One of the more pressing initiatives over the past year relates to the updates to the Uniform Residential Loan Application (URLA) and the Uniform Loan Delivery Dataset (ULDD). For the first time in about 20 years, the URLA is getting a major overhaul and it isn’t just cosmetic. The URLA has a whole new format, with new fields, data points, and requirements. With the initiatives going into effect in February 2020, it’s high time that lenders take a look at how they will impact the industry’s approach to loan quality going forward.
ULDD AND URLA: WHY AND WHY NOW?
Before we jump into the implications of these two initiatives, let’s do a quick review of the ‘whats’ and ‘whys’. The URLA has seen virtually no changes for the past 20 or so years. Why is it being changed now? Good question. In short, it’s because this should have happened a long time ago. This change was inevitable. We can’t go on conducting business he way we did 20 years ago when our industry has changed so dramatically and technology has evolved so much. Think about what personal computing was 20 years ago, and where we were with data collection and analysis. We’re capable of so much more. Our industry needs to start implementing a way to leverage the technological capabilities that are now at our fingertips. And the ULDD and URLA help us do just that.
DON’T GET FOOLED:
THE NEW FORM ISN’T JUST ABOUT WHAT GOES WHERE
From a quality control (QC) and compliance perspective, the big stressors are about data integrity, which, technically speaking, isn’t terribly recent news. The new URLA enhances the type of data collected and the placement of that data within the loan application. That means anytime you deliver a dataset to the GSEs, you need to have that dataset in a specific format, one that’s been laid out and mapped out in a certain way. But more importantly, you need to make sure of the integrity of that data.
This is one thing that lenders regrettably forget. They’re so concerned with how to deliver the new data, that often they forget about the quality of the data itself. This can be a costly mistake, because data quality has been a big industry focus the last couple of years as both Fannie and Freddie have been very critical on data integrity issues. You also have CFPB, which monitors the industry for data integrity issues with HMDA. Now that they’re asking for new data, you can rest assured they’ll be checking to be sure that data is accurate, not merely delivered in the proper format.
A critical focus with the new URLA and the updated ULDD, with their enhanced data sets, is making sure that the data is accurate. A real push has been around how to perfect that data—where the systems are and who’s delivering the data. The LOSs are all getting up-to-date with providers creating testing models with the URLA included. Their goal is to have this fully integrated by sometime this summer with the new URLA being required in February of 2020. Lenders should rest assured that most LOS platforms will cover these updates. However, lenders do need to be prepared for an even deeper data integrity review. That’s how you eliminate the problems that come out of core data.
When it comes to the ULDD, one of the big changes is the capture and reporting of the new HMDA required data fields. So, it’s not only about making sure that data is in the right field or correct format, but even more importantly, making sure the data is accurate. One of the focal points for the ULDD changes is fair lending. If it’s important enough to add a way to collect this data, you can bet that accurately reporting your data is vital. So now that we’ve established that data integrity is of the utmost importance, we need to figure out how lenders can most effectively accomplish that.
THE ULDD/URLA ERA: PREPARE FOR PROACTIVE VS. REACTIVE QC
The way lenders approach QC for ULDD and the new URLA will determine a lot about whether or not they thrive as the industry moves into more robust data collection and analysis.
Can lenders survive with their old processes when the ULDD and URLA go into effect? Sure. But I’m not sure for how long. Lenders that want to thrive and succeed in the new era need to be proactive in assuring data accuracy. If you’re still asking whether you should use technology to verify and validate data, or if you can get away with doing it manually utilizing the “stare and compare” method, the same process as they have for past 20 years, you should prepare for other more proactive lenders to pass you by.
Manual processes are risky and using technology to merely identify data integrity issues is short changing your potential for success. The real question you should be asking is whether the data in your system is accurate, and more importantly, how technology can help assure that data is accurate. Focus on the number of errors found within the data and identify the root cause of the data integrity issues. Then implement corrective action planning to track and correct these data issues once and for all. This will turbocharge your QC and risk management path into true proactive operations, a key component of industry leading lenders.
Using technology that requires manual processes, when a more advanced alternative exists, is risky. Lenders need as systemic a process as possible for data validation. One of the big things we see repeatedly is that lenders have all this data, all this information, but the systems don’t talk to each other. They receive an abundance of the required documents to support a mortgage loan, but often fail to obtain the digital data that created the document. One of the great things about utilizing the abundance of digital data that is at our fingertips is, when technology is utilized to its full extent, this digital data allows systems to talk to each other. Otherwise, at some point along the manufacturing process, somebody needs to go back and review that actual document and verify that the data in that document was actually brought into the system accurately. It’s always best to use technology that enables all systems to communicate and share data. Otherwise, you’re risking errors and even fraud.
The appraisal is a good example of this data vs. document process. The vast majority of loans these days have some form of an appraisal to provide a verifiable opinion of value. Most all appraisals come with the actual XML data file attached. Some tech providers offer the ability to conduct a form of data validation using that data file, and not require a person to handle the appraisal or input data points into the LOS. Without that technology, each data point is typically manually entered into the LOS by a loan processor. And of course, every time you key in a data field, you have an opportunity for error.
If you want to be proactive in improving data integrity, the key is to bring as many data fields as possible into your LOS in a digital format. And if you bring in the data digitally from your actual source document, the appraisal in this case, you’ll know the data fields on that appraisal are 100 percent accurate. Authority documents can be used to help validate data fields from other documents referencing the appraised value. Technology allows lenders to compare and validate all the data within their LOS and other key documents to that authority document with a click of the button
DATA QUALITY IS ONLY BECOMING MORE IMPORTANT
Data quality has always been important, but the ULDD and URLA initiatives are only going to make it more so. It’s important for the lending community to be even more critical of their data integrity. Don’t make the unfortunately all-too- common mistake of believing that delivering data in accordance with the ULDD assures that your data is right. In reality, the ULDD makes sure all their data fields are there, but does not validate the data is accurate.
Financial risks are high when compromising data integrity. Inaccurate data can cause a multitude of problems, from delays in loan delivery, extended warehouse dwell times, and, in many cases, pricing adjustments. In the worst cases, repurchase obligations result. If you think the GSEs won’t say they don’t want to buy a loan because the data you provided doesn’t match what you indicated you were selling, guess again. It’s the worst-case scenario, but it happens. Regulators, like the CFPB, for example, aren’t going away anytime soon. They’re closely scrutinizing HMDA data and have already levied numerous fines, some in the millions, for data validation errors.
So, no matter what changes are coming in May for the ULDD and in early 2020 for the URLA, a critical piece in preparation is to make sure your data is accurate. Technology can relieve a number of loan delivery headaches while bringing a substantial reduction in overhead. You just need to decide whether you will accomplish your goals by reacting to market conditions, or by proactively managing them.
