Tuesday, July 16, 2019

UCDP and UAD Five Years Later & Lessons Learned: Better Data, Better Decisions, Better Loan Performance

By Chuck Rumfola
May 2, 2017

Fannie Mae & Freddie Mac

Chuck Rumfola is Senior Vice President of Strategic Initiatives with Veros Real Estate Solutions, Santa Ana, Calif. He is responsible for building strategic relationships within the mortgage industry, enabling Veros to create and implement solutions to reduce lender and borrower risk. Before joining Veros, he worked at Fannie Mae for nearly 20 years and most recently served as vice president of single-family strategic initiatives. In his tenure at Fannie Mae, he held positions in single-family business, capital markets and mortgage operations.

rumfolaFor most in the mortgage business, March 19, 2012 came and went just like any other day.

For me, and the teams at Fannie Mae, Freddie Mac, the Federal Housing Finance Agency and Veros Real Estate Solutions that I had the privilege of working with, it marked the five-year anniversary mandating the Uniform Collateral Delivery Portal (UCDP). More specifically, this is the date when Fannie Mae and Freddie Mac mandated that the Uniform Appraisal Dataset (UAD) be a part of the loan delivery process for each GSE.

This may not sound like a big deal or cause for celebration, but mandating these new appraisal-related requirements has had a significant impact to the mortgage industry. Not only is that impact felt today, it has laid the groundwork for similar work, attempting to have similar results.

On the five-year anniversary, I took a moment to both celebrate and reflect. My thoughts immediately went to "Man, we put a lot of effort into this," but then they oscillated to the impact of what we did, where we are now as an industry, and what this means for the future of housing. We changed the infrastructure for standardizing appraisal data and built the platform for the transmission of that data. Our efforts solved a major issue (valuation) that significantly contributed to the mortgage market crash, set the foundation for modifying the (rep and warrant) GSE business model, and forged the path for other data-related initiatives that will also improve the GSEs' ability to analyze risk.

But, to get to this point, we had to go through some very tough times.

Pre-Market Crash
Before the mortgage collapse, times were very good. Everyone in the mortgage industry was making a lot of money and looked the other way as shady practices evolved. Interest rates were low and everyone was looking to build wealth. Homeownership was well over 70% nationally, hitting all-time highs.

The subprime market was exploding and feeding the private label securities and derivatives markets. There was huge growth in the no- or low-documentation, negative amortizing, option-ARM, and stated-documentation loans. The GSEs had aggressive affordable housing goals, and were losing market share to the Private Label Securities market. FHA's market share was in the low single digits. Cash-out refis were a big part of the market.

The credit quality of the borrowers had declined. Loan-to-Value (LTV) programs enabled no down payments and LTVs went up to 100%. Appraisals were overstated, and there was sharp home price appreciation in many areas of the country. In fact, the country had not seen national price depreciation since the Great Depression. Finally, the GSEs relied heavily on the rep and warrant policy for protection. The rep and warrant business model had worked in the past, but was soon to be tested in a stressed market.

Appraisals were a big part of the pre-crash puzzle. Money was easy before the crash; it was easy to buy a house, refi an existing loan and even take out large sums of money based on the perceived equity in the home. We all know what happens here. Hollywood has even brought the spotlight onto us with blockbusters such as "The Big Short," depicting the global financial crisis and inevitable market collapse.

Most people think that the market collapse was based on borrowers getting into homes they couldn't afford. This is true; however, the reality is that cash-out refis, combined with inflated appraisals, did a tremendous amount of damage.

For simple math, take a borrower who had a house worth $100K and a $60K loan balance. Now assume this borrower wanted to take out their equity in the home. The reality pre-crash is that the appraiser enabled an inflated value (well over $100K), the lender was more than willing to make the loan to just about any credit quality borrower, and the borrower enjoyed taking out as much cash as possible. When this happened, and it happened a lot, the borrower ended up with a lump sum of cash, but also an asset that was underwater. Tack on depreciating home values and you have millions of borrowers that owe more on their house than what it was realistically worth.
And, add on a life event such as death, divorce, job loss, or even an interest rate adjustment on their loan, and the borrower was left without any options to refi or move to improve their financial position. As we saw, too many borrowers just stopped making their house payments.

