Each year Rabbet surveys commercial real estate lenders and developers to uncover the latest industry trends in CRE. In 2019’s report, we revealed a huge shift towards automation and uncovered... Read More
Three Loan Servicing Challenges and Process Automation Solutions
We recently attended the American Banker-hosted webcast Addressing Market Needs for Business Process Automation. The focal point of the panel discussion revolved around research conducted by Dana Jackson, Vice President of Research at SourceMedia and Canon U.S.A. They captured data from 300 loan servicing managers and executives. The insights from the study impact lenders including those curious about automation and machine learning in banking and loan servicing.
Our key takeaways:
1. Loan servicing costs are dramatically impacted by regulation
Lending and loan servicing costs have increased due to regulatory compliance instituted during the “great recession.” According to panelist Craig Hughes from CC Pace, the fear of non-compliance has led to 2-3x more work, and regulations have led to a tripling of costs for servicing performing loans. He went on to say that improving employee efficiencies is a top priority as the workforce represents 40% off all costs associated with loan servicing.
2. Lenders are desperate for process automation
Desperate may be an exaggeration, but technology has mostly bypassed loan servicing. According to the data, 63% of the participants plan to implement technology solutions as a top priority for increasing competitiveness in loan servicing. They see the investment in technology as vital to improve efficiencies, mitigate risk, and reduce menial workloads so employees can focus on customer-facing and profit-generating initiatives.
“The information flow is dependent upon manual input that could be automated to speed up the process.”
Even today, much of loan servicing work is performed manually by way of data entry and massive, error-prone spreadsheets. For example, 30% of respondents indicated that workflows to manage lien releases and documentation processes were “All Manual” or “Mostly Manual.” Moreover, manual payment processing has the highest reported impact on servicing loans.
Top reported examples of lending management work that is done manually include:
- Lien releases
- Billing and payments
- Document and data management and record-keeping
- Regulatory compliance management
Automation isn’t a silver bullet, but the current amount of manual and menial work impacts cost containment and the drive to profitability. Automation absolutely reduces menial tasks, as seen in these several machine learning in banking use cases. Also, machine learning and associated predictive analytics provide the information employees need to perform better at their jobs.
“The current state has poor workflows and too many manual steps. We need to improve payment processing, communication, and documentation.”
Overall, respondents are interested in automation tools that can streamline processes, improve accuracy, mitigate costs, improve compliance, and enhance customer service.
Oh, and there’s the whole issue around incidents.
3. Payment Processing and Incidents
According to 47% of the participants, time spent manually reviewing documents and verifying loan data has the most significant impact on payment processing and onboarding. A top challenge for lenders relates to the additional administrative work and lousy customer experience associated with incidents like short pays or overpayments due to human errors. Whatsmore, reducing the frequency of incidents related to payment processing and billing was a top goal of lenders participating in the study. As a side note, have you read about the recent $35B overpayment? Talk about an incident.
Lenders see the automation of approval workflow process as a way to cut time and measure employee performance. For example, Canon and Rabbet offer tools to give loan administrators flexibility to create digital approval hierarchies. They can also quickly develop workflows to generate pre-programmed notifications when authorization is required.
To further drive engagement, when notifications go unanswered, alerts are created and can be used to escalate time-sensitive issues. Also, insights from these workflows help identify bottlenecks or individuals impacting processing speed. Some stakeholders are stronger managing some tasks while others’ strengths lie elsewhere. These insights allow leadership to allocate work based on employees’ strengths or provide additional training to develop necessary skills.
“Systems that are user friendly and don’t allow/mitigate human error as much as possible help banks function efficiently and effectively – better servicing the end customer.”
Loan Servicing Software as a Solution
In real estate development, we’re seeing a rapid rise in modern built technologies. For example, fellow Austin company, Kasita, has taken modular design to a new level with stackable homes for underutilized land plays and urban infill developments. In addition, ICON, a construction technologies company, can print a 650-square-foot house out of cement in under 24 hours, a fraction of the time it takes for new construction.
But, what about advances in loan servicing and loan management software for commercial lenders?
The leading software for construction loan management is now able to perform robotic process automation (RPA). According to Mr. Hughes, RPA in a lending environment can address many of the challenges faced by lenders attempting to cut costs and improve profitability as it:
- Assists with decision making and staying ahead of the curve
- Analyzes a portfolio to recognize trends
- Provides ongoing assessment of potential exposure
- Builds datasets to use for predictive analytics
- Automates data comparisons
- Automates data entry
There are several examples of RPA and machine learning in banking use cases. However, Rabbet is the first to apply it uniquely to the post-closing construction loan management process.
Rabbet’s industry-first machine learning and process automation technology applies digital workflow automation to capture and organize documentation. It leapfrogs current technology, that’s limited to optical character recognition to create searchable PDFs, by also using computer vision, machine learning algorithms, and rule-based predictive modeling.
So, in addition to storing and providing digital document access, the technology parses information from the myriad of emails and PDFs included within a construction loan draw request. It then analyzes the data and performs “searches” to identify errors. Spreadsheets and forms capture the clean data, and the software creates recommendation reports and highlighted alerts based on rules and the data. Example highlights for a construction draw review may include a warning that an invoice doesn’t reconcile with a receipt or that the approval of a draw is on hold pending the evaluation of an inspector.
Cost containment is a priority for loan servicing. To get there, lenders that adopt new technologies like robotic process automation and machine learning solutions are realizing dramatic efficiencies. A process that had taken hours of human interpretation is now streamlined and done in seconds with improved accuracy (and fewer instances).
The latest in loan servicing software creates a single and secure portal that hosts post-closing loan servicing information in one location. This type of collaboration and archival mechanism means loan administrators can easily audit an entire portfolio to ensure compliance. They can also run an instant project and portfolio level report like loan composition, cash flow projections, draw processing, and missing lien releases. Moreover, they have access to real-time information about payment status. All of this information helps lenders better communicate internally and with customers.