During a recent panel at the Bipartisan Policy Center 2014 Housing Summit, Darius Bozorgi, CEO at enterprise risk management and collateral valuation service provider Veros, indicated that he felt the technological approach of Fannie Mae, Freddie Mac and the Federal Housing Administration should be extended to the rest of the housing industry.
I needed to know more, so when I caught up with Darius later that afternoon, I asked him a few questions on what exactly he meant. After all, technology companies in the mortgage space, or any other for that matter, aren't exactly known for universally sharing technology.
Darius remains convinced of the idea, so I asked him to give me more on the concept and how it might be implemented.
Here is the result of that conversation in the form of three questions, plus a bonus.
Me: At the BPC 2014 Housing Summit, you talked about the idea of expanding the appraisal portal to other aspects of the mortgage origination process. What other aspects of the loan file do you think can be enhanced by this technology and how do you see it unfolding?
Darius Bozorgi: Certainly the entire loan file can and should benefit from the expansion of the appraisal portal, but we need to be careful not to try and tackle it all at once. Doing so would result in a major burden on industry stakeholders and would take too long to implement. One reason UCDP and UAD have been so successful was the narrow focus.
The effort started with the appraisal file and that has proven to be successful. The next logical expansion of the portal would be the UCD (Uniform Closing Dataset), especially since the GSEs’ proposed final version to standardize HUD-1 was published earlier this year.
Beyond that, the portal should ultimately be expanded across the entire loan file. We can start by looking at all the datasets under the UMDP (Uniform Mortgage Data Program). You've got UCD, ULAD (Uniform Loan Application Dataset), an effort to standardize the 1003 – essentially the complete loan application. At some point following these initial efforts, the logical progression would be to complete the standardization of the servicing dataset. That is being pursued under an initiative termed SDTI (Servicing Data & Technology Initiative).
Ultimately this type of system, if extended to its logical conclusion, should be the system through which loan-level origination data will flow, much like the appraisal. The data from this portal would then be in a position to feed the CSP (Common Securitization Platform). This would complete the circle bringing much needed transparency and repurchase to the system.
Me: If this layer of uniform technology can be applied to all the different components of the origination process, what are the big picture benefits?
DB: Many people cite two major issue categories in reforming housing finance: economic and structural. By structural I mean transparency or accountability and the ability to understand whether origination standards have been met. One of the common themes at this year's BPC Housing Summit was that massive structural issues still exist. Not only do they exist, they are actually the primary issues that require the industry’s attention and those who have an interest in housing finance reform.
In terms of attracting private capital, speakers reported that discussions around raising g-fees and lowering loan limits will not materially move the needle. Instead, panelists argued that more focus needs to be on infrastructure and structural-related issues. That falls squarely into what Veros is doing.
What has been established with the electronic appraisal delivery portals should be spread across the entire secondary market; not just Fannie Mae, not just Freddie Mac, not just FHA. If you want to get true industry standardization, and you want to get private capital flowing more reliably back into the marketplace, everyone needs to understand the facts of the loan during the loan manufacturing process and prior to the loan being closed. This provides originators with the ability to cure deficiencies prior to the loan closing.
That's the primary benefit if you do everything that we are proposing. Doing so establishes a foundation which will provide lenders and issuers with re-purchase certainty, meaning lenders will know with greater confidence that they aren't going to have to repurchase loans. That translates into much greater access to credit because lenders are going to be more comfortable originating. All secondary market stakeholders such as investors and insurers will have more confidence in knowing that industry-wide standards have been met because these standards have been validated. You essentially provide full transparency and 100% due diligence up front at the beginning of the process. It is the foundation of what everybody says they want.
Me: Once the GSEs got appraisals flowing through the portal, how did it impact their operations?
DB: If you look at the progression of what has happened with appraisal processing and risk management from the point at which the portal and its mandate went into effect through to present day, the very first effort was doing something that had never been done before in that space: standardizing the appraisal dataset and then providing a vehicle by which standardized data could be collected.
Thinking about it in phases, the first phase was all about establishing that initial goal: effectively collecting the information while focusing on data integrity – and making sure that what was being collected was good and could be efficiently analyzed by the relevant stakeholders. That is one of the reasons why the portal efforts started off on a joint level with a limited number of fatal edits or hard stops at the beginning. It was important not to create too much of a shock to lenders and, ultimately, to the housing market.
Having fully wrapped their arms around the question, “can the appraisal data be received with a high level of integrity,” we believe the focus then shifted to taking a strong look at this data and what stakeholders want to message around, all with the ultimate goal of driving increased appraisal quality.
What impact has been seen? The number of fatal edits has increased over time and it has been a very calculated and deliberate transition step-by-step. The GSEs retired legacy non-MISMO data formats, they’ve retired PDF extraction, again with an effort of driving increased data quality and integrity, while getting everybody to speak the same language – all of which drives appraisal quality.
Messaging back to the lender has been enhanced, giving the lenders additional insight into what the GSEs are seeing as it relates to various issues of appraisal quality, and that has led to a significantly higher-quality appraisal. Just the fact that the appraisers, AMCs, and lenders know that this is being watched and being messaged upon is a very strong deterrent to bad behavior. So there has been a very positive evolution, and that evolution continues today. With more and more appraisals being collected – closing in on 19 million appraisals to date – the GSEs’ ability to understand risk is increasing dramatically over time. FHA will pursue the same benefits as they continue work on the rollout of the Electronic Appraisal Delivery (EAD) portal. These benefits extend across a very large segment of the mortgage market as it exists today but is only the tip of the iceberg. The same positive results must be extended across the entire loan file and across the entire secondary market in order to achieve stated goals and objectives.
Me, bonus: So what were your takeaways from the BPC Housing Summit?
DB: I thought it was a very well-run event and I was very pleased to be a part of it. There seemed to be good representation of members of Congress, regulators and other government-related stakeholders who have been engaged in the discussion on housing finance reform. And what we have seen historically, six years post-crash, is that while there have been several housing finance reform bills over this time period, none of them have gotten much traction. I would argue that the reason they aren't gaining momentum is because there have been many people talking about the end game without talking about or understanding the means to get there – in essence, putting the cart before the horse.
What BPC Housing Summit has done is provide a forum for DC policy insiders, members of Congress, their staff, agencies, etc. to be exposed to and discuss a number of critical core housing-related issues.
There was a lot of discussion about risk sharing and how to get private capital back into the space. I heard good dialogue around what portions of risk should be shared and what portions should be guaranteed as well as how and when it should be shared. However, while great discussions, they still focus on the ends and somewhat ignore the means. One of the key discussions still missing is how to get there.
Until we get down to the primary risk issues, housing reform will continue to be a challenge. For example, if we want to talk about who, how and when mortgage risk should be shared and guaranteed, don't we first have to ask ourselves whether we are properly identifying risk in the first place? How do you identify and quantify what should be shared if you don't know what 'that' is? It is an important part of the dialogue that needs to have more discussion. We need to spend more time on developing the framework to properly identify risk at origination and ensure that stakeholders - investors, GSEs, etc. - down the line can feel confident that those origination expectations, standards, and requirements were in fact met and adhered to. This is an essential part of the discussion. All in all, I thought it was a very successful event and I applaud the BPC for its efforts. I hope the dialogue continues following this event.