Mortgage servicing is a scale business, meaning the economics of scale can be achieved with larger servicing portfolio by spreading the fixed costs among more loans being serviced. Such scaling; however, hasn’t achieved the expected results as indicated by both the increase of servicing cost on a per-loan basis and on loans serviced on a per-employee, basis according to the research for the past decades by the Mortgage Bankers Association.
This trend is more distinct for the non-performing loans whose servicing costs quadruple from below $500 before the housing crisis in 2008 to more than $2,000 in the past few years. Apparently, such an increase is largely due to the compliance requirements posed by the regulators. The servicing industry should reform by adopting new technologies and data-driven approach to automate the compliance process cost-effectively.
Track the right metrics
When you use your data to track the right metrics, you are empowered by such insights to focus on the most important things for your business and provide smart scaling. Quickly profiling at-risk borrowers in different situations can be a very useful technology in servicing. Using forbearance plan under the CARES Act, for example, a wave of borrowers come to forbearance exit that significantly stretch a servicer’s operation capacity limit.
Servicers can use aggregated monthly servicing data to narrow down the borrowers in every stage of forbearance and prioritize resources for those that need help the most. By doing so, you can cut down wasteful expense and maximize your employee capacity.
Profiling these at-risk borrowers involves charactering them using social, economic, geographic and monthly loan-performing information benchmarked against the national and regional statistics. For instance, a combination of a borrower’s credit score, loan payment history, employment, loan-to-value ratio, location, local income level, and many other characteristics can be used to infer the borrower’s ability to repay.
When you use machine learning and artificial intelligence on top of this profiling, you can make — at scale — personalized recommendations of remediation options. Your borrower outreach can be more targeted and hence more effective. And, you can avoid the mistake such as offering a 40-year modification to a forborne loan with a 5-year remaining term.
Data on top of machine learning and AI gives you the advantage
Servicing rules change fast and they have short implementation timelines. You can pinpoint the faulty areas by running rule-based exception monitoring functions. Since servicing is all about timing – when things start and when things end – constant monitoring and tracking loan performances and regulatory change is critical in compliance management. You can only become more proactive in mitigating compliance risk as well as other risks quickly and effectively with speedy information processing and quick to action on the information.
Similarly in financial portfolio management, such profiling techniques are used for adequately and timely evaluating borrower’s default and prepayment propensity under changing market dynamics. This can have a tremendous impact on a servicer’s bottom line and MSR valuations. With the data at your side, you can actively manage your risks and boost your profitability with corresponding hedging actions and customer outreach.
Diagnose the health of internal processes
Data can also be used to diagnose the health of your internal process. Every borrower touchpoint, from payment collection to customer complaints represents a data point in servicing process. By tracking each stage in this process, you can gain a better view of the inefficiencies and bottlenecks of the servicing operation, such as employee productivity, client service performance and others.
These analyses can build operation optimization and identify ways to grow smarter without incurring huge outlays of hiring and capital investments. For example, a customer call history may show a few common topics that could have been answered more easily by making that information available online or through written communication. This can free up time and resource for customer calls on more important issues.
Data won’t replace humans; it will make them smarter
At the core of this digital success is data technology. Technology is not to replace human but to make human smarter. It can free up human to do what they are good at by automating part of work that machine can do best. Instead of spending 99% of the time working on getting the data right and 1% of the time understanding the information from these data and make human intelligent decision, it should work the other way around by using machine to automate and reduce the data processing time from 99% to 1%. So you can get the best of both worlds. At the end, it will be human to discover all the whys and tell a good story.
This will require the data management system to be capable of analyzing big data. Big data means not only the sheer volume of the data, but also the variety and velocity of the data. The system should be able to pull in data of all different formats from all different sources and generate results on an almost real-time basis.
Technology can now be scaled for small companies
The good news is that this has been a reality in modern SaaS solutions thanks to the scalable cloud native infrastructure. Cloud technology evolves in a way that smaller companies can access large datasets and the same level of technology infrastructure that was used to be exclusive to only large companies. The technology access to scale has been democratized.
Billions of data points can be processed in matter of minutes or even seconds. Data can be segmented and analyzed at very fine granularity in multi-dimensions and quickly rolled up into different hierarchical levels. Scanning of the loan performance data can walk back and forth in time in terms of selecting historical look-back and projecting future forecast.
More importantly, elastic pricing schemes in cloud computing minimizes fixed cost and allows variable cost cutting of computing resources on a per-minute basis which is certainly more palatable than that on human resources. Therefore, the industry can become more stabilized without seeing large personnel turnover due to the cyclical nature of this business.
Going forward, servicers will likely face more regulatory scrutiny as they have learned from the last housing crisis. Staying compliant is more costly than ever. Investment in data technology to put the effective risk and control in place can help scale the business better in light of these challenges. Servicers should keep this in mind when growing the business – not only to grow faster, but also grow smarter.
Howard Lin is president of mortgage risk analytics company Cielway.
This column does not necessarily reflect the opinion of RealTrends’ editorial department and its owners.
To contact the author of this story:
Howard Lin at [email protected]
To contact the editor responsible for this story:
Sarah Wheeler at [email protected]
The post Opinion: Use data to scale your mortgage servicing business appeared first on HousingWire.