Home >> Daily Dose >> Why Repurchase Risk is Growing—And How to Manage It
Print This Post Print This Post

Why Repurchase Risk is Growing—And How to Manage It

This piece originally appeared in the February 2022 edition of MReport magazine, online now.

It’s been some time since repurchase risk was a focal point of the mortgage industry. But those who were around during the last housing crisis can remember quite well the days when, if you could sign a note, you could sell a loan. It left a big impression on me because it made me realize the impact of sound underwriting practices and loan due diligence, which is what my career became focused on.

As we see the market expanding in purchase originations and new loan products, many are asking whether we are digging ourselves into a hole once again. Personally, I don’t think so, since underwriting standards are significantly higher than they were 15 years ago. However, with the market transitioning away from refinances toward purchase loans, repurchase risk is likely to become a focal point in the new year.

For this reason, now is a great time for originators to consider whether they are deploying the right tools and strategies to minimize risks, so they can hit the ground running in 2022 with confidence. The key is viewing quality control not as a cost center, but as an investment that ensures you’re going to be successful in the years ahead.

Repurchase Risk Will Grow
The good news for lenders is that purchase mortgage originations should increase next year. The Mortgage Bankers Association, for example, predicts these originations will grow 9% to a record $1.725 trillion. Yet according to a recent STRATMOR Insights report, lenders are concerned about how they are going to keep pace with volume while maintaining their profit margins, as purchase loans are invariably more complex and therefore more costly to produce than refinances.

One factor behind the growing origination costs is compensation. Loan officers who specialize in purchase loans are usually paid more, because they have spent years building relationships with buyers and referral partners. Many lenders have also been ramping up their recruiting campaigns to find these top performers and take advantage of their relationships, which translate into additional volume. Yet these days, with fewer quick-turnaround refinances and housing supply strained to meet demand, a loan officer’s relationships may take longer to develop into home sales and closed loans.

With housing inventory so tight and intense competition taking place between buyers in many markets, it can take up to six months to work with someone before they become a buyer. When they do, purchase contracts are often written with shorter turn times so that borrowers can be more competitive with cash borrowers—which brings additional pressure on lenders to close quickly and on time. The readiness of a lender’s staff to enable this, with no increase in defects, may be challenged as lenders shift resources and ramp up purchase underwriting skills, which may open the door to additional risk.

Currently, a growing area of risk for purchase loans are appraisals, which is understandable given the rapid rise of home values and the pressure to turn appraisal reports around quickly. Furthermore, with Fannie Mae and Freddie Mac both incorporating desktop appraisals into their selling guides, it becomes important to ensure that post-close quality control include the completion of a Collateral Risk Assessment questionnaire (or similar review method) by a skilled subject matter expert on appraisal theory and that it is included as part of the loan-level reporting package.

Another area of concern stems from the number of recent streamlined refinances that have been underwritten with limited borrower credit documentation and underwriting. For many lenders, it will be a challenge to break out of that streamlined refi mindset and make sure they have the skills to calculate income and perform the kind of diligent underwriting that is required for purchase loans. Automated income calculation tools can certainly help to create more consistent decisioning and accommodate skill gaps.

The rapid rise of non-QM loans is yet another factor contributing to repurchase risk. McKinsey recently identified non-QM loans as one of five trends that will shape next year’s housing market, as lenders look for ways to serve the growing number of borrowers who are self-employed or have income streams that disqualify them from traditional mortgages. Since the early days of the COVID-19 pandemic, when non-QM loan volume virtually dried up, more non-QM lenders have been entering or re-entering the market. Meanwhile, the number of securitizations of non-QM loans grew steadily throughout 2021 and is expected to continue in 2022.

Certainly, there are many self-employed borrowers who have the ability to repay. Problems occur, however, when lenders start working with borrowers but lack controls to document their underwriting decisions. It will be important for lenders to focus on all facets of credit, income, and assets to make sure the data they are using has been validated. The use of automated tools that can take some of the complexity out of the process will certainly help. Also, performing a more thorough credit analysis and increasing audit reviews for these types of loans can minimize issues. Non-QM loans by nature come with a certain agreed-upon risk, yet it’s important for lenders to have safeguards in place to make sure they are not exceeding these levels.

