1557-0331 6-8-16

Office of the Comptroller of the Currency

Supporting Statement

Domestic First Lien Residential Mortgage Data

OMB Control No. 1557-0331

A. Justification

1. Circumstances that make the collection of information necessary:

Comprehensive mortgage data is vital to assessing and monitoring credit quality and loss mitigation activities in the residential mortgage market and the federal banking system. This data is important and necessary to support supervisory activities to ensure the safety and soundness of the federal banking system.

The Dodd-Frank Wall Street Reform and Regulatory Improvement Act of 2010 requires the OCC to collect this mortgage data. 12 U.S.C. 1715z-25.

2. Use of the information:

The loan-level data collection is being revised to include aggregate values to be calculated from data that is currently reported in loan-level format. These aggregate values will be industry standard measures of portfolio performance, including but not limited to: outstanding loan count and unpaid principal balance; delinquency and liquidation ratios; and the number of loss mitigation actions completed. Aggregate values generally will be reported at the total portfolio and state level, with some values also reported by portfolio segments including, but not limited to: borrower credit class and type and execution date of loss mitigation action.

The reported data items will still be calculated from loan-level data that includes: bankruptcy or foreclosure status; and other detailed loan information. Banks would not be required to report this data to the OCC monthly, but would be required to provide it upon OCC’s request.

The current collection of monthly first lien real estate mortgage loan-level data will continue after OMB approval of the aggregate data collection and will cease on November 30, 2016.

The primary method used by the OCC to provide effective oversight of national banks and Federal savings associations is through ongoing supervisory activities and targeted exams.  

Local bank examiners review board and management reports, which provide aggregate and detailed information on all aspects of a bank's mortgage lending activities, on an ongoing basis.  They also conduct targeted examinations of various aspects of mortgage operations in which the teams review specific loan information to evaluate underwriting, default management, and other risk issues on an individual loan as well as a portfolio basis.  As a result of post-crisis bank actions and OCC expectations for risk management, these internal bank reports have become significantly more robust in the last 5+ years.  

During ongoing supervision and targeted reviews, exam teams request specific information and customized reports from the bank to achieve the OCC's bank-specific supervisory objectives.  Examiners summarize their assessments in conclusion memos (internal documents), risk assessments (internal documents), and supervisory letters (communication to banks).   Exam teams also complete an annual credit underwriting assessment of major portfolios, including mortgages. The OCC's process for completing these assessments has become more robust since the crisis and currently includes examiner assessments of new products and changes in underwriting practices.  When examiners identify concerns, they communicate them to the bank and record them in the OCC's supervisory information systems.  Since the financial crisis, the OCC has expanded its reporting on concerns so it can use this information to identify risks and trends from a systemic perspective, including risks arising from new products and changes in underwriting.  

 

Lead Expert teams (created since the financial crisis) participate in the aforementioned supervisory activities and conduct horizontal analyses using the examiner prepared documents and “concerns” information described above to conduct horizontal and trend analyses.   The aggregate data currently derived from the loan-level collections is used to supplement and cross check the information provided by the examiners.  As long as we have aggregate data, we can continue to do this.   If we see discrepancies between the aggregate data and the information provided by examiners, we can investigate through the local teams.  

As needed and requested, local teams are given direct “view”' access to specific bank information systems that house loan-level information.  This gives the OCC access to more robust information than what is included in the existing mortgage metrics report.  

In addition to the above, the OCC's Lead Expert and Risk Analytics teams (both created since the financial crisis) periodically meet and share information with counterparts at the Federal Reserve and FDIC to help identify and respond to emerging and systemic risks.  Discussions are based on aggregate data.  Loan level data is used to drill down and investigate anomalies, when needed.   In addition, the Federal Reserve's data collection provides sufficient information to meet these needs.  

3. Consideration of the use of improved information technology:

The collection of information will be available electronically. Any improved information technology may be used to meet the requirements of this collection.

4. Efforts to identify duplication:

The OCC has an information sharing agreement with the Board of Governors of the Federal Reserve System (FRB) covering any sharing of information from their domestic first lien residential mortgage data collection. FRB accesses account-level data from 61 respondents, including nine financial institutions, for first-lien residential mortgage data.

5. Methods used to minimize burden if the collection has a significant impact on a substantial number of small entities:

Not applicable.

6. Consequences to the Federal program if the collection were conducted less frequently:

Less frequent notice would substantially impair the effectiveness of the program.

