ReportTitle_SRQ&A.jpg

Q&A: Data analytics and credit decisioning capabilities

May 2021  |  SPECIAL REPORT: BUSINESS STRATEGY & OPERATIONS

Financier Worldwide Magazine

May 2021 Issue

FW discusses data analytics and credit decisioning capabilities with Chris Neale and Damian Hales at Deloitte.

FW: How would you describe the uptake of data analytics technology among financial institutions (FIs)? To what extent are you seeing growing awareness and demand for innovations that can improve credit decisioning in particular?

Neale: Data analytics technology plays a critical role in financial institutions (FIs) and is constantly evolving to meet the new requirements and needs of the business. Historically, data and analytics was seen as a technical capability, but there is now an awareness that to deliver the right customer experience and develop customer intimacy, there is a need for analytics and insight across every customer journey and every customer interaction, elevating the role of data and embedding analytics as a core capability across every function. Credit teams have always been at the forefront of analytics innovation, and this is still true today – they have an insatiable demand for faster, better and more timely data and insights, leveraging new technology platforms and services to develop more accurate models, forecasts and predictions, helping FIs to generate more revenue, manage risk and enhance the customer experience.

Hales: The uptake of data analytics among FIs has increased year-on-year. This uptake has increased exponentially, however, due to coronavirus (COVID-19). More FIs are gathering more data and using more sophisticated analytics techniques to understand the credit, impairment and capital risks that COVID-19 has caused. They are also turning to data analytics to streamline their credit decisioning processes so that there is less reliance on humans and because more FI customers are becoming more and more familiar with digital interactions and processes. Therefore, there is a huge growing awareness and demand for further credit decisioning innovations to push this even further.

FW: What are the key advantages of data analytics for FIs? In what ways can artificial intelligence (AI) enhance the credit decisioning process?

Hales: The key advantage of data analytics for FIs is the potential to turn raw data into insight. Artificial intelligence (AI) techniques, if appropriately controlled, have the potential to turn more raw data into greater insight, on customers, their spending and borrowing habits and risk profile. This in turn can lead to enhancements to product and customer propositions, customer targeting, pricing, limit management and portfolio management strategies.

Neale: As technology has evolved and with the advent of big data and more sophisticated analytics, FIs are now able to process a greater volume and variety of data, in real-time, than ever before. This enables FIs to understand their customers, their behaviour, their detailed financial position and their market in more detail and turn this data into actionable insight that allows for better credit decisioning, faster disbursements and ultimately better customer outcomes. For example, FIs can now assess a customer’s general ledger in real-time by ingesting data through application programme interfaces (APIs) from cloud-based accounting packages, providing an up-to-date and accurate view of their financial position for credit decisioning. AI can be used for automation and to remove all the friction from the customer journey through analysing and extracting data and insight from financial statements and invoices to feed into the credit decisioning process. Credit models can also leverage AI to learn and adapt over time, removing biases and making more accurate predictions on probability of default, loss given default and exposure at default.

“If you do not understand your customer or their circumstances, then the credit decision that you make will be by definition sub-optimal.”
— Damian Hales

FW: With data key to making the right credit decisions, what steps should FIs take to ensure the quality and integrity of the data they collect?

Neale: Data management and governance will become a key differentiator for FIs – those that do it well will thrive and generate maximum value from their data, and those that do not will fail and make poor credit decisions. As credit models are enhanced to analyse a broader set of data, it is critical that FIs have an appropriate data governance and control framework in place to provide accountability and ownership over the data, as well as agreeing standards, definitions and data quality metrics and reporting. This needs to be business driven and owned, rather than through the chief information officer or chief technology officer – the responsibility of data lies with business owners.

Hales: The term ‘garbage in, garbage out’ as an adage could have been coined for the credit decisioning use case. If you do not understand your customer or their circumstances, then the credit decision that you make will be by definition sub-optimal. Therefore, it is critically important that the quality and integrity of data used in the decisioning process is as good as it could possibly be, and to ensure this, a robust data control framework needs to be in place

FW: What legal and regulatory considerations need to made with regard to the collection and processing of data for credit decisioning? How can FIs demonstrate they have robust compliance procedures in place?

Hales: The three key legal and regulatory considerations when it comes to collection and processing of data for credit decisioning are: use of data, bias and accuracy of data. It is essential that the data is being used for the agreed purpose based on the terms and conditions of the product. It is essential that the data being used does not create bias in decisioning, avoiding discrimination based on sex and race, among other things. The data must also be accurate, with an individual having the ability to correct any data inaccuracies that may adversely impact their credit. To address these risks and demonstrate compliance, it is essential that FIs have a well-defined control framework in place, that is monitored on an ongoing basis, combined with an outcome testing framework.

Neale: As with any data initiative, FIs need to consider a number of regulatory and compliance requirements. Firstly, they need customer consent that they are able to use the data for modelling and analysis purposes – in the EU, this is required under the General Data Protection Regulation (GDPR). Secondly, regulators will want evidence that FIs understand the data lineage and data controls for all data used in credit models, so the FI needs to ensure the data being used for modelling and analytics is of the right quality, is fit for purpose and comes from a trusted source within the organisation. Finally, where AI is being used to automate credit decisioning, FIs will need to ensure that the algorithm is understood, is operating as expected and is delivering a fair outcome for customers, and not introducing bias or discriminating against specific customers.

