AI and the Transformation of the Mortgage Industry

Your Source for What Matters in the ML Mortgage Industry

Introduction of AI in Mortgage Industry 

The global economy is expected to shrink by 3% due to COVID-19 pandemic. Unemployment rates are progressing on one hand, and businesses are cutting pay on the other

US GDP is estimated to touch $22 trillion in 2020. The mortgage industry contributes 10% of the GDP through new loan origination. However, the current slow-down will impact the mortgage industry as well.

“No doubt “Artificial intelligence” is the new buzzword in the current pandemic world. Organisations have started exploring out this technology to automate or perform tasks accurately and elevate their revenues. AI has percolated every aspect of human life and aims to do even more in the coming years.

According to a survey conducted by Gartner, organisations that have deployed artificial intelligence grew from 4% to a whopping 14%. This rapid wave of transformation in the organisations, from “Non-AI” to “All – AI” has not left the mortgage industry untouched.

1.2 Mortgage Industry’s Evolutionary Pressure  

The US mortgage market has changed dramatically since the global financial crisis. While much of banking has continued its decades-long consolidation – the mortgage industry has moved in a different direction.

Instead of consolidating, the mortgage origination market has been delineated by increased fragmentation. As the industry moves forward, players must consider strategic positioning and technological advancements in an environment characterised by an accelerating pace of innovation. 

With numerous demanding and tech-savvy borrowers, continued Mortgage Tech interest across the value chain, and a pronounced rise in technology investment, the mortgage market is sure to look tremendously different than today.

The mortgage industry is facing five (5) key challenges that are setting pressure towards a paradigm shift. 

Lack of transparency  

There is no easy way for consumers to identify, at an early stage, those products for which they qualify”, which impedes both the consumer’s ability to find a suitable deal and intermediaries’ ability to find them the best deal promptly. 

This lack of transparency in lending criteria is one of the reasons why about 30% of consumers missed out on being able to use lower-priced mortgage products for which they would be eligible. The most obvious way to make more information available is using tools intermediaries could use to identify, early on, those products for which the consumer is or is not likely to be eligible.

Operational stress    

The current origination process across lenders in the mortgage industry is lengthy and frustrating. The typical lending process in the mortgage industry requires around 40% manual intervention, with multiple weeks as an average time to offer (from start to finish). With high time to offer, frequent cancellations, revenue leakage, poor colleague-broker-customer experience and high manual errors, the mortgage lending process is cumbersome.

Cognitive malaise    

Mental exhaustion can happen to anyone who experiences long-term stress. It can make you feel overwhelmed and emotionally drained, and make your responsibilities and problems seem impossible to overcome. Many underwriters experience mental fatigue and stress due to a lengthy mortgage lending process – in gathering and mapping all the documents manually.

Data processing obsolescence   

The old-fashioned traditional paper-based manual process reduces operational efficiency and makes it difficult to ingest and capture relevant data.

Escalating complexity    

The escalating complexity of mortgages – right from loan application to funding is time-consuming.

Automation dysfunction  

Lack of tools to automate the mortgage lending processes, that make the origination process complex.

As these challenges converge and build up operational pressure, companies are aggressively seeking innovative solutions. New technologies developed from AI is redefining the nature of the business.

1.3 The AI Leverage  

Artificial intelligence has become a game-changer in the mortgage industry. 

As more and more, mortgage organisations start investing in modern innovations and technologies, only a handful of these organisations understand the concept of these emerging technologies, and how this development can change the ball game forever 

Tech Enabler #1: Capture 2.0   

Capture 2.0 also referred to as “intelligent data capture” is an amalgamation of computer vision solutions, machine learning models that help lenders to precisely identify and classify more cases or loan documents and extract more data, accurately from them. Machine learning models can be trained to identify and recognize patterns and make sense of massive data.

The intelligent data capture and the subsequent smart validation ensure for automated data completeness, correctness and consistency checks across the documents submitted as part of the application thus reducing the processing times and significantly reducing time to offer.

Tech Enabler #2: Conversational AI   

Conversational AI provides businesses with consumer-grade experience of consuming insights generated leveraging visual and conversational (both text and voice) techniques that can be shared across discussion forums and groups.

