Quality Control Systems for Loan Servicing - The New Paradigm

THE CHANGING LANDSCAPE OF SERVICING QUALITY

The landscape for Loan Servicing Quality Control has changed dramatically in the last few years. Prior to the mortgage crisis quality control was run by in-house quality departments of servicers with a pure a process improvement perspective. In contrast the role of servicing quality control departments today is to be the “last line of defense” for the servicer entrusted with detecting and mitigating a myriad of downstream risk.

The servicer’s actions are now being observed and analyzed in by a number of stakeholders and observers such as investors (both GSE and non GSE), federal and state regulators, borrowers and borrower advocacy groups and media. This necessitates the need for stricter standards and focus on quality control on all aspects of servicing with special emphasis on default servicing.

GSE’s directed by FHFA’s Servicing Alignment Initiative have created the FNMA STAR program and FHLMC Servicing Success program as servicer performance management directives. A key directive of these programs is the servicer’s responsibility to develop and execute an effective internal quality assurance program. Private investors are following suit with their own set of quality guidelines and scorecards which will many cases be expanded and/or modified versions of the GSE directives.

It is apparent that the same investors that previously took a somewhat passive approach to loan servicing performance and quality control, letting servicers to self-enforce terms of the Loan Servicing Agreement, are now interested in taking a more active approach towards the oversight of their assets. Investors are demanding a real time view of the performance of their portfolio with loan level detail rather than be dependent on monthly investor reports at the portfolio level. They would like to monitor quality of servicing and be able sanction non-compliant servicers with fines or in extreme scenarios terminating the servicing contracts for cause. Under this climate of increased scrutiny the servicer should attempt to create a culture of “doing it right the first time” by paying careful attention to quality and ensuring nothing falls through the cracks that can cause undue downstream risk in the short or long term.


CURRENT CHALLENGES FOR SERVICING QUALITY CONTROL SYSTEMS
Let’s discuss a few of the process and technology barriers that the new breed of Quality Control systems has to overcome.

Process Challenges
It seems these days that some new regulation, restriction, compliance or consent order is being placed on loan servicers on almost a daily basis. Rapid process changes have to be implemented by servicers along with corresponding changes in quality control systems in order to respond. HAMP and Loan Modification Audits, Robo-Signing Audits, Foreclosure Consent Order Audits, Single Point of Contact Audits, etc. are just a few of the new audits required by servicers.

In addition, recently added processes are creating numerous sources of errors and omissions that need to be detected and mitigated. If we look at the loan modification process, servicers are being challenged to act more as originators to run a production pipeline of modifications from borrower contact through closing. A large number of errors are cropping up due to immature processes and improper interpretation of guidelines. According to MHA, the biggest areas of problem in HAMP Loan Modifications are in income and expense miscalculations, missing documents, miscalculated DTI ratios, missing of potential signs of fraud, etc.

Servicing Quality Control systems are challenged to remain relevant as these new process changes are implemented and to provide actionable information on the root causes of errors and exceptions in the underlying processes. An adaptable and flexible system that does not need onerous IT intervention and is under the full control of the quality department is essential.

Technology Challenges
The top 3 loan servicing systems used by the industry today (with more than 80% market share) are monolithic main frame based applications written more than 30 years ago (an eternity in technology terms). These systems are merely green screen (sometimes with a graphical user interface slapped on top) front-ends to gigantic mainframe databases. These core systems lack key features such as workflow management, rules engine, extensibility, ease of data access, etc. Key workflow steps are captured in unstructured notes that cloud the chronology of events and prevent the creation of a clear chain of accountability.

To add to these problems, a plethora of loosely integrated, bolt-on solutions have been created primarily in the default arena such as loan modification, REO, short sale, foreclosure and bankruptcy management systems to make up for the shortcomings of the existing servicing systems.

Getting a holistic view of servicing quality is extremely challenging in this environment of patchwork systems to construct a coherent “quality story” across the organization. It is not surprising therefore the quality control at most servicers today are mostly run on spreadsheets requiring a lot of manual data integration to get at actionable results.