The Crash
Markets have a way of correcting swiftly; the mortgage market found out the hard way. Home prices fell nationally. Many borrowers were underwater and couldn't refi out of their "toxic" mortgage. The subprime market dried up overnight. The PLS market disappeared. Major firms went out of business or had to be merged with other companies. We entered the Great Recession. The GSEs frequency of loan default and severity of credit losses forced them into conservatorship. They borrowed nearly $200 billion from the Treasury to make good on their mortgage-backed securities obligations. FHA's market share soared to more than 30% because they were the closest thing to subprime. Now FHA had an issue as their reserves fell below mandatory thresholds. Mortgage credit was tighter than it ever was before. The GSEs had to figure out how to stop the bleeding.

The New Market
For the GSEs, which had to restore confidence in the market, the post-crash market became vastly different than pre-crash. First, the GSEs went back to basics to control the "frequency" of default. They tightened their credit policies requiring higher down payments, full documentation loans and higher credit quality borrowers. Second, they introduced a new infrastructure to ensure accurate collateral valuation with the UAD and UCDP. This helped them control the "severity" of losses. All this was an attempt to modify their business model. Yes, the GSEs would still rely on reps and warrants, but now they wanted to know more about the loan pre-purchase or at least be in a position to identify issues immediately after purchase.

By the way, the GSE's post-crash book of business is the best they've ever had.

The Results
The government entities (GSEs, FHA, VA) have nearly 80% market share. Sound credit policies have returned, protecting borrowers, lenders and investors. Yes, there is still work to be done, but it is hard to argue we're not in a much better place today.

On the collateral valuation front, we've standardized data, built the portal to collect the data and issued policies mandating delivery of the data to Fannie, Freddie and FHA. These efforts have enabled drastic improvements in risk analysis. We've seen nearly 40 million appraisals flow through UCDP, resulting in a huge database for risk modeling. There are 100% automated collateral review, targeted appraisal reviews and proprietary risk analysis in place to protect the GSE's and loans being sold in the secondary market.

The GSEs created their own proprietary tools in the appraisal space, including Fannie Mae's Collateral Underwriter and Freddie Mac's Loan Collateral Advisor to perform appraisal quality reviews crucial in managing collateral risk. Fannie has Day 1 certainty and Freddie will soon follow suit. And, FHA, after implementing the Electronic Appraisal Delivery portal, will soon be in a position to emulate what the GSEs have accomplished.

Unlike before, where facts didn't matter, competition is now based on who is better at managing the risk associated with the facts of the loans. This was always the agreement between Fannie and Freddie when developing UAD and UCDP.

What's Next
What's next, given the success of UAD and UCDP? The GSEs are following a similar recipe with the Uniform Closing Data set. The UCD is targeted to be mandated Sept. 25 to ensure data used for underwriting the loan match the loan actually purchased by the GSEs. The closing dataset has been standardized. Then, similarly to UAD and UCDP, the GSEs will have data necessary for risk management, lender messaging and possible policy changes in the future.

The Uniform Loan Application Dataset will come in around 2020. The process of standardizing data has already begun. You can bet loan application data will be used by the GSEs for risk management or they wouldn't have it on their data roadmap. The snowball keeps getting bigger and bigger. And, all of this derived from the appraisal-related activity that was mandated five years ago.

After that? Who knows. What we do know is that the industry learned the hard way that sound credit policies and standardized data matter. Because of this, I'd like to think that we will not repeat some of the mistakes of the past. All signs seem to indicate that the data revolution will continue. Why not look at the entire real estate process where we can make improvements through data, technology and transparency?

UAD and UCDP marked the beginning of tighter regulations, new sound quality control processes, and improved risk analysis for the industry. So, is there a cause for celebration? Absolutely. We survived, and not only that, we're all better for it. More importantly, real lives and real borrowers are better for it too.

(Views expressed in this article do not necessarily reflect policy of the Mortgage Bankers Association, nor does it connote an endorsement of a specific company, product or service. MBA Insights welcomes your submissions. Inquiries can be sent to Mike Sorohan, editor, at

Share this article