PERSON OF THE WEEK: Becky Walzak Talks Quality Control, Dirty Data And Industry Standards

Clean or dirty – what's your data? That's the question confronting many lenders and servicers today, as compliance concerns mount and loan-file integrity undergoes greater scrutiny than seen previously.

This week, MortgageOrb catches up with American Society for Quality Mortgage Expert Becky Walzak, CEO of Walzak Risk Analysis LLC, to discuss how mortgage professionals can ensure their data quality is unassailable.

Q: The deadline for mandatory compliance with the Fair and Accurate Credit Transactions Act's Red Flag Guidelines is fast approaching – Nov. 1. What advice do you have for mortgage banking professionals as they put the final touches on the development of their Red Flag programs and ready them for implementation?

Becky Walzak: There are several areas I would focus on. The first is the cohesiveness of the program. While the requirements cross many areas, if the overall implementation plan is not coordinated across all of these areas, there may be gaps that have been missed. I suggest that if it has not already been done, a process map be developed and accountability for each area be assigned to ensure that every action is coordinated with every other.

The second is not to forget quality control (QC). The recommended method for implementing processes is to plan, do, check and then act. If you have established how all aspects of the program are to be implemented, run a few test loans through the process and let QC check to see if it works. Once fully implemented, QC should also include these activities in their ongoing QC plan.

For those individuals still struggling, the upcoming Mortgage Bankers Association's Quality Assurance Conference will include ideas from various lenders on how to implement and test for these Red Flag Guidelines.

Q: In the spring, you lobbied for the inclusion of a lender scoring system in a mortgage rescue bill – the idea being that such a system would demand accountability on the part of lenders for the loans they underwrite. What is that proposed amendment's current status?

Walzak: I am pleased to tell you that there are two requirements in the recently passed housing legislation that, once implemented, will serve as the foundation for such a scoring system. Once the scoring system is in place, individual lenders can have their loans scored for overall process risk and compared to other lenders in the industry.

This comparative type of analysis will provide a means to allow consumers, regulators and investors to identify which loans carry the lowest risk and which lenders are able to control their processes so that consumers can pick the best lender to use.

Q: Are the extant sets of guidelines and regulations that apply to the mortgage industry sufficient or, aforementioned scoring system aside, do you encourage stronger standards be put in place?

Walzak: I do not believe that external standards are the answer. All external standards do is create jobs for auditors. Mortgage lenders need to set their own internal standards as to how they will produce quality loans, put those processes in place and make sure those processes are being followed. This is how other industries have improved the overall quality of their products.

However, one aspect of this is missing for lenders. Other industries expect – and typically receive – higher value for a better quality product. Today, there is nothing that rewards a lender for ensuring the quality of their loans. They are dumped in a pool with every other loan and paid the same. So why should they be focused on anything but production at the lowest cost?

If we really want to avoid onerous external standards, the industry needs to have a method for measuring loan quality that can be used to effectively quantify the value added for doing a good job so that "good quality" lenders will be paid more for their loans and poor-quality lenders will be paid less.

A new report from the Mortgage Asset Research Institute found that the incidence of mortgage fraud in the U.S. increased by 42% in the first quarter of 2008 from a year ago. Combined with the Miami Herald's recent investigative news series focusing on predatory lending, and the innumerable antifraud efforts touted by law enforcement agencies, the focus on mortgage fraud has never been stronger.

Walzak: Fortunately, that is true. The crisis we are now addressing is partly the result of the inclusion of individuals into the industry whose whole purpose was to steal money from lenders. Just because they didn't use a gun, doesn't mean they weren't crooks. Florida recently found over 10,000 individuals licensed by the state to be convicted felons. These individuals are estimated to have stolen over $80 million dollars in Florida alone.

Unfortunately, these individuals do not account for all of the fraud. A recent report found that while the incidences of external fraud have decreased, internal fraud has increased. This means that staff members are responsible for a large portion of the fraud that is now occurring. And while their motivation is more likely fraud for income, it does not lessen the negative impact this has on the industry.

Q: But in your podcast, LoanWrap, you note the elimination of "bad apples" from the industry doesn't necessarily equate to the elimination of "dirty data." Is there an imbalance in the attention being paid to those who originate bad loans and the inclusion of just plain bad data itself?

Absolutely. Loans from the largest to the smallest lender that are run through our data analysis product show numerous loans with missing information, incorrectly formatted information and just plain illogical data. For example, for one lender who ran 1,000 loans through the program, we had 17,000 data mistakes.

They include four-digit zip codes, property addresses with no house numbers, loan amounts that we formatted as millions instead of thousands, and fields such as loan purpose completed with information such as "condominium." This doesn't even include such mistakes as inputting a credit score of 496 as 694. Until we get data quality under control, we cannot assume bad data means fraudulent data.

What are the biggest challenges going forward as the industry attempts to put itself back on track?

Walzak: I think there are a number of challenges, but I believe that the most critical is to gain back the confidence we lost from the secondary market, consumers and legislators. A friend was telling me the other day that he went to his cousin's wedding recently and no one in the family wanted to introduce him to the other guests since he is in the mortgage industry. That's about as low as it gets.
So how do we get it back? I think the best way to provide proof that we have tackled the problems that led to the current crisis, which include poor data quality; processing, underwriting and closing mistakes; the lack of data transparency in loan pools; incomplete default predictive models; and greed.

Although we can't do much about the greed, we can tackle these other issues. Reviewing these issues results in one overriding problem, and that is the operational processes and controls used to produce loans. Only when the data is accurate, the calculations are completed correctly and the guidelines are applied as required can we be certain that we have a quality loan that can be correctly priced in the market.

Lenders have to be accountable for ensuring that the loans they produce meet these standards. Once that happens, default models can become much more accurate, pooled loan data can become more transparent and the effectiveness of credit policies can be measured.

Unfortunately, we have always attempted to do this one loan at a time, and that just doesn't work. What lenders need is a way to measure how effectively the processes and procedures they put in place are working. This means a measure that can score all loans so that both the purchaser and the consumer can determine how well the lender does its job and the lender can identify where his process is weak or failing.

Once we can demonstrate that lenders have controls in place to ensure the quality of the product, we will once again begin to gain back the confidence we need to revitalize the industry.


Please enter your comment!
Please enter your name here