Quality in manufacturing is calculated based on performance, the quality of components and the process by which those components are assembled. That works for building automobiles, but it does not quite work for originating and securitizing home loans. For the mortgage industry, the components are data, and the product is a decision made based on that data. Including the manner by which the data is processed from origination through securitization, these are the indices of long-term performance.
Big data and cloud analytics are terms that conjure up Matrix-like images for many mortgage bankers. For most, these ideas and methods seem very far from loan production's reality and day-to-day challenges. This being said, however, they are terms our industry needs to pay attention to because the technologies behind both are facilitating meaningful change.
Technology is evolving so rapidly that it is transforming the way mortgage bankers and investors perceive data. While ‘big data’ is truly the purview of giants like Facebook, Google or Twitter, the tools and strategies developed to define, parse, interpret, standardize and, last but not least, monetize data, will evolve mortgage banking.
Although the mortgage industry does not use big data in the same way Twitter does, monetizing the data captured from origination through securitization will play a pivotal role in the return to a healthy private securitization model.
Data will provide analytics, business information, transparency and accountability for loan quality and performance, all of which will drive informed investment decisions. Or stated another way, the technology of analyzing data will redefine how the industry measures risk.
Investors will always rely on loan-term performance as a macro index, but technology now facilitates access to an increasingly granular level of analytics. It is this emerging accessibility to data that will allow investors and regulators to properly evaluate loan production, compliance and securitizations. The market has been steadily moving in this direction since the ‘crash’ of 2008.
In 2009, the Federal Housing Finance Agency (FHFA) began directing Fannie Mae and Freddie Mac in a cooperative effort to develop the Uniform Appraisal Dataset (UAD) and Uniform Loan Delivery Dataset (ULDD) standards. Since then, the ULDD and the Uniform Mortgage Servicing Dataset (UMSD) have been steadily evolving to migrate origination today, and servicing into the future, into a uniform dataset.
These standards, in combination with a MISMO XML file format, will govern the creation of a recognized mortgage ‘language.’ Once the vocabulary and grammar of this new language is formalized, documented and adopted, the general assumption is that governance can become more transparent and manageable.
Can data itself cure all ills? No, it cannot. But bad data is worse than no data, and frankly, we as an industry have not done a great job with our data. The erosion in data integrity is well documented as files progress from origination to administration.
Where do those data errors originate? The most common source is an originating lender that hit the implode-o-meter several years ago. This is a direct consequence of an industry managed and analyzed through the prism of sales – neglecting quality assurance, redundancy and even technology – in the pursuit of raw volume.
A lesson we have all learned, in perhaps the harshest way possible, is that business risk and loan performance are ultimately predicated on the quality of data input. If we absolutely know errors exist in the data, then we absolutely know there are flaws in our risk analytics.
As much as many in the industry chafe at the idea, we are compelled to adopt a manufacturing mentality to address the data-integrity issues that continue to plague us. Data is our primary raw material. When a company is manufacturing cars, for example, there is a quality assurance process in place for both the raw materials and the process by which those raw materials are used in the creation of the final product.
Although this industry has a colossal impact on so many disparate parts of the economy, we really do not have a clear mandate when it comes to technology and data analytics. However, change is imminent as regulators enact data-driven regulations. Already, the FHFA and the Consumer Financial Protection Bureau have announced their intention to collaborate on the creation of a National Mortgage Database, whose primary purpose will be to gather and analyze mortgage industry data. This data will then be used to formulate housing finance regulations in the future.
As an industry driven by financial assets, we must come to terms with the reality that data is an asset that is bought and sold time and again throughout a mortgage's lifecycle. It is no less important than a skilled underwriter, a loan origination system or the very brick-and-mortar building in which you operate.
Data is the foundation of our industry. It is time we treat it as such.
Ruth Lee is executive vice president of Titan Lenders Corp., headquartered in Denver. She can be reached at email@example.com.