SERVICING QUALITY CONTROL SYSTEMS – THE NEW PARADIGM

It is obvious that a new class of Quality Control Systems is needed to tackle the challenges facing servicers today. The key features these systems need to provide are:
a. Flexible Integration – For servicing quality control to change from a batch process to a real time review of production performance it is imperative that quality control systems accept data from core systems easily. Web service based integrations where the core systems either push or the quality control system pulls data at specific production thresholds or time intervals are ideal. If web services are not supported by the core systems, database level transfers might be done for a less elegant but effective integration. If file import based techniques are used, data should be imported into the quality control system at regular intervals rather than in large data drops. Additionally the system should allow ease of integration with third party systems such as credit reporting agencies, valuation firms, fraud checks, etc. to be able to re-verify input data.

b. Template Based – To keep pace with rapid changes, the quality system needs to empower subject matter experts and quality managers to rapidly create new audit checklists, sampling strategies, reports and scorecards with minimal help from IT. Quality managers can start from a pre-defined library of templates and then be able to quickly enhance or modify it for their needs. The quality system should be a delivery platform onto which subject matter experts can supply “Add On Packs” for various areas of the business. With a template based system a servicer can rapidly respond to changing audit needs without having to acquire or learn yet another quality system when the next new audit need arises.

c. Agent Scorecards and Accountability – Along with focusing on the loan that is being audited, important information about the servicing agent(s) who performed the particular function needs to be tracked. If a quality control effort is unable to pinpoint the cause of a failure to an individual or group of individuals who performed the function, much of the benefit of finding the quality issue is lost. Agents who are being audited also need to be provided real-time scorecards on their performance and required to undertake remedial training to allow them to prevent committing such errors in future. If the source of the errors is found to be due to ambiguous or incorrect policies and procedures such issues should be corrected expeditiously.

d. Auditor Management - Auditors performing the Audits should be similarly accountable for the audits they have performed through detailed scorecards and reports. Auditors are typically asked to find issues in the work of operational personnel that possibly have far more experience in performing the particular function being audited. To maintain the value of the audit process it is thus important that auditors are assigned work based on clearly identified skill sets where they have demonstrated knowledge and experience. Auditor workload should be managed proactively based on auditor tenure, skillsets, workload and other dynamically calculated factors based on historical data stored in the quality system.

e. Risk Management - The purpose of a Quality Control Audit is to mitigate risks by ensuring adherence to documented Policies and Procedures. It is important therefore that each audit checklist item reference the appropriate risk item and its corresponding Policies and Procedure entry that mitigate the risk. This reference is not only used by an auditor to perform the audit correctly but also used by a regulator reviewing the results of the audit in future to certify that the audit was performed per documented guidelines.
f. Workflow Based System – A typical audit workflow should allow the steps of Audit, Audit QC, Agent Challenge, Auditor Response and Arbitration. The quality system should have a flexible workflow engine that manages the audit queue and routes work to the participants of the system. Workflow steps should be timed so there isn’t a scenario of a never ending audit waiting on responses from Agents or Auditors and the audit can be “closed” and definitive results obtained from the exercise.

g. Actionable Intelligence – Just performing an audit is a worthless exercise unless the results of the audit can be utilized to remediate the issues that were identified. A flexible and user driven reporting engine is important but the emphasis should be on drilling down to the root cause of an error or exception. Dashboards, Data Mining and Online Analytical Processing (OLAP) tools need to be utilized by the system to provide a drill-up or drill down view of the Root Cause and the Source of it to the key stakeholders of the organization. As an example, the CEO of a servicing organization might just be interested in knowing error rates are trending up across the organization, the Servicing Manager in knowing that Loan Modification of HAMP loans are showing increased issues and the Loan Modification Underwriting Manager in knowing that most errors are arising from the calculation of income for self employed borrowers.

CONCLUSION
Servicing as we know it is changing dramatically and new and innovative ways of managing servicing quality enforced through the use of technology are needed to allow servicers to deliver a reliable and consistent service to its borrowers, investors and other stakeholders.


