Statement of Objectives.pdf

STATEMENT OF OBJECTIVES

FOR

Archival Credit Data I. Purpose

The purpose of this project is to provide the Department of Housing and Urban Development with access to archival credit data for U.S. mortgagors for use in research on an as needed basis in three broad areas:

Federal Housing Administration (FHA) Analysis, Affordable Lending Analysis, and Primary and

Secondary Mortgage Market Studies.

II. Background

The Department of Housing and Urban Development continually monitors and evaluates the performance of FHA, the GSEs, Lenders, Mortgage Borrowers, and overall housing and mortgage market conditions.

Effective analysis of mortgage origination and performance requires access to archival credit data. The requisition seeks to provide on demand access to historical archive credit data for use by the Department of

Housing and Urban Development and its agents for mortgage market research, monitoring, and evaluation.

III. Scope

The scope of work for which archival credit data may be used includes the following

Federal Housing Administration (FHA) Analysis

TOTAL Recalibration & Updating: Scoring systems require periodic re-estimation to assure that they continue to perform well for changing populations and conditions. The re-estimations are made using samples that have been augmented with data on more recent borrowers for whom the associated credit data must be acquired.

Monitoring scoring systems: In the interval between scoring system re-estimations, the models need to be monitored for significant changes in performance with respect to approval rates and an examination of the types and characteristics of borrowers who are accepted or referred by the systems, to guard against disparate impact.

Stand-alone default studies: Scoring systems are designed to exclude factors such as local or national economic conditions from the underwriting decision. Examination of the effect of these factors on default is important for issues such as actuarial soundness of the FHA insurance fund. Credit data would enhance the predictive power of default models.

Analysis of delinquencies: The FHA TOTAL mortgage scoring system is based on prediction of mortgage default. HUD may want to focus on a model based on delinquencies in the future in which a delinquency scoring system might be used for loss mitigation purposes as opposed to loan approval.

Affordable Lending

In recent years there has been substantial focus on affordable lending, both by FHA and conventional lenders. HUD is interested in extending loans to currently underserved borrowers. Analysis of the credit characteristics of this population will be helpful in designing loan products and underwriting standards to serve them.

Primary and Secondary Mortgage Market Studies

Mortgage markets are continually evolving in response to changes in the economy, the financial industry, and housing market conditions. Trends toward automated underwriting, risk-based pricing, new products, and capital conduits, such that credit scores are now critical to monitoring and understanding the structural changes and, their impact with an eye to developing responsible regulatory responses.

Authorized Archival Credit Data Users

Archival credit data and scores will be used by the Department of Housing and Urban Development and the following entities could be extended access to the analysis files with coded archived data subject to specified conditions:

1. Agents of the Department, such as research contractors or academic institutions or researchers that have entered into a contractual agreement to furnish their research product to the Department.

2. Oversight & coordinate agencies, including GAO, OMB, CBO, DOJ, and the U.S. Congress.

3. Other Federal agencies involved in public purpose research, such as the Treasury Department, Federal

Reserve System, Federal Housing Finance Agency, or Consumer Financial Protection Bureau.

Provision of data to other entities would be on a case-by-case basis, with approval from the Credit

Repository.

Permissible purpose is governed by the definition in Section 604 of the Federal Credit Reporting Act

(FCRA). In some cases, the credit data HUD will request will involve a permissible purpose for uncoded credit information, such as information on current FHA borrowers. For these cases, HUD would be provided credit data with identifiers. In other cases, HUD will purchase anonymous (i.e., coded) data for individuals for whom HUD does not have permissible purpose. The contract will clearly specify the circumstances under which permissible purpose exists.

IV. Program Objectives

The Department shall request credit history data from archival files that will be processed through individual work orders. Each work order will be issued by the GTR and will request raw credit data, attributes, and/or summary indices, such as FICO or Vantage scores, as needed for the specific work order.

Credit data may be requested for a national sample of FHA borrowers or loan applicants or for all borrowers/applicants in a few MSAs. Alternatively, coded credit data may be requested for samples of non-

FHA borrowers.

The Contractor may be requested to perform analyses, such as frequency distributions. The Contractor may also be asked to link coded credit data with other data, such as geographic identifiers, HMDA, or Census data, and to update coded files with subsequent changes in the status of some variables, such as claim/termination status.

V. Performance Period

The contract shall include a period of performance of three years.