Unlocking the power of transactional data to enhance “early warning & collections” strategies

Why is transactional data important right now?

Firstly, the potential of Open Banking within the credit industry is huge and the value of transactional data can support decisions right across the customer lifecycle to enable lenders to improve strategies and credit decisions.

Early warnings and collections are a focus for many organisations due to the impact of the pandemic. The consequences of Covid-19 and the recession, coupled with the ever-changing financial behaviour of customers, has made it harder to develop a solid strategy in early warnings and collections. Existing models have generally underperformed and therefore businesses are looking for more contextual data to improve analytical performance. This is where Open Banking has created an opportunity.

What problem does transactional data solve?

Put simply, it provides more accurate indicators to develop effective strategies. This all starts with the process of categorisation. At Experian we’ve developed Trusso, our AI powered categorisation engine, which classifies transactional data with high accuracy (+90%) and this allows us to better understand the spending behaviour of consumers and SME businesses.

With categorisation, it is possible to generate actionable insights specifically designed to support the generation of early warning strategies, and to develop KPIs that analyse and predict early vulnerability and identify how to implement collections.

We’ve now developed a set of specific key performance indicators to create early warnings through observing the health of the account, the evolution of the balance, and the number of days in overdraft or below a pre-defined threshold. Accurate warnings stem from the analysis of the key categories where payment frequency is combined with the trend of monetary means.

It is also possible to generate a set of useful ratios to monitor the payment capacity, the savings capacity, and the debt-to-income ratio plus other indicators that are recommended by the local regulators. This is of course very useful for underwriting teams when assessing files.

How is the transactional data collected?

Data is obtained in two ways. It can be analysed from the client’s own internal customer data, or thanks to a secure end-to-end Open Banking platform, we can collect consent from the customer in a one-off permission or with recurring access. One-off consent allows us to download the full historical depth of one or more current accounts to generate a comprehensive view of past vulnerability signals and then use this to forecast the future to identify early warnings and even the best dates to charge instalments based on the account performance. And as the name suggests, recurring consent can be used to monitor the user on a regular basis so you can become much more accurate and proactive in supporting the customer.

What are the key strengths of using transactional data for early warnings and collections?

The key strength is related to the capability to provide a truly holistic view of the spending behaviour of a consumer or an SME, evaluating a large number of parameters that help establish the probability of failure when it comes to the re-payment of an instalment. This is done not only by looking at the past, but by forecasting the future outlook of the aggregates that inform the affordability or cashflow assessment. You can also execute stress tests to understand and predict the vulnerabilities if one or more income sources decline.

The precision of the categorisation provided by Trusso means we can ensure a holistic view of all aspects of the net income of a consumer, a sole trader, or a small business, to clearly define when income is highest and therefore the best time to charge instalments, helping to implement a high performing collections strategy.

How can the risk department consume this comprehensive set of insights?

Experian have designed a system that can be easily used in any context and adapted to suit different policies. The early warning and collections KPIs and aggregates can be ingested simply via JSON or XML, and the insight can be visualized in an intuitive dashboard together with the transactional score, the meaningfulness index and other essential transactional data-driven information to assess the credit risk. It can even be provided as a pdf fact-sheet.

Alternatively, we provide a pre-configured decision engine or a data sandbox platform where models can be created and enhanced with this additional data and easily exported onto the client’s own system. And of course, we also support clients with our consultancy practice to help them get the maximum value from transactional data for analytics and decisioning.

Download the use case brochure using the link above or explore our Open Banking page.