By John Levonick
The quality revolution: Leveraging technology is one sure way to restore faith in the private-label securitization market.
The industry is abuzz over which industry participants are best suited to shoulder the appropriate level of risk to ensure quality loan products are originated in a responsible manner without limiting access to credit. There are many new ideas that attempt to restore faith in the private-label securitization market. They include promises of new regulatory bodies, collateral data transparency and credit-risk retention; however, only rarely is there mention of quality loan origination practices and loan-quality performance metrics.
The ability to track the performance of a loan should create a world where a lender's success is driven by its ability to consistently price, underwrite and deliver quality products to the secondary market. In an ideal competitive lending environment--where production quantity does not inhibit loan quality--mortgage lenders should be able to gain a competitive advantage with investors through quality origination and performance metrics obtained from a smart rules-based loan production platform. Such a platform would provide front-end point-of-sale, loan pricing, processing, underwriting, closing and commitment fulfillment.
Leveraging the proper technology can significantly reduce institutional and loan-level risk by ensuring compliance with loan credit, collateral guidelines and regulatory compliance requirements while reducing overhead through the operational efficiencies a smart platform can create.
The need for a fully transparent origination process that produces quality loans should motivate those that retain risk either through "skin in the game" or "first-loss class retention" to create incentives that motivate the move to the necessary origination technology. Such technology will produce the level of quality investors demand. The bottom line is this: For the private-label securitization market to flourish, the lenders selling new originations into the secondary market must be held responsible for the quality and overall performance of loans they are creating.
Lenders must be held to a higher standard and ensure that each loan that passes through their origination platform meets all the investor and regulatory, credit, collateral and compliance requirements while delivering fully transparent loan data to all of the players in the delivery of a private-label securitization.
Data controls to reduce institutional risk
For established lenders to effectively compete in today's lending environment, their first step toward quality is to shed antiquated and inefficient origination platforms and processes. These loan origination systems (LOSes) are comprised of a number of different manual and automated processes that contain various workarounds, gaps and temporary fixes. These often lead to manual override or system bypass capabilities by individuals that have a significant financial interest in the loan closing versus an interest in the overall quality of the loan being originated.
Technology available today can provide a solution that contains rules-based underwriting tools with configured workflows that "lock down" data to ensure the necessary credit, collateral and compliance requirements are met as the loan progresses through the origination process.
A rules-based underwriting system is a preconfigured underwriting engine that has built-in calculations and threshold validations to enforce the specific investor or product-specific guideline requirements. This system automates the traditionally manual processes associated with loan underwriting. It transforms worksheets into objective electronic calculations, memorialized in an electronic format that will be archived with the originator in anticipation of electronic audit delivery and passed on to the ultimate Investor as part of the entire closed-loan file.
This rules-based system is locked down to ensure that no single individual lacking proper administrative authority may alter or amend any credit, collateral or compliance data in an attempt at subterfuge. This locked-down process restricts any deviation from, or violation of, guideline requirements (intentional or unintentional) as the loan progresses through the underwriting process, delivering a loan that has met all specific guideline elements.
Workflows are automated operational task progressions that not only drive the route of the loan through the origination process but specifically permit and restrict user access to particular data fields based on the lender's workflow. Further, these tools automate the ordering of service providers (i.e., appraisal, title), automatically task specific users with single or multiple responsibilities and, most notably, escalate credit, collateral and/or compliance issues for review when specific tolerances fall within a preset risk range.
Workflows may provide for secure uploads directly from borrowers or third parties (i.e., banks, employers), thus reducing manual processes and eliminating errors, lost documents and fraud potential. The use of secure direct uploads when used in conjunction with optical character recognition (OCR) software creates a second level of data referencing and data validation for increased accuracy while permitting upstream risk-management reporting to senior management through a transparent, real-time reporting process.
Technology that provides an automated rules-based underwriting solution, with preconfigured workflows that restrict the progression of the loan without the fulfillment of specific conditions is reinforced by secondary data validations to effectively create an electronic underwriting file and an auditable trail detailing accountabilities and their ultimate fulfillment. This is the first step in creating an origination process that will produce loans that meet all investor and regulatory requirements. In turn, this will reduce the lender's exposure to put-back risk as well as meet all regulatory burdens it faces as a regulated entity.
Front-end quality control
The need for a strong quality-control program is central to all origination practices. Lenders are acutely aware of the risks associated with the origination of loans that do not meet strict compliance standards. Those risks include exposure to repurchase risk from investors and/or insurers, federal and state regulatory inquiries and audits, as well as the ever-present potential of civil litigation.
