Innovating to Cut Costs For Mortgage Lenders, Save Time and Money for Borrowers, and Expand the Availability of Credit

Andy Higginbotham, Chief Operating Officer, Single-Family, Freddie Mac

“Big Data” is the buzz phrase in the business world. For the mortgage industry, responsible capture and analysis of data can uncover hidden patterns that reveal key insights that can be beneficial to the industry, its regulators, and its most important stakeholders—the consumer.

In the mortgage industry, cutting-edge algorithmic models fueled by data analytics stand to help lenders underwrite mortgage loans more quickly and cost-effectively and, in turn, sell them to investors such as Freddie Mac. The bottom line: A better borrower and lender experience and a more liquid and financially sound housing market.

What Does Big Data Mean to the Mortgage Business?

For mortgage lenders, data needed to underwrite loans and for modeling or other loan-related purposes comes from many sources. Those sources may include:

• In-house records such as loan files, bank statements and brokerage accounts.

• Third-party information such as credit scores and other traditional sources of underwriting data.

In our business, data analytics translates this information into models that analyze and draw conclusions from the data and update these conclusions as new data become available. These models, in turn, allow lenders to make better decisions in a wider range of cases than they could otherwise do.

  In the face of rising interest rates, declining volume and high production costs, controlling costs will enable lenders to save money, at the same time as they expand their capabilities to serve a wider customer base​  

Looking More Broadly at Non-Traditional Mortgage Applicants

By using the insights gleaned from big data, to which the borrower has allowed access, lenders can learn more about consumers who apply for a loan but who don’t have a significant credit history to present as a basis for prudent underwriting.

In the same way that lenders can build alternative credit profiles for millennials, they can also do this when working with borrowers in underserved communities, many of whom lack traditional credit histories. That’s true of millennials too, many of who don’t take out car loans, use credit cards, or work as salaried employees the way their parents did.

Accelerating the Mortgage Approval Timeline

In the face of rising interest rates, declining volume and high production costs, controlling costs will enable lenders to save money, at the same time as they expand their capabilities to serve a wider customer base. By integrating big data into loan origination and underwriting systems, lenders can more fully digitize application processing, speed up underwriting and bring borrowers to the closing table sooner.

For example, lenders can expand the data they use to calculate and validate a borrower’s income and assets, and then analyze the customer’s credit history more effectively and rapidly. In the last couple of years, some industry leaders are prequalifying in real time borrowers who request the service and authorize limited access to bank, credit card and brokerage account statements, employment information, and other relevant material. This can save the borrower the headache of gathering information from multiple sources and can reduce the underwriting timeline significantly. That’s a real benefit to busy borrowers and additional value that the lender brings to the transaction by leveraging the power of information.

In addition to enhancing data integrity, machine learning can help customers avoid last-minute obstacles by flagging early a data point on which an under writer will need additional information. For example, if the system identifies a large deposit or withdrawal in a mortgage applicant’s bank account, the lender can ask the customer about it early—via an account status alert—and feed the answer into the underwriting model, avoiding potential delays that may hinder a customer’s pending transaction.

The speed and efficiency of loan production—from originating to closing a mortgage—can cut lenders’ processing and underwriting costs, making them more competitive and saving the borrowers money. With big data, processors can prepare higher-quality loan files for underwriters who are then freed up to focus on the big-picture credit profile of a borrower rather than all the initial “stare and compare” work. This kind of efficiency can also shave days off the loan approval process and improve customer service.

Examples of Big Data in Action

Lenders know that success with their customers depends on improving customer service. In 2018, a Minnesota credit union used data analytics to focus on 1,400 members after calculating the amount of money they’d save by converting to short-term mortgages. Using data to quickly identify and refine its market, the lender maximized its marketing dollars, wrote nearly $30 million in new loans and saved money in the process.

In early 2018, a Georgia-based mortgage lender integrated data technology into its loan origination system to source employment status, income and other information from a major credit bureau. The lender said that within six months, it had automatically validated borrower financials tied to nearly 19,000 loan applications valued at $6.5 billion. In doing this, the company said it approved these mortgages in two thirds of the time it usually took and reduced closing times by about five days, definitive support for the mutual benefits achieved by lenders using data to reduce the burden on borrowers who’ve traditionally had to manually gather and provide records to loan officers.

Other solutions draw upon historical and public records data, the Multiple Listing Service, repeat sales and prior appraisals to assess the condition and marketability risks tied to a home’s value. This can help determine whether a lender can issue a mortgage without requiring the borrower to obtain a traditional appraisal. Without an appraisal, lenders can close a mortgage up to 10 days earlier and save the borrower $500 to $750.

Looking forward, the mortgage industry should continually explore and understand how it can make responsible, productive use of data and machine-learning analytics. This can help lenders and other mortgage professionals solve the challenges and issues of today and those of tomorrow. In doing so, we will improve the country’s housing finance system, save consumers money and ultimately reimagine the mortgage experience by delivering innovative solutions to help millions more Americans become sustainable homeowners.

Check out: Top Proptech Solution Companies