Created:

04/06/2015

Last Revision Date:

 

Actors:

  1. Line of Business user(s) or system(s)
  2. FIBO Integration module
  3. Rule interpretation manager
  4. Regulator submission reporting reviewer
  5. Regulator

 

Description:

Multiple lines of business submit native data files related to HMDA covered loans that have been processed through their systems in the applicable date range to the FIBO integration module, where the files are rationalized against the required regulatory requirements, and integrated to create one regulatory report – the Loan Application Register-- for required HMDA reporting to the regulator.

Trigger:

Lines of Business create native data files for HMDA reporting, and submit them to FIBO Integration module.

Preconditions:

 

  1. Rule interpretation manager has created data terms, parameters, and process requirements in compliance with Rule to create FIBO integration processing module.
  2. Lines of Business have created raw data files in compliance with native data terms acceptable to the translation function.

 

Post conditions:

 

  1. If successful, the Integrated LAR report is returned
  2. If not successful, exceptions are returned to the Regulator Submission Reporting reviewer.

 

Normal Flow:

 

  1. Lines of business process loan applications in accordance with normal processing in native systems
  2. Lines of business produce native data files for HMDA covered loans and loan applications
  3. Files are received in FIBO integration module.
  4. FIBO Integration module translates native data terms to rationalized common data terms for each file
  5. FIBO Integration module appends each file to the universal corporate output file.
  6. FIBO Integration module outputs exceptions by record for review.
  7. FIBO Integration module outputs universal corporate output file in compliant LAR reporting format.

 

Includes:

  1. FIBO interpretation of HMDA Data File Format from Rule (spreadsheet entitled (Proposed HMDA DDF Draft 0.v MASTER external yyyymmdd.xls) where “v” represents current version.
  2. Finalized Rule mandated and distributed by CFPB (regulator)
  3. HMDA Data File Format from Rule (spreadsheet entitled (Proposed HMDA DDF Draft 0.v MASTER external yyyymmdd.xls) where “v” represents current version.)

 

Frequency of Use:

The LAR report will be executed in compliance with the CFPB Final Rule, which may be annually or at other periodicity as requested by the CFPB (quarterly).

Special Requirements:

The LAR Report must be returned in the defined regulatory format, and must comply with data quality and formation rules specified by the final Rule and documented in the HMDA Data File Format from Rule (spreadsheet entitled (Proposed HMDA DDF Draft 0.v MASTER external yyyymmdd.xls) where “v” represents current version.)

 

Assumptions:

  1. Lines of business prepare files related to HMDA covered loans for HMDA LAR reporting in expected formats.
  2. HMDA processing requirements and data terms are interpreted in the FIBO Integration module from the HMDA Data File Format from Rule (spreadsheet entitled (Proposed HMDA DDF Draft 0.v MASTER external yyyymmdd.xls) where “v” represents current version.)
  3. The FIBO Loan model is based upon MISMO standard concepts

 

Notes and Issues:

The current HMDA Data File Format has been created from the CFPB Proposed Revised Home Mortgage Disclosure Act (Reg C), issued by the Consumer Financial Protection Bureau in the Federal Register in July of 2014. An updated spreadsheet will be created when the Rule is finalized in late 2015.

 

2        Integration and Analysis of Published HMDA Data

 

2.1    Feature Process Flow

 

C:\Users\f358915\Documents\Architecture\EDM\FIBO\FIBO MISMO HMDA Ontology Use Case.jpg

2.2    Use Case(s)

 

 

 

Use Case ID:

HMDA2

Use Case Name:

Integration and Analysis of HMDA Data from CFPB published data

Created By:

J. Cooper

Last Updated By:

 

Date Created:

04/08/2015

Last Revision Date:

 

Actors:

  1. Data Providers
    1. Consumer Finance Protection Bureau (CFPB)
    2. Census Bureau
    3. Department Of Commerce
    4. Market Research Corporations
    5. National Association of Homebuilders
    6. Other data providers
  2. Ontology Providers
    1. EDMC/OMG
    2. GeoNames
    3. Census Bureau
    4. Other Ontology providers
  3. Business Analyst

Description:

An analyst uses ontologies to integrate multiple data sources, including HMDA data, Census Data, Housing Market Data, and geographic data to infer important business information that is not available from any single data source.