It is time for lenders to take advantage of the technology that is on the market and make it work for them, as opposed to figuring out a way to make their business processes work within the technology parameters. Individually configured workflow tools combined with a rules-based underwriting engine effectively provide a built-in quality-control solution for lenders that will identify credit, collateral, compliance threshold, tolerance and guideline violations as well as any potential fraud-pattern recognition prior to origination.
A comprehensive quality-control program consists of more than merely identifying violations. It consists of the preparation for audit and ensuring that each loan file contains all necessary documentation to support underwriting decisions.
The ability to electronically archive vital asset and income documentation to prevent lost or misplaced paperwork is extremely important to support a defense to a claim by an investor or auditor that the underwriter did not properly validate a vital income or asset element. For example, the Federal Housing Administration (FHA) recently implemented measures to increase monitoring of lender performance and compliance with FHA guidelines/standards and plans to publically report lender performance rankings to improve transparency and accountability. Areas of particular concern for FHA lenders (based upon recent audits by the Department of Housing and Urban Development [HUD]) must be addressed in a way that guarantees that the underwriting tasks will be fully completed and documented, such as validation of source of funds to determine a borrower's ability to repay and documentation of compensating factors for loans with low qualifying ratios.
These required processes that traditionally consisted of manual assessments have been memorialized on worksheets and supported by documentation provided by the borrower's employer, bank or asset manager. Should any of these worksheets or supporting documentation become lost or removed from a hard loan file, the lender will have a difficult time defending against a claim by the regulator or investor. Automating these processes and electronically archiving the calculations and supporting documentation permits an originator to eternally provide proof that tasks were completed, calculations were done and proof was obtained.
Lenders must leverage technology to lock down loan origination platform workflows and rule sets in order to alleviate themselves from the burden of managing multiple pre-funding and post-closing vendors and effectively instilling a front-end quality-control validation. The reliance on post-closing loan audits as part of a quality-control plan has failed and the process must be integrated earlier into the automated origination workflow.
Lenders must focus on a quality-control plan with the necessary automated controls to guarantee guideline compliance during underwriting. This will permit the reallocation of ineffective manual quality-control overhead toward revenue-increasing efforts to increase profitability and deliver real-time data on 100 percent of loan production that moves into the secondary market.
The beneficiaries of this direct data delivery will be those that have significant interest in overall loan quality, such as investors with skin in the game, rating agencies, credit-risk insurers, independent third-party diligence firms, and federal and state regulators.
Loan data distribution and industry initiatives to the secondary market
The capability to distribute data directly to the investor would greatly increase the ability of the investor and rating agency to assess the credit risk for an entire loan pool, especially considering that the investor may be required to carry some of the overall credit risk.
Loan data are distributed to the investor directly from the originator's platform and are supported by all ancillary electronic documentation to validate the accuracy of the data. With a properly formatted electronic and searchable file format, the loan-level due diligence can be conducted on a rolling basis or in a static review pool, reducing the scope of the necessary diligence.
Certifications of loan compliance and fraud review from commercially accepted vendors will accompany such loan data, reducing the redundancy of testing on the back end to effectively reduce the per-loan due-diligence cost for the investor while permitting each loan to have some level of subjective data verification for accuracy purposes.
Clearly, this level of transparency is the key to the ultimate determination by investors of whether to purchase the mortgage asset or not. Traditionally these data have been collected, verified and distributed by independent third-party due-diligence providers auditing a sample of loans from the loan pool that is set for potential acquisition.
The Association of Mortgage Investors, a national trade association representing an array of institutional investors and asset managers, released a set of guiding principles to Congress and regulators on April 1, 2010, detailing (in the association's opinion) the steps needed to restore the struggling securitization market. The first recommendation is that financial institutions need to "[p]rovide loan-level information that investors, ratings agencies and regulators can use to evaluate collateral and its expected economic performance, both at pool underwriting and continuously over the life of a securitization."
It is abundantly clear that the traditional due-diligence practice of merely sampling loans is fundamentally flawed. All of the interested parties involved in the private-label securitization industry can no longer rely on broad sweeping assumptions of quality. Trading desks, credit-rating agencies and insurers can no longer rely upon data from a loan sample to make purchasing decisions.
In order for investors to regain faith in the securitization model, 100 percent of all loan data must be available for each individual asset that will comprise the securitization collateral in order to ensure that risk is known and understood by all parties involved. Electronic delivery of loan data to investors and rating agencies will permit the necessary credit-risk analysis on each individual asset that is intended for delivery to the securitization market. This will effectively provide the necessary data for a fully transparent credit analysis while also providing the investor a look at a prospectus that contains not only the credit rating, but the data, schema and all representations of the loan-level data in that particular trust.
