Latest Fannie DU version limits risk layering

Latest Fannie DU version limits risk layering

Written By: Joel Palmer, Op-Ed Writer

Less than a year after enabling higher debt-to-income (DTI) ratios for certain mortgage borrowers, Fannie Mae is adjusting its underwriting standards to address the risk associated with many of these loans.

Last July, Fannie implemented a change to its Desktop Underwriter (DU) that permitted a DTI up to 50 percent, an increase from 45 percent. Loan ratios above 45 percent would “rely on DU’s comprehensive risk assessment,” and the system“removed specific rules that had previously set maximum loan-to-value ratio and minimum reserves requirements for those loans.”

Many in the industry applauded the change as a way to help many first-time homebuyers, especially those carrying heavy student loan debt, qualify for a mortgage. 

At the time, Fannie Mae officials said they were comfortable with the change. Their research indicated that many borrowers with DTIs between 45 percent and 50 percent have good credit and not prone to default. Many have adequate downpayment and/or reserve funds to offset any risks from having a higher DTI ratio. Plus, Fannie Mae pointed out that having a DTI of 50 or under doesn’t automatically qualify somebody for a home loan; income, credit scores, and LTV ratios will still be strongly considered.

But according to the GSE’s recent 10-K filing, the percentage of single-family mortgages purchased with DTIs higher than 45 percent have significantly increased since the change.

In the fourth quarter of last year, 20 percent of Fannie’s mortgage acquisitions had a DTI higher than 45 percent. For all of 2017, the number was 10 percent. For the full year of 2016, only 5 percent of Fannie’s mortgage purchases had a 45-percent or higher DTI. 

According to both the 10-K filing and the release notes for DU 10.2, Fannie has assessed the profile of loans delivered since the change went into effect. As a result, Fannie is “fine-tuning” DU’s risk assessment to limit risk layering. This refers to the acquisition of loans with multiple higher-risk characteristics, including high LTV ratio, credit profile with a history of delinquencies, debt-to-income ratio above 45 percent and no or low levels of reserves. 

“With DU Version 10.2, we expect fewer DU Approve recommendations on loans that have multiple higher-risk characteristics; however, we expect to continue to acquire a higher proportion of loans with debt-to-income ratios above 45% than we have in previous years,” read the 10-K filing.

“This update supports our commitment to prudent risk management, helping lenders to better manage default risk while continuing to provide sustainable homeownership options to borrowers,” stated the release notes for DU 10.2.“There will be no change to the risk factors evaluated by DU.”

Private mortgage insurers are also raising their standards. Essent has implemented a minimum FICO score of 700 for loans with a DTI higher than 45 percent. MGIC and Genworth Mortgage have implemented similar policies this month. Radian Guaranty has disallowed high DTI loans in cases where a downpayment is less than 5 percent.

In addition to the risk assessment changes, DU Version 10.2 will also include the use of the Experian trended credit data received in the credit report as part of the credit risk assessment. The DU risk assessment will now include trended credit data from Equifax, Experian, and TransUnion.


About the Author

As an NAMP® Opinion Editorial Contributor, Joel Palmer is a freelance writer who spent 10 years as a business and financial reporter and another 10 years in marketing for the insurance and financial services industries. He regularly writes about the mortgage industry, as well as residential and commercial real estate, investments, and retirement income planning. He has also ghostwritten books on starting a business, marketing, and retirement income planning.


Opinion-Editorial (Op-Ed) Disclaimer For NAMP® Library Articles: The views and opinions expressed in the NAMP® Library articles are those of the authors and do not necessarily reflect any official NAMP® policy or position. Examples of analysis performed within this article are only examples. They should not be utilized in real-world application as they are based only on very limited and dated open source information. Assumptions made within the analysis are not reflective of the position of NAMP®. Nothing contained in this article should be considered legal advice.