Delivering first-time right applications
Mortgage Finance Gazette
Mortgage Finance
3 years ago
Img

Introduction  

The Covid-19 pandemic had a significant impact on the UK mortgage lending market. Is your organisation overwhelmed with too many mortgage applications? Are you aiming to deliver more first-time-right applications? If yes, then the Digilytics AI-enabled SaaS product is here to save the day.  

A quick status update

Life before the pandemic was easier when it came to mortgage origination:

  • Keep a good credit score.   
  • Provide accurate details and documents while applying for mortgages 

Satisfying these two conditions was enough to secure a mortgage for a home or business on most occasions.   

However, since the pandemic, other factors also play a role in deciding if the application will be approved or rejected.   

According to a survey by Aldermore bank, four out of five mortgage applications by first-time buyers were rejected since the pandemic. That is, only 19% of applications were accepted compared to the 48% before the pandemic.   

While poor credit history is the main reason for loan rejections, administrative errors by the lender is also a significant contributor. For example, the administrative error rate was only 14% pre-pandemic, before rising to 35% in March 2021.   

Reason for lender errors

Over the years, the UK government has introduced many loan origination schemes to help and support both the buyers and the lenders. The Right to Buy scheme, mortgage guarantee scheme, and Bounce Back Loan Scheme are some schemes that help home-buyers and SME businesses.   

These schemes increase confidence among the public, which leads to an influx in mortgage applications. Lenders and case handlers in banks and other lending organisations then have to verify all the data. In the traditional loan origination process, two scenarios occur at this point: 

  • Humans are prone to administrative errors, especially when they have to process vast volumes of data. In addition, a small workforce may also take a lot of time to complete the processing.  
  • Hiring more lenders and case handlers to verify information from documents may be expensive, leading to more losses.   

 It’s not always the lender’s fault

Sometimes, the borrower makes errors as well. These errors may be in the form of:

  • Lack of complete information provided in the application  
  • Lack of documents supporting the mortgage claim  
  • Lack of information in the supporting documents   

Self-employed professionals mostly face issues with incomplete data and lack of documents. As many as  80% of self-employed UK citizens have faced mortgage rejections. Out of these citizens, over 39% felt that the mortgage origination process should be simplified, with the requirement of a few documents only.   

Understandably, lenders cannot shift to processing applications with a few documents only, as they have to verify if the borrower will be able to repay the loan. This verification can only be done with the help of sufficient information.   

To put it in a nutshell, these are the requirements that you must fulfill to process mortgage applications and deliver first-time-right applications,  

  • Reduce or eliminate any administrative errors  
  • Deliver higher accuracy in data extraction and verification from the documents   
  • Find other sources of information if the provided data by the borrower is insufficient.  

But how can you meet all these requirements and quickly process all the applications at the same time?  

Digilytics RevEl: A product that efficiently checks all your boxes  

Digilytics RevEl is a product that enhances the UK loan origination system using technologies like Artificial Intelligence (AI), Machine Learning (ML), and other advanced techniques.   

With Digilytics, you can accurately extract and verify borrower information and deliver first-time-right applications in a single automated mortgage origination journey.   

Our product, RevEl, is a SaaS-based solution that makes lending easy by,   

  • Helping lenders check the completeness of the application  
  • Case handlers extract the information, and  
  • Underwriters verify the information and make decisions on loan approvals.   

Document data extraction   

Manual processing involves applications submitted in the form of physical documents. The data is then manually entered into the computer by case handlers or extracted using OCR (Optical Character Recognition) tools.   

OCR tools are only limited to structured documents. However, supporting documents like bank statements, tax returns, payslips, etc., are sometimes unstructured. To accurately read and extract data from these documents as well, RevEl utilizes One-shot Learning techniques.   

OSL combines machine learning, computer vision, and natural language processing (NLP) to extract data with 95% accuracy. Using these techniques, RevEl enriches the data captured from OCR systems to read all the words in the document.   

Document data verification   

Once all the data is accurately extracted from the documents, RevEl’s artificial intelligence processing can help verify all the information for the underwriter.   

RevEl performs the 3C checks on the data that include checking its completeness, correctness, and consistency. It also cross-validates the information with data from other documents.   

RevEl can generate predictive models using the borrower’s credit history and buying patterns to help underwriters make better decisions on loan approvals and rejections through all these checks and validations.   

As the entire process is automated with Digilytics, the frequency of administrative errors by members in your team is significantly reduced, thereby increasing the rate of first-time-right applications.   

Verification through online data sources

Now that accurate data extraction and validation through documents is possible, what about processing applications of borrowers who cannot provide enough data?  

Online data sources are beneficial in such situations. For example, Digilytics has partnered up with AccountScore, a platform that provides complete bank transaction data and analytics directly from the bank. This is made possible with the help of Open Banking.  

RevEl offers Intelligent Affordability Service (IAS) with AccountScore, which enhances the income and expenditure analysis and provides real-time analytics and support so that you can scale your operations.   

Other online data sources also help you verify the authenticity of the documents submitted, verify the borrower’s employment data, etc.   

Conclusion   

Spending a lot of time processing each application and lengthening the feedback loop between borrowers and lenders can increase the operating costs and affect the productivity of the mortgage origination team.   

Digilytics can help you reduce your operating costs and time to fund by delivering more first-time-right applications using its AI-enabled SaaS product on documents and online data sources. 

 

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