Fortunately, I see most lenders are taking a more serious view of risk in recent years. I remember a time when we would board a new client for our QC services, and the only person we spoke with at the lender was the Director of QC. We rarely spoke with the people they reported to. Today, I’m having far more conversations with Chief Risk Officers and other senior leaders—in fact, every one of our engagements is driven by these discussions.

Lenders are also spending more resources on loan quality as well. For example, Fannie Mae only requires that lenders perform reviews on a sample of 10% of their files. Lenders have a choice to perform additional discretionary and target samples in addition to this 10% requirement. Our clients have always done these discretionary reviews, but only a minimal percentage of them. Recently, that’s flipped around, and they’re doing much more, because they realize the importance of maintaining good loan quality.

Still, QC is viewed as a cost center, and is rarely given the same priority as loan production. It’s also a widely outsourced function. But given today’s repurchase risks, lenders must make sure their third-party QC providers are dialed in to areas of growing concern. For instance, I mentioned appraisal quality earlier—this is an area in which QC vendors must leverage their underwriting expertise to make sure appraisal competency is baked into their loan audit services. Yet traditionally, this has not been a strong point among QC providers.

Putting Plans Into Action
Whether lenders outsource QC functions or address them in-house, the ability to identify, handle, and clear loan defects quickly and cost-efficiently will be critical. For many, the best path forward will be leveraging automated QC technologies that are able to identify loan defects through automated audit rules, focus audit staff on these exceptions, leverage portal technology to resolve conditions with the area of responsibility, and allow the portion of loans that are defect-free to sail through the process across pre-close, post close, TRID, HMDA, or pre-funding reviews.

One of the newer options that lenders have are technologies that leverage machine learning-driven automated document processing technologies that can scan thousands of data points across hundreds of pages in loan file documents for defects, enabling the bulk of loans to be quickly cleared while identifying a smaller subset of files that need remediation.

Not all of these technologies are the same, but the superior ones are those that work with a broad set of examples across document types that their machine learning tools are able to “learn” from. Those that have robust APIs can empower lenders to integrate these tools with other systems and pause the process at any time to address errors or other findings as they appear. Lenders can also leverage confidence scores, the degree to which the machine thinks it got it right, to create their own thresholds for certain critical or non-critical document types. For example, this could be used to focus attention on documents that are most likely to contain defects which could result in a repurchase.

Automation is the key for having the ability to automatically and quickly handle and clear defects as they are found, saving days and days of resource time. With more automation, you no longer need to wait for your outsourced QC provider to complete the review of all your post-close loans and only provide you with a static report. With newer technologies, defects can be identified and addressed as they are found in real time, often within hours.

The bottom line, one way or another, is that every lender needs a sound strategy for minimizing risks in what will be a growing purchase market in 2022. That strategy needs to include technology automation that can catch defects early, optimize secondary market profits, and minimize future repurchase risk. The good news is that today’s lenders have many more options to choose from when it comes to defining the strategy that works for them. With so many market variables and challenges affecting their profitability in 2022, the best time to choose the right strategy is now.

About Author: Rick Hall

As SVP Client Services for LoanLogics, Rick Hall has responsibility for managing LoanLogics’ Account Management and Loan Audit Operations teams. His role creates greater alignment and coordination across these teams, enabling better service for clients. Through his leadership and focus on continuous process improvement, he looks for opportunities to optimize productivity in the company’s audit operations, drive new audit best practices, and ensure an exceptional customer experience for clients using LoanLogics technologies and services through Account Management support.

Check Also

Examining Homeownership Gender Gaps

Single women own an estimated 2.64 million more homes than single men nationwide, according to a new LendingTree study, and are more likely to own their homes. Click here to see where the homeownership gap between genders is the largest in the U.S.