Data & Automation in Mortgages – Underwriting Can Be Overwhelming

Post • 6 min read

Mortgages are the most profitable product line within retail banking. Although net interest margins have been relatively low for a while, volumes have been resilient and continued to grow. Hypostat puts a value of 7.83 trillion on outstanding mortgage volumes across the EU27-UK markets (+3.0% y/y). 

It’s big business. So many neobanks who initially entered the market with current account propositions looking for scale have extended or announced plans to move into the mortgage space and other lending areas in search of better margins. This includes the likes of Bunq, Monzo, Starling and Revolut. The list goes on.

The vast amount of data in mortgage journeys

But building a modern mortgage proposition in a profitable way requires a fast, secure and efficient cloud-native banking engine under the hood. Mortgage lenders collect over 5,000 individual data attributes of borrowers during a typical mortgage journey. Yes, you read that right - 5,000. It’s a vast amount of data. 

Mortgage lenders collect over 5,000 individual data attributes of borrowers during a typical mortgage journey.

And it’s not just the sheer volume of data that poses a challenge for mortgage underwriters; it’s the wide variety of structured and unstructured data that needs to be collected and processed. This includes data related to the borrower’s identity, employment status, income, credit history, proof of deposit, property details and much more.

An outdated reliance on email

Despite the prevalence of digital channels, many lenders and mortgage brokers still rely heavily on email to collect data and communicate with customers. While email is versatile, it’s not ideal for transmitting large amounts of data or engaging in lengthy exchanges. 

Furthermore, it’s risky to ask borrowers to send sensitive information to a generic email address like, as this could expose them to fraud and other security threats. Relying on email also slows down the mortgage process unnecessarily, as recipients have to sift through a large volume of messages.

Lenders and mortgage brokers can streamline their operations and improve the customer experience by shifting to more efficient and secure digital solutions.

Relying on email also slows down the mortgage process unnecessarily. Lenders and mortgage brokers can streamline their operations and improve the customer experience by shifting to more efficient and secure digital solutions.

It’s time for automation

There’s vast potential for further automation across the entire mortgage journey, from the mortgage-in-principle and application stage to offer and completion. Application to completion can take up to 40 days. For a nervous homebuyer, this can feel like a very long time. 

Admittedly, this isn’t all down to underwriting, but collecting and handling all the data required to process a mortgage application is time-consuming. It’s estimated that 90 per cent of an underwriter’s time is spent on processes that can be automated. It’s getting there, just slowly.

90% of an underwriter’s time is spent on processes that can be automated. 

Mortgage underwriters are typically well-educated, well-qualified and possess a solid set of mathematical, analytical and critical decision-making skills. Their role is to assess the risk of lending money to businesses and individuals and to ensure that lending is in line with the organisation’s lending policy rules. 

Why automate data processes for underwriters?

With all that in mind, why would you have them spending their time doing routine or repetitive tasks that can be automated? 

For example, suppose you could reduce the time an underwriter spends on data collection, verification and processing from days to seconds. That would enable them to spend more time on higher-value and complex applications like mortgages for the self-employed or those with a thin credit file.

A modern mortgage origination platform based on public cloud services can be a game changer. It digitises documents, manages structured and unstructured data, provides real-time analytics and machine learning, and offers customisable and automated workflow tools with API management and orchestration. 

This means that cloud-based services can handle the scale and flexibility that modern organisations require, especially in a constantly changing mortgage market landscape.

Where should mortgage lenders focus their efforts?

Here’s a few things to consider when thinking about automation within a mortgage origination platform:

1. Automating data extraction from documents

As mentioned above, lenders gain lots of different data for various reasons. For example, they’re required to confirm the identity of every mortgage applicant to regulatory standards. However, without commonly accepted and widespread digital ID systems, borrowers typically provide scanned copies of their passports (yes, even via email!). 

To help solve this, AWS, Azure, and Google Cloud have developed their own solutions that use optical character recognition, computer vision and natural language processing to extract text from images and turn it into structured data. 

2. Exposing yourself on the internet

If a recognised digital ID system was widely adopted, mortgage applicants would no longer need to scan and email passports. In the UK, an industry-led approach to digital ID is being piloted, while other countries like Finland and Germany are taking a broader, government-led approach.

Irrespective of the approach, the technical implementation will be very similar. It’ll allow financial services firms to connect to the identity provider via an API call in a way that protects the individual’s privacy. 

Modern mortgage lenders must design their platforms to support the secure integration of new data sources via APIs. However, this exposes the platform to security risks when connected to the internet. Fortunately, standard APIs like Open Banking are being further developed, eliminating the need for applicants to send sensitive data by scanning and emailing. 

This can enable a truly digital mortgage journey, eliminating the need for paperwork and allowing the platform to integrate with cloud services and other digital ecosystems. Such a shift creates exciting opportunities for innovative new mortgage propositions.

3. We’re all technologists now

For lenders, underwriting is a core capability crucial to their long-term profitability. Outsourcing mortgage underwriting to external service providers can lead lenders to lose control over the quality of underwriting and data security. But as technology increasingly automates underwriting processes, lenders mustn’t just outsource underwriting to their own technology and data science teams. 

As the underwriting and technology teams work together more closely, upskilling will be needed on both sides. At the very least, lenders will need underwriters who can understand the algorithms performing the mortgage decisioning and the technologies that underpin the automated workflows. 

In addition, the lender’s technology and data teams will need to invest in explainable artificial intelligence for regulated activities such as mortgages to help the business understand and explain to its stakeholders how automated decisions are made.

Fear not, dear underwriters; you may not need to learn python data structures and algorithms just yet. There’s still a lot to do before the floodwaters of email attachments recede from your inbox. Ultimately, the mortgage customer should see the real benefits of automation in terms of speed. The technology is already there to reduce the mortgage underwriting journey from 40 days to 40 seconds for most mortgage applications.

Time to automate?

Speed could be the real differentiator for lenders who get it right in a very price-sensitive market where customers routinely switch to find a better rate. Who knows? Maybe at that point, we may even need to talk about slowing down the mortgage process, giving borrowers more time to consider further the most significant financial commitment they are ever likely to make. 

Our data services are specifically designed for companies adopting a cloud native approach to infrastructure and application development. We’re here to help you maximise value from data flowing into and out of  digital services, so you can enhance performance. 

And we’re also focused on putting the right people, processes and culture in place to help you get maximum value from your cloud investments, fast.

Are you set up for digital resilience in the cloud?

For financial services, basic continuity planning has always been part of keeping up with regulations.

But two factors are driving business continuity up the agenda for FSI right now:

  • The current geopolitical instability in Europe
  • New regulations like DORA and NIS that require businesses to leverage digital technologies like the cloud to keep running in the event of a disruption or emergency

We have a cloud-based strategy to achieve business continuity - helping you use cloud to safeguard critical data and keep essential services running during disturbances.

It's called the Digital Vault. And our handy guide explains things a little more, and breaks down whether or not you're really resilient to potential risks. Read it here.

Get in Touch.

Let’s discuss how we can help with your cloud journey. Our experts are standing by to talk about your migration, modernisation, development and skills challenges.

Kris Soderlund
Kris Soderlund LinkedIn
Global FSI Lead