Big data unlocks funding for mid-sized SMEs
Today, leaders in every industry must understand the implications of big data and the right ways to harness it.
Data can drive an abundance of opportunities, from fuelling growth and innovation to underpinning strategic decision-making. But in alternative finance, data is crucial to the business’s purpose. And for the mid-sized SMEs that ThinCats serves, the impact of big data can be felt even more powerfully.
Although ThinCats’ analytics team is relatively small, we have already had a significant impact on several areas of the business. We work with “big data” - meaning both granular and aggregated public data on half a million mid-market firms – and “micro data” meaning more detailed information that we collect on our portfolio of around 200 borrowers within this mid-market. The combination of both datasets enables us to provide the best outcomes for our clients.
The big data universe: accuracy in credit scoring
My team’s main output is PRISM, our proprietary credit risk model. Used predominantly in origination, it supports the business development managers (BDMs) by enabling them to firstly avoid wasting time and effort on firms that would not meet our risk appetite and then, for those that are credit-worthy, to instantly calculate an indicative price that is risk-based and so equitable to all borrowers.
The BDMs must be confident that the indicative price will in most cases reflect the actual price that makes it onto the term sheet – only if significant new information comes to light should the price change. PRISM allows us to screen out high-risk businesses and - through APIs - instantly pull in up-to-date financial and Credit Bureau data (from Experian and Companies House) and turn that into a credit score within milliseconds.
This forms the basis for our risk-based pricing. Because we have built the model inhouse – and optimised it for mid-sized SMEs – we can give much more detail to the underwriting team, explaining why a particular company is scoring as it is. What are the positive and negative factors contributing to that score and where should they focus attention in their underwriting?
The borrower-level data: deeper insights build relationships
Then there’s the borrower-level information, where we have got a lot more granularity on the population that we lend to.
We are developing a toolkit for the relationship managers who look after the borrowers once they have been onboarded, which will enable a rapid understanding of the progress of each borrower and their strengths and weaknesses at any point in time.
Traditionally, each relationship manager would manage the accounts of their portfolio of borrowers by receiving and analysing management information and having the necessary conversations if there are breaches of financial covenants or missed payments, for example.
The goal is to move to a much more proactive approach. Using our models built from credit bureau data, we can give relationship managers foresight of potential issues that they may otherwise not be aware of – for instance, if a borrower gets into difficulty with one of its other credit agreements, goes into arrears on another loan, has missed payments or gone overdrawn on a current account
We are also automating the capture of management information that borrowers supply to us and tracking trends in key metrics. If cash or profit margins are trending below forecast or are lower than the equivalent period last year for example, or approaching the minimum covenant that has been set, we will flag up a red or amber warning to investigate the circumstances and have discussions with the borrower if necessary.
The next stage is to use real-time open banking data which will drive even more timely alerts.
What does this mean for our clients? Of course, in the current economic environment, the sooner we can identify any potential problems, the more scope there is to work together to take corrective action. But we also want to use this toolkit to drive positive interactions – for example where a borrower is growing profitably and ahead of forecast, there may be opportunities to explore further lending to facilitate acquisitions or accelerate organic growth even further.
The bigger picture
While it is operationally important for us to flag issues up at the individual borrower level, once we have this borrower-by-borrower view, we can also aggregate up to a portfolio level. This allows us to generate deeper sector-based insights by understanding segments that are performing particularly strongly and those that are lagging. We can also carry out benchmarking: being able to advise our clients how their performance compares against their peers is a powerful tool that will help us deepen relationships as advisors as well as finance providers.
Through data and analytics, we can operate in a more systemic way, which should mean we never miss out on an opportunity to have conversations that provide genuine support to our clients.
Although data analytics provides statistical robustness to our lending decisions in this critical part of the SME market, it will always be used alongside the judgment of our human talent. Our clients’ businesses are complex - they need human beings to fully understand their needs - but by equipping those human beings with accurate, insightful data we can deliver the best possible results for our clients.