FW: How would you characterise the relationship between data analytics and product innovation? What new capabilities do technology platforms offer FIs?

Neale: I have seen the clear link between data and product innovation. There is a virtuous circle between developing and launching products, customers using those products and generating a digital ‘exhaust’ in the form of data which, in turn, can be analysed and turned into insight then fed back into product innovation and to improve the customer experience. As with all things data-related, the greater the volume of data, the more accurate and useful the insights. FIs now have access to virtually unlimited computing and storage through leveraging cloud platforms, as well as a range of powerful services that help turn this vast data into useful insights.

Hales: Data is both an enabler of any successful product and proposition innovation, but also a driver for innovation. The innovation in invoice discounting and some small and medium-sized enterprise (SME) lending has been enabled by real-time access to data in client enterprise resource planning (ERP) systems. Open banking and the ability to obtain real-time transaction data has transformed credit decisioning for new customers.

“Data analytics is already an indispensable part of the credit decisioning process, however we are entering a new paradigm where data will play an even greater role.”
— Chris Neale

FW: What essential advice would you offer to FIs looking to integrate data analytics to improve their credit decisioning capabilities? What are some of the common risks and challenges that need to managed to ensure a successful process?

Hales: FIs need to think about the proposition – is this digital only, or does the customer have channel flexibility? This will materially impact the decisioning capabilities and data architecture required. Back testing is another consideration. How will additional data be assessed to ensure financial prudence? In this regard, FIs should start small and thoroughly test the end-to-end architecture. FIs also need to think about the operational risk associated with the end-to-end process and develop a digital control framework that supports it, rather than relying on the existing control framework.

Neale: The integration of data analytics to improve credit decisioning capabilities should not be approached as a technology challenge – that is the first and most common mistake. Credit decisioning is there to fulfil a customer need, so think about the end state vision, the end-to-end customer journey and customer experience that you want to deliver, and then determine the appropriate technology components needed to deliver that experience and vision. In my experience, the other common challenge is data architecture – FIs will need a modern, scalable, event-based architecture to meet their data and analytical needs. As the use of AI increases, one of the key challenges firms will face is making sure their algorithms and analytical models are fair and not introducing bias or discrimination into the decision-making process. Having explainable AI will be key for both customer trust and regulatory compliance.

FW: What are your predictions for data analytics uptake by FIs in the months and years ahead? Do you expect it to become an integral, indispensable part of the credit decisioning process?

Neale: Data analytics is already an indispensable part of the credit decisioning process, however we are entering a new paradigm where data will play an even greater role. In the future, FIs will move away from analysing traditional financial data sources which provide limited value, to a world where they have a whole ecosystem of alternative data available to enable real-time credit decisioning and even proactive, pre-approved credit offers to customers, delivered at exactly the time they need it. This ecosystem will enable FIs to ingest social, news, internet of things (IoT) and industry data, and combine it with richer financial data to make better credit decisions and to more accurately predict losses and failures. One theme that I expect to continue to grow in significance is environmental, social and governance (ESG), and my prediction for the next few years is a customer’s ESG credentials will start to influence credit decisioning and pricing, with FIs offering more attractive pricing to those customers with a better ESG rating.

Hales: Data has always been an integral and indispensable part of the credit decisioning process, but what has changed and will continue to change is the breadth and depth of the data, new analytical techniques to get insights from this data, and speed of decisioning and disbursement of funds. The breadth of data is likely to increase further as alternative data sources, such as news and social media, become standard credit sources. The depth of data is likely to increase as more and more transaction analysis will take place with less focus on aggregated variables, such as current account turnover. Artificial learning models were the realm of fraud systems historically but are now commonplace in credit decisioning models and will likely extend into impairment and capital models. The speed of decisioning and disbursement has moved from days to hours and will firmly be measured in the seconds in the future.

 

Chris Neale is a partner in the analytics practice and helps clients generate value and deliver innovative transformational change through the use of information and data. He leads Deloitte’s commercial banking analytics team and has 19 years’ experience in a range of roles across the UK, EMEA, North America and APAC. He has experience in transforming data organisations, deploying digital and disruptive solutions, generating new insights through analytics and AI, and risk and regulatory reporting. He can be contacted on +44 (0)20 7007 7201 or by email: cneale@deloitte.co.uk.

Damian Hales is a partner within the risk and regulation practice and has over 25 years’ experience in the financial sector, specialising in credit risk management across the full credit lifecycle from origination to collections and recoveries, IFRS 9, capital management and programme management. He leads Deloitte’s credit risk transformation offerings, including documentation intelligence, web-based credit risk sensing, corporate KPI stress testing, and credit and impairment regulatory reporting managed service. He can be contacted on +44 (0)20 7007 7914 or by email: dhales@deloitte.co.uk.

© Financier Worldwide

THE PANELLISTS

Chris Neale

Damian Hales

Deloitte

©2001-2024 Financier Worldwide Ltd. All rights reserved. Any statements expressed on this website are understood to be general opinions and should not be relied upon as legal, financial or any other form of professional advice. Opinions expressed do not necessarily represent the views of the authors’ current or previous employers, or clients. The publisher, authors and authors' firms are not responsible for any loss third parties may suffer in connection with information or materials presented on this website, or use of any such information or materials by any third parties.