How much should we pay for quality? Is focusing on quality the right choice when the cost exceeds the return? These questions are being asked a lot these days but very few if any answers are forthcoming. Most companies seem to be so concerned about the consequences of any problem that does or might exist in any one loan, they are doubling or even tripling the number of file reviews being done. Is the cost of these additional reviews actually worth it?
Because of the quality failures that were at the very heart of the Great Recession, we are now experiencing the repercussions from Fannie Mae, Freddie Mac and FHA as all have come out with new Quality Control requirements that are designed to produce loans with zero defects. In addition, the new regulations and standards that CFPB have placed on both servicing and production require strict adherence to the standards established for the quality of these operations. To the industry this has meant adding additional reviews, more frequent reviews and more rework on loans. And still the risk of repurchase or consumer action remains. All of these activities result in more costs but so far we haven’t seen any financial benefit for these efforts. Is this normal? Is this what we should expect? Or should we, as one frustrated manager put it, start making buses instead.
Stefen Heinloth, in the February 2000 edition of The Quality Digest asked a similar question by querying “Are quality related efforts worth their cost?” In answer to this question he described two things that have to exist for the answer to be true. One is that quality has to be measurable. Only then can a company determine if what they are spending to meet standards, theirs or others, is really worth it. The second thing is that there must be a cause and effect relationship between quality and financial results. He goes on to say he sees that “companies are taking a return on quality approach, viewing quality as an investment and holding quality efforts accountable for bottom-line results.” Of course this was for other industries such as manufacturing, not ours.
The concept of value.
Unfortunately Mr. Heinloth’s statement was focused on the application of quality management techniques and concepts that this industry has yet to learn. Despite the labeling of ideas and dictates as “manufacturing quality’, what we have been directed to do or what we have scared ourselves into doing is not in any way shape or form, Quality Management. In order to better understand what they are and how we, as an industry can implement them in a manner that allows us to value quality improvements several of these concepts need further explanation.
One mainstay of Quality Management is the ability to gain value by the improvement of the product/service being provided. However, improvement requires measuring the process and finding out what is not working right; what requires rework and what process failures result in products that have to be “scraped”. In order to do that we must be able to measure our processes. Edwards Deming stated that if you can’t measure it, you can’t improve it. So if we are going to identify what parts of our processes are causing us problems we have to measure them. Of course, the quality control process is supposed to be doing that, right? Unfortunately there are numerous flaws in the programs that Fannie Mae, Freddie Mac and FHA require. But more importantly, when lenders are dissatisfied or question the results of their QC programs they fail to make any changes that would be beneficial. One reason is that most don’t know what is wrong or how to fix them. So here’s a quick guide to QC that anyone can follow.
1.) QC should be measuring where you have a risk in your process. For example, if a processor or underwriter doesn’t calculate the borrower’s income correctly, that is a risk to the process. In addition, if they enter the wrong amount of income into an automated underwriting system, that is also a risk. These process variance or “defects” should be measured for every loan yet there are far too few measurements that tell the number of times this type of process failure occurs.
2.) Process variances cannot be labeled by some subjective measure of precision. In other words there are no “critical” defects or “minor” defects. A defect is a defect. Since it is only when this defect has proven to cause a default can it begin to be labeled as a critical issue. As we all know there are very few, if any defects that impact a loan to this degree.
3.) Random events do occur. We seemed to have ignored this concept. Way too many underwriters and production staff spend time defending a defect in a loan when it is simply a ransom mistake. Way too much time and effort is spent on reworking a loan file when QC identifies an error. This is particularly true of “curable” items. Over and over QC reports identify defects rated as curable. What does this even mean? There is no lender in the industry whose process says, generate this document but DON’T put it in the file so that when we discover an error we can “cure” it. It sounds like some type of disease.
In reality failing to put documents in the file is part of the process and if it doesn’t get done when the process says it should, it is a defect. Spending time going back to find the document or creating another one, is plain and simple just unnecessary rework and rework costs money.
4.) Which brings us to another basic issue. How do we know if the defects identified in a review are really worth spending money on to fix? Here is where statistical knowledge is at its best. Since we are not doing a review of all loans, we need to know if the frequency of a defect is really a problem. Identifying this with just one review is probably tough, but if looking at multiple reviews, QC staff should be able to tell management the probability that the defect found is not random.
5.) Since the majority, if not all QC staff do not have the statistical knowledge to achieve this, another way for a lender to determine if something in the process needs fixed, is to look at the industry as a whole. Having a benchmarking tool that can be used to compare rates of occurrence on a specific issue within the industry to your defect rate would immediately tell you if you have some type of systemic problem that is driving this excess number of mistakes. If your percentage of errors is much larger then it would normally be, there is a good bet that something needs to be fixed.