7. Special circumstances that would cause an information collection to be conducted in a manner inconsistent with 5 CFR Part 1320:

None. The information collection is conducted in accordance with OMB guidelines in 5 CFR part 1320.

8. Efforts to consult with persons outside the agency:

On March 7, 2016, the OCC published a notice for 30 days of comment concerning the collection, 81 FR 11904. One comment from an individual was received.

The commenter believed that the revision to the collection was unnecessary and that the burden of the collection could be minimized if the OCC calculated the aggregate values instead of the banks. The commenter asked the OCC to explain why it is not less burdensome for the OCC to calculate aggregate values from the data it already collects.

The OCC believes that no longer requiring monthly submission of loan-level data reduces regulatory burden to the reporting financial institutions. In addition, given changes in mortgage industry concentrations and the economic climate, the OCC believes the revised format of the MMR and streamlining the data collection processes that support it further reduce the burden. The OCC is decreasing the number of data elements collected and plans to limit the number of metrics collected to those  required by Section 104 of the Helping Families Save Their Homes Act of 2009 (12 U.S.C. 1715z-25) (the Act) supplemented by a limited number of essential performance metrics that are industry standards.  The data collection would be aggregate metrics reported by the bank quarterly directly to the OCC. The OCC believes that because banks have this data readily available and that having them provide it to the OCC is most efficient and the least burdensome method of obtaining it.

The commenter stated that the issue of burden is especially important because the OCC does not need the aggregate value information on a set schedule and the banks will provide the information only upon request. The commenter also stated that if the OCC cannot aggregate the loan-level data due to data quality problems, then aggregate values on poor-quality datasets submitted by banks would have no practical utility. The commenter stated that the OCC should explain how any aggregate values on a poor-quality data helps the OCC achieve its mission. The commenter asserted that poor data quality can increase costs by causing resources to be misdirected and that the OCC should explain why it is not addressing any such problems. The OCC maintains that there is no indication that banks produce poor quality datasets. The current OCC data aggregation contract is set to expire in November 2016 and loan level data will no longer be available for aggregation. Therefore, the collection of aggregate metrics is necessary in order for OCC to report and comply with Section 104 of the Act. In addition, the OCC believes there will be an increase in data quality since banks will perform data aggregation services as well as data integrity and quality checks will be used to verify data based on internal control processes.

9. Payment or gift to respondents:

None.

10. Any assurance of confidentiality:

The information collected will be kept private to the extent permitted by law.

11. Justification for questions of a sensitive nature:

Not applicable. No sensitive information is collected.

12. Burden estimate:

Aggregate Data Collection:

Estimated Number of Respondents: 61.

Estimated Annual Responses per Respondent: 4 per year.

Estimated Burden per Response: 120 hours per month/per bank.

Estimated Total Annual Burden: 29,280 hours.

Cost to Respondents:

29,280 x $101 = $2,957,280.

To estimate average hourly wages we reviewed data from May 2014 for wages (by industry and occupation) from the U.S. Bureau of Labor Statistics (BLS) for depository credit intermediation (NAICS 522100). To estimate compensation costs associated with the rule, we use $101 per hour, which is based on the average of the 90th percentile for seven occupations adjusted for inflation (2 percent), plus an additional 30 percent to cover private sector benefits. Thirty percent represents the average private sector costs of employee benefits.

Loan-Level Data Collection:

Estimated Number of Respondents: 61.

Estimated Annual Responses per Respondent: 12 per year.

Estimated Burden per Response: 430.

Estimated Total Annual Burden: 314,760 hours.

Cost to Respondents:

$ 1,028,000.00

To estimate costs associated with the collection, we used the total contract cost for the information collection. OCC staff believes that national banks’ and Federal savings associations’ reporting costs are substantially lower.

Total: 344,040 hours.

13. Estimate of total annual costs to respondents (excluding cost of hour burden in Item #12):

Not applicable.

14. Estimate of annualized costs to the Federal government:

None.

15. Change in burden:

Prior Burden: 314,760 hours.

Current Burden: 344,440 hours.

Difference: +29,680 hours.

The increase in burden is due to the overlap in the collections. The loan-level collection will be discontinued on November 30, 2016.

16. Information regarding collections whose results are to be published for statistical use:

No publication for statistical use is contemplated.

17. Reasons for not displaying OMB approval expiration date:

Not applicable.

18. Explain each exception to the certification statement:

Not applicable.

B. Collections of Information Employing Statistical Methods

Not applicable.

6