Data is the life blood of today’s mortgage process. Without data, the origination and servicing of mortgage loans would be impossible.
While that statement has always been true – even in the old days of completely paper-based mortgages, people still had to fill out applications with numbers and information – it is especially true today, as lenders use multiple sources of data to process loans.
And there is exponentially more data available.
Managing all that data effectively has become a new focus. Lenders must ensure that they are using the best and most accurate data available when processing loans, or face the risk of buybacks and increased defaults. What’s more, they must ensure that they are using their data assets effectively, in order to reduce costs. Similarly, servicers are tasked with utilizing their data to their best advantage for their investor clients, as well.
To learn more about how lenders and servicers are leveraging data and what data-related challenges they face, MortgageOrb recently interviewed Souren Sarkar, president and co-founder of Nexval, a provider of technology-enabled mortgage services and fintech innovation for the financial services industry, and a pioneer in mortgage business process outsourcing (BPO).
Q: What are some issues the mortgage industry struggles with when it comes to leveraging available data?
Sarkar: There are several major issues at play. The first is that lenders and servicers manage so much data from many different sources, much more than any other industry. Borrower data, loan data, property data, appraisal data, title data, investor guidelines—the list goes on and on. The second issue is that the information systems used in the mortgage industry have been built up over decades, so there’s a lot of data sitting in databases that have not been updated in quite some time. Thirdly, as in any evolving industry like ours, things change. The basic mortgage process is the same but so many operational aspects have changed, from regulations and compliance to how the secondary market operates, so the ways we use that data change, too.
Q: Why do you think the industry continues to struggle with their data issues today?
Sarkar: The mortgage industry moves slowly, but at the same time, there are a lot of transactions. This leads to data monopolies, in which a handful of mortgage banking systems control the vast majority of borrower data. In fact, the leading origination platform and the leading servicing platform were recently bought by the same company. Having a monopoly on data is great for business because the controlling entity can charge people for their own data. At the same time, there’s little incentive to change, which is why most of these monolithic systems have not evolved fast enough to keep pace with new regulations, investor guidelines, and the market.
Within the industry there are also many drop-off and pickup points where data is further broken down into chunks and sitting in various multiple systems. Some technology providers, in desperation, will build subsidiary systems to hang off their larger systems. But this often makes working with data even more difficult. This disparate situation of large monolithic systems with a lot of duct tape systems hanging from them has created a very complex scenario that is difficult to break away from.
Making any changes to how the industry uses data is extraordinarily difficult because of the overall lack of innovation in our industry. Operationalizing data and making it strategically valuable is usually an afterthought. In most cases, data is handled by people and teams who are focused on an individual problem, whether it’s borrower verifications or escrow. But no one stops to consider the overall impact of adding a new field of data on the overall operation. Everyone is constantly spinning their wheels trying to keep up with a massive deluge of data.
Q: What exactly is a smart data fabric?
Sarkar: A smart data fabric is basically an architecture of data services and standardized data management practices across an organization, and typically involves cloud environments and intelligent and automated systems. Instead of using individual data storage systems, smart data fabrics leverage data services and APIs to pull data from different areas of the business to give an organization a more holistic view of its data.
Smart data fabrics are built on the concept that data is the core asset of any company. The idea is to build the data architecture from the ground up. With a data warehouse, for example, a lender is pulling data from everywhere and trying to make sense of it. With smart data fabrics, the lender is building the ability to organize the data in a usable way. The participants inside the organization are all able to freely share their data and with some level of standardization. It’s about building functional data in an organization and, ideally, across the entire industry.
Q: How can smart data fabrics help the mortgage business?
Sarkar: Smart data fabrics have benefits for any mortgage organization, but they are particularly valuable for servicers. For example, most servicers have a core system that they use for all their operations. When onboarding loans, loans come in and are registered and messaged into the system. However, servicers may have a specific system or module for default management, another for foreclosures, and one for third-party provider communications, any one of which may be vendor-driven.
Under current processes, all these systems and their users are constantly pulling data or data fields from the same core system for different uses. Not only is this a lot of work that must be repeated manually, over and over, but every time someone touches the data, there’s a chance for an error to be made. Smart data fabrics allow servicers to create distributed databases that release servicers from this gridlock and organize data in a way that enables them to keep up with changing data points and data fields.
Ultimately, smart data fabrics are what will propel the industry to the next level.
That being said, the push for a smart data fabric needs to start at the C-level. If a company has innovators on staff, those innovators need to have a voice and be given the opportunity to pursue their ideas. With most lenders, the head of underwriting usually has more power than innovators. That’s why many mortgage companies are better off accessing a third-party service provider who can take a holistic view of the organization and build accelerators and a proof of concept for a smart data fabric.
Q: How can smart data fabrics bolster AI in the mortgage sector going forward?
Sarkar: Smart data fabrics and AI go hand in hand. The structuring of data in a usable fashion is important for any AI system, so the system can access that data and use it to make actionable decisions and perform predefined processes. However, any AI system depends on having good data. If a lender has bad data, it’d better off deploying humans. But that can of course be problematic, as well, because humans too often rely on gut feelings to make decisions.
The bottom line is that it’s not data that makes organizations thrive – they thrive when data becomes information. Organizations can grow when they are able to make actual forecasts and view problems at high and low levels, without having to create another system or strain one’s IT resources to build what is essentially another rat trap. With smart data fabrics, it’s about moving forward with what a lender building today, making small changes, and slowly transforming it over time.
The key is understanding data is the business. Once that realization kicks in, change can happen.
As published on Mortgage Orb, November 7, 2022