Author:
Souren Sarkar is CEO of Veritiq (www.veritiq.com) a provider of Quality Control and Process Management software for the mortgage industry.
Email: souren.sarkar@veritiq.com
Phone: 786-348-5595

Definition of Loan Servicing is Changing Dramatically

Excerpt of a recent interview I gave to Mortgage Newslink Magazine - http://www.mortgagebankers.org/tools/FullStory.aspx?ArticleId=20978


I think the definition of a servicer is changing dramatically in this day and age. First of all, people are realizing what a servicer does, and it has become a public persona. So we hear a lot about reputation risk or things like that which was not so prevalent in the industry. The servicer is in the crosshairs of a lot of regulatory and compliance kind of issues which has changed the industry dramatically, and we also see the servicer taking on new roles.

The servicer is considered almost like a new originator. When you do a loan modification, you are negotiating rates and trying to look at DTI ratios and things like that which they have never done before. We also see the changing role of the servicer as an asset manager looking at property portfolios, thinking about landscaping, property preservation and things like that which was always there in servicing, but not to the extent that it is today.

With the increased load of non-performing assets and regulatory and public oversight, the servicer is in the crosshairs of a lot of these new roles but they’re locked into servicing fees that they had gotten into a long time back. There has to be a dramatic change in how servicers are paid for the work that they do. There has to be a gradation in the scale, and that’s what in some of the new propositions it is being proposed that all servicers need to get paid commensurate to the amount of work that needs to be done on that loan and not in a general portfolio basis. We almost have to go down to the loan level and say, how much work will this loan need? Well, this is a sub-performing loan, the cost here will be significantly higher and there has to be a more interactive approach. As much as the servicers want to comply, they also cannot lose their shirt on it. ......

To RFP or Not To RFP That Is The Question

When your house is burning down do you send out an RFP to the all the fire stations requesting data on their staffing levels, quality of equipment, fleet management, capability matrix of firefighter skills, etc. No, you pickup the phone and dial 911 to dispatch fire fighters from your nearest fire station - you need action not perfection. The same analogy can be made for projects that are sent out to RFP versus those that are not.

My observations of this befuddling process of participating in enterprise RFP's has been that they are:

• A future looking request for an anticipated future need
  
• A tool to validate pricing and process for a pre-selected vendor by opening the proposal out to bid to a broader vendor base
  
• An attempt to validate/ benchmark the process flows of internal teams by investigating industry best practices
  
• A means for sourcing groups to keep busy during down times when there is not an internal need for sourcing - giving them a reason to exist
  
• A strategic initiative by sourcing groups to learn more about a growing trend with competitors such as offshoring, to evaluate the applicability of the trend to their organization
  

  So what then is the magic formula for winning RFP's. There is none but here are a few suggestions:
  

• Know the stakeholder(s): Key relationships with stake holders is the single most important factor in winning RFP's. As stated in the problems above, many if not most RFP's are "inside" jobs where the stakeholders have already decided or are leaning towards favorites. You want to be in that in crowd. Relationships have to setup prior the RFP stage independent of Sourcing intervention. Observe industry trends and shortlist the stakeholders of the future, then systematically reach out and network with them and crystallize relationships long before the RFP is inked. The holy grail is when an RFP is written with your key capabilities in the back of the mind of the stakeholders.
  
• Establish subject matter expertise: If you have not had the opportunity to establish key stakeholder relationships, the RFP is also you/ your company's resume. Establishing subject matter expertise and thoroughness/ completeness of responses is of paramount importance. During the question period ask relevant questions geared towards the key pain points of the customer not operational or technical details that can be figured out after you win the RFP.
  
• Ask key discovery questions: Don't assume key questions that provide insight into the playing field for the RFP will not be answered. ask them anyways. Questions like how many participants are there in the RFP, how many will be selected, who are the top participants, what are they key qualifications for the winning bidders, etc. You will be surprised how much clients will be ready to answer if only you asked.
  
• Stay away from RFP's that smell that RFI's: if the RFP is too broadly defined without much details asking questions like "state your key capabilities in << fill in the blank>> , gracefully decline or have your analysts provide broad answers to the RFP enough to satiate the client. Move on..
  

  So to conclude, revisiting the fire in the house analogy, the best fires to tend to are those that are burning right now. The customer is ready for a purchase and fast. For future fires (business) invest time and effort now on identifying the trends that will cause fires in future and establish relationships with that user base.