The backdrop to 2008 crisis

The unsustainable home mortgage derivatives market was a key contributor to the 2008 U.S. financial crisis. Many lenders/bankers did not closely assess borrowers’ ability to repay before providing housing loans. Regulations of subprime lending were too lax and fraudulent interactions were common- such as overstating a borrower’s income and over-promising investors on Mortgage-backed-securities (MBS) products in the secondary market

The new age of lenders

The COVID-19 pandemic created a boom in the new home purchase loans and home refinancing markets as the interest rates dropped sharply. More lenders have come up to supply this demand, especially the new-age digital lenders. Meanwhile, millennials who constitute the majority shares of the mortgage market demand speed, personalization, convenience, and seamless experience.

Along with growing customer expectations and falling mortgage rates, escalating cost of loan origination, and the high prices of homes, the profitability of the mortgage industry is under pressure. The outstanding residential mortgage has still not matched the crisis level. Though the space has not grown since but has been through a lot of changes. Non-Banking loan origination now accounts for more than 60% of the market share as opposed to 25 % in 2008.

Technology-based lenders are gaining prominence in this space owing to their personalized offering, faster loan processing, and end-to-end digital process. Loan underwriting has become a central piece of loan processing as the data-backed loan approval process is a key differentiator for digital lenders instead of the traditional way of local broker’s discretion.
Traditional Financial Institutions like Wells Fargo, JP Morgan, Bank of America have started to launch the digital platform for mortgage lending to meet the changing demands. Digital lenders have showcased the potential of technology to disrupt the loan process, reduce operational costs and gain a competitive edge.

Over the coming years, we will see Artificial Intelligence becoming commonplace in the Mortgage Industry and the majority of lenders will be using it in their business

Growing need for Artificial Intelligence

Mortgage lending is a data-intensive business, it showcases the tremendous opportunity for AI/ML capability to use data to generate insights, make accurate and reliable decisions in seconds.
Mortgage lenders are now relying on technology to drive the lending process from loan origination, processing, underwriting, closing, and funding.

Typical mortgage lending process:

Artificial intelligence(AI) / Machine learning(ML) has recently gained prominence and is paving its way in the mortgage industry. It has enormous potential to improve the operational efficiency of document-intensive home loan processes by analyzing and organizing all the paperwork.

AI in loan origination:

The average time to close all types of loans is still close to 50 days, which is way more than what customers expect. AI can reduce the turnaround time of mortgage loans by effectively speeding up the data gathering, reviewing, analyzing, verifying, and digitizing. Let’s talk about how AI can help lenders provide a seamless experience to borrowers and speed up the touchpoints of borrower’s journey

Pre-approval

Pre-approval requires minimum basic details regarding the mortgage and borrower to estimate the loan requirement and suggest products. AI can enable the lender to use the data to personalize the loan products and ask more relevant questions

Mortgage loan application

It is a crucial step in the loan origination process. Borrower submits a different set of documents ranging from various income documents, asset documents, insurance documents, and debt documents. Pre checks on the quality of documents and type of documents can help expedite the complete process. Recognic (API-first SaaS product) which uses proprietary Machine learning models to check the quality and classify the documents automatically streamlines the process and digitize the documents which can be further fed into the underwriting process to take a quick, accurate, and reliable decision

Loan processing

Once the borrower submits documents, the next step is to gather all the submitted documents, review the information in loan files, validate the information and create a complete package for the underwriter to give final approval. Recognic’s AI models can digitize the documents and let the lender have complete control of the documents with minimal manual intervention.

Underwriting

It is the most crucial step of the loan application process where the risk assessment of the application is carried out before finalizing the decision of whether to approve or reject the loan application. While it is the most crucial step, it consumes a day to underwrite an application. AI-assisted underwriting combines the traditional sources and the borrower’s digital footprint to perform the risk assessment for more accurate creditworthiness. Reducing the turnaround time (TAT) of underwriting is dependent on the “right” information availability in the “right” way. Recognic’s advanced IP allows the pre-trained models to extract the relevant fields from the financial documents and help underwriters take fast and accurate credit decisions

Fast adoption of technology can be a differentiator in creating an improved customer experience and bringing down overhead costs.

Stayed tuned for our upcoming articles on some of the technologies under the Artificial Intelligence umbrella, which can disrupt the traditional way of lending.