6.) Finally there has to be an established cause and effect relationship between the defect and the risks inherent in the product. If there is none, then why are you spending time and money testing for it, “curing” it or trying to redo the process? Well of course you say, that only makes sense. So then, what are the defects that cause an unacceptable level of risk in a mortgage loan product or a loan servicing outcome? Unfortunately, we don’t know, or at least most of us don’t know. Sure we say LTV is a risk or DTI. And what about insufficient reserves? But where are the statistics that validate that?
While the industry has lots of data on performance and some of these static elements we attribute to poor performance, we have no way of knowing if the data we have used is accurate. In reality the only work done that is publically available are the results of investors and others whose interest is selling the product not improving the process.
Ultimately we do not have a standardized way to test our processes or relate them to product failures. We have failed to develop any meaningful relationship between process failures and risk and we spend all of our time and money on rework and unproven changes. No wonder recent MBA data identified that total loan production expenses increased to $6,932 per loan in the second quarter of 2014 from $5,818 per loan in the same quarter the previous year. This number is only going to get worse as QC staff increase the number of reviews for TRID and other regulations and agencies changes.
Where is the value?
Based on all the issues identified above, it is hard to comprehend that the industry has any idea of the value of their operational processes. Since we haven’t correlated defects to repurchases or rework let alone done any real work of identifying the causal relationships of these errors, it is impossible to know the difference between the cost of a product produced by a satisfactory process or the value of a consumer/investor that is satisfied with a successful servicing program. However, in order to determine that we do in fact know some of these numbers, let’s discuss the return on investing in quality improvements.
The return on the investment for any process improvement is calculated as the ratio of two financial estimates:
ROI=Net returns from improvement actions/ Investments in improvement actions. The numerator and denominator are defined as follows:
- Net returns from improvement actions is the financial gains from the implementation of the improved actions, which are generated by new changes in quality, efficiency and utilization of services, or in payment for those services.
- Investment in improvement actions are the costs of developing and operating the improvement actions.
Looking at an ROI calculation from this perspective it is fairly straightforward as to what the quality, efficiency and utilization of resources are involved in producing loans. If a produced loan does not follow the guidelines or processes in place to ensure a quality loan, then it is defective. This can mean that if the defect(s) are discovered by an investor, the loan may be rejected by the investor. If discovered prior to sending the loan to the investor it may have to be placed in portfolio, may have to be held on a warehouse line for an excessive amount of time increasing costs or sold as a “scratch and dent” loan. If problems are discovered during the process or by QC, the problem will have to be “cured” which involves rework by staff or even asking the consumer to supply additional information or replace documentation that was provided earlier. Again, this is more cost. And we can’t forget to include the time that production resources are dedicated to fixing something that has already been through the process rather than on generating additional income by producing additional loans.
However, we can develop a hypothetical example of how the ROI on quality can be determined. For our purposes let’s say that a review of the defects identified above cost around $2,000 per loan on average. Using the MBA production cost of $6,932 per loan we would add these additional costs of $2,000 per loan which raises the overall average cost production $7435. The cost of producing 100 loans per month is therefore $743,500.
Once we identify the operational controls that failed in the origination process that increased these production costs, we know what we have to fix. For our purposes we hypothesis that a change to the system will prevent the loan from moving to the next stage of production when these errors occur and training of the processors should correct the error. The cost of these fixes is $25,000. The change is then implemented.
Once implemented, the QC staff measures the results. They find that the problem has been eliminated by these changes. In addition, the efficiency created by the revised process flow has taken an additional $10.00 off the cost of producing a loan. Therefore we have achieved a net return of $2,010 from these improvement actions. This change reduces our cost per loan to $5,425. Our monthly production cost, on average for 100 loans is therefore $582,500 or a difference of $ 2,010. Using the ROI formula for quality improvements our return is 6.44%.
While this example is somewhat simplistic in nature, it demonstrates that if individual companies were to understand and apply the concepts of quality management to their organizations, the value would be clear. So instead of complaining about the costs of quality control and the perceived cost of reviewing an excessive number of loans to ensure that they meet investor and regulator QC standards, lenders would be wise to redefine quality and begin to implement these previously validated, achievable returns on the investments they make.
About The Author
Rebecca Walzak is a 32 year veteran and Industry Expert on Operational Risk Management and Organizational Control. She is a leader in developing Operational and Control automated assessments for lenders, rating agencies and investors. Walzak has expert knowledge in all areas of the mortgage industry including production, servicing and secondary.
Barbara Perino is a Certified Professional Co-Active Coach guiding her clients who are executive leaders and their staff. Barbara has been trained through The Coach Training Institute (CTI) located in San Rafael, CA. She completed a Coaching Certification Program through CTI and the International Coaching Federation (ICF). Prior to becoming a coach, Barbara was a 16-year veteran of the residential mortgage industry.