Reduce Turn-Time And Increase ROI

In today’s challenging mortgage market, leading lenders are looking to streamline operations while creating greater efficiencies to lower the cost to originate loans. That is why they are looking at areas within their operations that can be automated. When it comes to automation, you hear terms like RPA, bots, OCRs, and whatnot; it can be a complex concept to demystify. By this article’s end, you will better understand what all these buzzwords mean in a mortgage automation configuration. At the same time, you will have a fair idea of calculating the TCO (total cost of ownership).

Benefits of Automation

Lenders can obtain many benefits by investing in suitable automation projects; the following are some of the top benefits that customers we have worked with experienced.

Faster Loan Processing

Borrowers want to get into their homes in the shortest possible time; at the same time, they want this process to be simple and transparent. The typical loan processing timeline is labor-intensive and tedious; by implementing automated processes, lenders can close more loans in less time. Automation leveraging RPAs and AI bots can significantly reduce loan production time.

Higher Accuracy

According to the “Institute for Robotic Process Automation & AI” (IRPAAI), Out of every 100 steps, a human, on average, is likely to make up to 10 errors. Rectifying the mistakes that occurred during the loan process is expensive and time-consuming.   By implementing automation, you can significantly reduce the number of these loan processing errors.

Greater Regulatory Compliance

With ever-changing rules and regulations on mortgage lending, it is vital to implement automation that can help lenders minimize operational risks while at the same time improving regulatory compliance. 

Lower Costs

Automation is not an all-or-nothing kind of initiative; you can automate a part of your process VOI or VOE or a more protracted process like automating disclosures. According to IRPAAI, you can reduce your costs by 25-50% compared to the conventional way of processing loans. In most cases, if the proper process is selected and the right vendor is selected, you can achieve positive ROI on your investment within six months.

Mortgage Automation Types

Before we dive into the actual use cases for mortgage automation, it is essential to know what automation categories are available to add the most significant value to lenders.

Rule-based automation:

Rule-based automation is often referred to as decision-based AI where rules are pre-defined instructions within the systems you use, whether LOS, CRM, or others. Usually, these systems are event-driven means; if “this” happens, then do “that.”

Systems Integrations (APIs):

Even in the age of digitization, different systems work in silos, and human intervention is needed to move data from one system to another. By implementing Systems Integration, you create direct lines of communication between two or more methods, reducing errors and increasing speed. So, you can integrate your CRM with your LOS for a seamless data flow between the two. Typically these integrations are done through APIs (language through which two systems communicate).

Intelligent Document Processing:

In the mortgage industry, data is at the heart of automation; approximately 80% of the data is in unstructured formats like documents, emails, images, PDF files, and whatnot.

Intelligent Document Processing (IDP) can read, extract, and process data ingested from a variety of sources mentioned above.

How is IDP different than conventional OCR? It does the following:

• Pre-process: which includes noise reduction, de-skewing, etc.
• Classification: based on machine learning / AI, which can identify mortgage-related documents like W2, 1099, etc.
• Data Extraction: this is where the actual data will be extracted and pushed to the appropriate forms in your LOS or other enterprise software like 1003 etc.
• Domain Specific Validation: enterprise-level IDPs also have information about the domain and make validation decisions based on the industry information.
• Human in the Loop: this helps the AI engine to have supervised learning, like items that the IDP cannot process.

Robotic Process Automation (RPA or bots):

RPAs mimic human interactions with the computer, the tasks which are repeatable and don’t require intellect or reasoning can be the best fit. Things like ordering Fraud or Flood, submitting orders, and getting a response from the websites are a few mortgage-related examples.

AI & ML:

Artificial Intelligence and machine learning are two buzzwords everyone is talking about. Leveraging the potential of these two technologies will provide long-term and multi-fold benefits for the dollar spent.

In combination with ML, AI has human intelligence capabilities, enabling these technologies to learn patterns and behaviors of data and make decisions based on the same. The unique property of AI & ML is that it gets better over time.

A few use-cases of AI in lending are reduction of non-performing loans, improving loan portfolio, including thin-file borrowers, customer behavior patterns, etc.

Different kinds of vendor engagements

Time & Material Model:

In this arrangement, you pay the development team for the time & material spent on developing the automation solution for you. In this case, most of the time lender owns the code.

Fixed Price Model (SOW):

The automation company offers a fixed timeline and cost to build automation for you. For example, if you want to automate your Initial Disclosure process, the consultant will provide you with a fixed-price solution.

Dedicated Team Model:

Most lending automation projects are lengthy and require the team to work for an extended period of time. The lender hires the entire team to build their automation solution in this arrangement.

Extended Team Model:

If your company has in-house automation development resources, companies can work as your extended team.

Automation as a Service:

Automation companies also offer a pay-as-you-go model where they charge per loan.

Examples of processes that lenders automate

Following is a list of a few processes lenders prefer to start with for their automation journey.

• Automation of Disclosure, including initial disclosure, COC disclosure, NOA disclosure, final disclosure, etc.
• VOI & VOE
• Ordering Title
• Denial Letters
• Round Robin tasks assignments to the users
• Placing eConsent requests and conditions
• Reviewing and clearing conditions
• Ordering services, including flood, fraud, automated underwriting system (AUS), credit, and appraisals
• Intelligent Document Processing with Classification & Extraction
• Ordering mortgage services
• Verifying AKAs

Total Cost of Ownership

Typically, the total cost of ownership (TCO) takes account of all the expenses related to a specific project, not only the direct costs. Generally, lending companies make the mistake of taking just the initial cost of any automation initiative. When the project is over budget, they have to stop due to non-approval of the expenses.

It is essential to calculate the TCO as early as possible before starting the project. We will discuss the variables that make up the total cost for any automation initiative, which will help you make a better business decision.

Startup costs:

This refers to all the costs incurred when taking the new automation initiative in production. This includes:

• Software usually up-front software cost plus per user license
• Hardware servers, storage, and companies often forget to take costs for backup and DRP if it is a critical application.
• Implementation costs are sometimes overlooked by a few vendors gimmick with the clients where they provide a minimal upfront cost and cover all the charges in the name of the implementation.
• Training is an essential part of automation initiatives which provides the user with an understanding of what they need to do and what not.
• In other system integration, your core systems to which your automation connects ask for separate API key costs or per-use costs for API.
• If you are going for a pre-built automation solution, it is mandatory to keep customization costs in mind.

Operational costs

After installing the software to production, there are operational costs you must keep in mind while making an automation solution decision.

• Maintenance & support is integral, and you should ask for SLA before you even sign the agreement.
• Patches the maintenance contract should include security and bug-fix patches.
• User licenses if it is recurring or per-use basis.
• A regulated industry like mortgage needs enhancements as and when compliance comes into play, you should be aware of ongoing change costs; usually, it is in the form of hourly rates.
• User & admin support is an integral part of operational costs; though it is most of the time taken from existing resources, a cost is associated with it.
• DR (Disaster Recovery) & HA (High Availability) sites are required for mission-critical automation, which can cause a halt in business operations.
• Cloud or data center costs need to be taken into account too.

Conclusion

Automation is moving from nice-to-have to must-have, and lending companies should consider starting their automation initiatives immediately. If you keep the above details in mind, we are confident that you will have successful automation and after the first success, we have seen the company follow the path and be more profitable.