Trigger:

The analyst poses a significant business question for which there is no single information source that can provide the answer.

Preconditions:

  1. The CFPB has published the HMDA datasets
  2. Other data providers have published their data
  3. FIBO, GeoNames, and other needed ontologies have been published

Post conditions:

  1. The business question has been answered.

Normal Flow:

  1. The analyst identifies data sources that contain the pieces of information that are needed to answer the question, along with the ontologies that describe that data.
  2. The analyst identifies common dimensions (e.g., geography) that can be used to integrate the data.
  3. The analyst may need to create an ontology to identify equivalences between existing ontologies, for the integration dimensions.
  4. A semantic tool integrates the information and infers new knowledge that answers the analyst’s question.
  5. The analyst is promoted to VP of Marketing

Includes:

HMDA, Census, and Market data

Ontologies that describe these data sets

Ontologies that describe dimensions that can be used for data integration

Frequency of Use:

Ad hoc.

Special Requirements:

Analytic tool set that is capable of semantic inference using ontologies

Assumptions:

Data is accurate, consistent, and timely within the criteria of the analysis

Notes and Issues:

 

 

 

Note:  FIBO would allow the ability to integrate between FIBO and other ontologies, but would not pursue a tight coupling with another ontology.  This Use Case needs to be updated to reflect this “enabling” versus an explicit integration to other ontologies.

 

Example:

 

Company A has a competitive advantage in the Mortgage Origination industry in that its employees, systems, and products are bilingual in English and Spanish.   It wishes to leverage this advantage by focused marketing to underserved Hispanic populations.

 

It needs to identify the top 10 US regions of underserved Hispanic populations, taking into account ethnicity, home ownership rates, population, income distribution, property valuation trends, and other factors.   It also needs to determine, for regions with well-served Hispanic populations, what

are the preferred mortgage types and characteristics among Hispanic borrowers.

 

Information about mortgages and the ethnicity of the mortgagees is available in the HMDA dataset; it is summarized geographically by CBSA.   Hispanic borrowers can be identified using HMDA Ethnicity Type

(Hispanic/Non-Hispanic/NA/Declined to Answer).

 

Information about population, income, geography, and ethnicity, with trending data, is available from

the US Census.   It is presented geographically by Census Block.   Ethnicity has dimensions of Mexican, Cuban, Puerto Rican, Central American, South American (Spanish), and many others that are non-Hispanic ethnicities.

 

Housing market data is available from the National Association of Home Builders.   It includes trending

data and projections.   It is presented geographically by ZIP Code.

 

The GeoNames Ontology describes the relationships between different types of geographic regions, including Census Blocks, Postal Codes, CBSAs, MSAs, Counties, Cities, States, and many others.

 

The analyst creates a small ontology that states that HMDA Race Type (Hispanic) is equivalent to the  union of the Census Ethnicity categories of Mexican, Cuban, Puerto Rican, Central American, and South American.  

 

The analytic tool set uses all this information to identify the top 10 US regions of underserved Hispanic populations, taking into account ethnicity, home ownership rates, population, income distribution,

property valuation, trending data, and other factors.

 

A similar analysis identifies the top 10 best served regions for Hispanic borrowers.

 

The analytic tool then examines the HMDA data for the best served regions to infer what mortgage products are most popular among Hispanic borrowers.

 

The Analyst proposes a new marketing campaign targeting the Hispanic communities in the 10 most underserved regions, with emphasis on the mortgages products that are most popular among Hispanic borrowers.

 

  1.  
  2.  
  3.  
  4.  
  5.  
  6.  
  7.  
  8.  
  9.  
  10.      
  11.