Electronic delivery of loan data to regulators already has become a reality. In establishing national examination guidelines for state regulators, the Conference of State Bank Supervisors (CSBS) and the American Association of Residential Mortgage Regulators (AARMR), both based in Washington, D.C., have identified that a sample-based audit is no longer a viable solution.
The new examination guidelines are moving to a process in which examiners can look at 100 percent of an originator's production. The Multi-State Mortgage Committee (MMC) is a representative body of state mortgage regulators appointed by CSBS and AARMR to represent the examination interests of the combined states under the Nationwide Cooperative Protocol and Agreement for Mortgage Supervision. Under the Protocol and Agreement, the MMC has implemented processes of uniformity and modernization for the examination of state licensees, including the use of lender data uploads to conduct electronic loan portfolio reviews.
In February 2010, CSBS and AARMR announced the availability of a new data-file format called the licensee examination file (LEF) designed to provide originators and software providers with uniform loan portfolio submission standards to state regulators for electronic examination. Beginning last year and continuing through 2010, the MMC is conducting pilot examinations using ComplianceAnalyzer� software provided by ComplianceEase�, Burlingame, California.
This pilot project is being led by Pennsylvania and was kicked off by notification to nearly 250 lenders in 2010 that will provide for a phase-in period for the technology-based examination process system integration. The MMC is making significant strides with software vendors to employ the LEF for system upgrades.
"Rather than each lender figuring out how to integrate their portfolios, the lender's software vendor should be able to push an upgrade out that simplifies the process," says Chuck Cross, vice president of mortgage regulatory policy for CSBS.
The MMC expects that by year-end, all multistate mortgage companies should be able to integrate with the standardized request for electronic file data. The key here is that regulators will have greater ability to target and review problem loan files than traditionally has been the case with random-sampling methods. However, the ability to scrutinize more thoroughly will likely be a good thing for many lenders. If the software review shows very little risk in the lender's portfolio, then it is quite likely that the traditional sample size will go down--not up.
"There is a win-win for everyone when an institution invests in good compliance and quality-control measures," says Cross.
The American Securitization Forum (ASF), New York, launched the Residential Securitization Transparency and Reporting (RESTART) project, an industry-developed initiative established in February 2008 to help restore investor confidence in the private-label securitization market. Project RESTART is aimed at increasing transparency and standardizing certain aspects of the private-label securitization market through several initiatives, specifically the ASF Model Representations and Warranties as well as the ASF Loan Identification Number Code (ASF LINC(symbol)(TM)).
From a loan originator's loan-performance-metrics perspective, the ASF LINC project is particularly relevant. The LINC is a 16-digit identification code that captures underlying loan type, origination date and country of origin, and randomized alphanumeric data to create a unique ID for loans that may be pooled and sold into the capital markets. The ASF LINC is assigned at no cost to issuers and is stored in a central data repository administered by New York-based Standard & Poor's Fixed Income Risk Management Services (S&P FIRMS).
The ASF LINC allows parties to track a loan from inception throughout the life of the loan. These loan-performance data offer the perfect opportunity for originators to obtain information on how well their loans are performing. The performance data on loans they originate will be central to their institutional loan performance metrics that will support the argument that their origination practices are strong and effective.
These loan-performance metrics, when used in conjunction with the results from the electronic loan file audit submissions, will provide originators the affirmation that they have a strong origination platform and that they will create good loans for private-label securitization.
Quality is the future
Technology, when fully utilized, can be a great enabler of institutions that seek to stay competitive and lower intuitional risk in a volatile industry. For private-label securitization to re-establish itself as a source of liquidity for the mortgage lending industry, lenders must originate loans without the repurchase risk and regulatory compliance risk, and stand behind the assets that are sold into the secondary market.
With the technology solutions available today, there is no disputing that loan quality starts with the originator's ability to control its origination process, lock down its data and deliver quality loans to investors.
Investors should seek to create relationships with originators that can prove they have the ability to produce loans that meet strict guideline requirements supported by performance metrics. These preferred originators should produce quality loans delivered through an origination platform that can lock down internal compliant processes, with the goal of creating a world where a lender's success is driven by its reputation for producing and delivering quality products.
In a perfect world, quantity will never inhibit quality; mortgage lenders that make the commitment to quality through technology will separate themselves with performance metrics. Those metrics will ultimately drive loan-sale pricing simply because investors will pay for the superiority of collateral that their respective prospectus will proudly identify. MB
John Levonick is chief legal and compliance officer for Mortgage Cadence Inc., Denver. He can be reached at jlevonick@mortgagecadence.com