An interactive visual & conversational user interface for consuming insights 

  • Dashboards: Real-time & predictive reports 
  • Chatbots: Adhoc query with text & voice 
  • An Interactive & collaborative medium with a default discussion forum & integrated with other communication

Tech Enabler #3: Machine Learning Models  

Documents such as mortgage illustration, application declaration, payslips, bank statements and affordability assessment form are a rich source of valuable information, that can be leveraged to gain meaningful insights.

AI for document processing is a powerful tool for streamlining workflows, minimising delays, and reducing errors caused by manual document classification.

Recommendation engine modules can be used to recognise & automatically classify documents based on structural features (layout-based document classification), textual features (content-based document classification), or both.

 It enables users to automatically classify various mortgage specific documents such as payslips, bank statements, legal documents, valuation documents, affordability assessments, correspondences, and others.

This ensures important information is easily available for intelligent decision-making, eliminating risk and cost associated in manual document management by improving the time to offer and time to fund significantly.

The models are also helpful in populating an activity list for the loan processors. The data extracted from the documents are fed as an input to models which populate a standard activity list for the processors to work. This adds uniformity to the processing and significantly reduces TTO.

Machine learning algorithms provide high levels of accuracy and reliability by handling messy inputs. There are different types of algorithms, that can be used to classify documents.

1.4 A World Beyond Digital  

The operational terrain in the mortgage industry requires substantial changes that extend beyond digital solutions. A more focused and intentional approach is necessary. 

Intelligent Digital is the ability to align business understanding with technology innovation and human insight to solve important business problems. An intelligent digital platform leverages digital connectivity, cloud computing and artificial intelligence to create a connected ecosystem.

Intelligent Digital = Digital Capability + Easy to use AI

Key attributes of Intelligent Digital Model

  • Data capture and input from multiple sources
  • Data extraction capabilities for multiple languages
  • Integration into existing business process management systems
  • Document classification using Natural language processing and text analytics
  • Intelligent automation and decision making
  • A secure and reliable system that protects the interests of multiple parties
  • Document storage capabilities
  • Best in class user experience

In the mortgage industry, the Intelligent Digital Model has contributed to seven business transformations:

  • Achievement of significant productivity benefits
  • Improvement of the origination process efficiency
  • Improvement of organizational relationships - borrower, broker, and colleague
  • Cognitive enhancement - Pre-trained to self-learn over time
  • Ease of implementation - Bolt-on to existing technology
  • Productivity enhancement
  • Revenue improvement 

With these intelligent digital transformations, the mortgage industry is entering an undiscovered yet exciting new territory

1.5 AI and the Future of the Mortgage Industry  

Leveraging AI to intelligently process scans of paper documents could significantly reduce dependence on paper and manual validations and accelerate transaction speeds. 

Further, being able to classify cases into different processing queues and maintaining a real-time status for remote-located internal staff and external partners, would significantly streamline operations and help underwriters with unprecedented cognition. 

Lastly, the explainability of AI, with reports would significantly improve compliance, QA activity and uncover new revenue streams for mortgage origination.

While you are here, other top articles you might be interested in

1. Tech Enablers in the Mortgage Industry

2. The Future of Computer Vision, Machine Learning and Artificial Intelligence in Mortgage Industry

3. How industry 4.0 principles can work in the favor of mortgage origination?

4. Top 5 Real Challenges in Building Predictive Models in Mortgages

5. 5 Ways in which Mortgage Lenders can Leverage Digital Lending for Good

6. The Power of Intelligent Digital in the Mortgage Industry by Digilytics AI

7. Whitepaper on Data Lakes: A new approach to managing data

8. Applying industry 4.0 principles to mortgages

9. Intelligent Digital Income and Expense Verification by Digilytics AI

10. Digilytics AI is one of the top 30 ML-powered start-ups

11. Top Challenges in Implementing a Learning Interface

12. First time right applications for mortgage lenders – Expectations vs Reality

13. Key Takeaways from the Webinar on Intelligent Affordability Services

14. Making First Time Right Applications a Reality in Mortgage Lending