Credit Kudos is a complete transactional decision toolchain that enables lenders to address a far larger borrower cohort whilst decreasing false positives.
Financial Aggregation is by no means a new concept. Personal Finance Management (PFM) tools using “screen-scraping” to collect financial information have existed since the early 2000’s. Whilst receiving traction with early adopters, PFMs suffer from notoriously low conversion rates with many users reluctant to share credentials with a third party. Open Banking will eventually remove the requirement for any sharing of sensitive information, putting in its place a consistent OAuth authentication and authorisation flow. Implementing the rest of the customer journey is however entirely the task of the TPP, which is expected to spawn a raft of varied and possibly inconsistent experiences.
Backed by hundreds of hours of research and testing, Connect packages our best-in-class aggregation process as a portable library that partners can integrate in minutes. Customers using Connect see a consistent co-branded experience which can be easily customised to work inside an existing lending journey.
In addition to providing a single point of integration, the Connect library unifies output from ASPSPs, yielding a single consistent Application programming interface (API). Normalisation and Named-entity recognition (NER) are all handled by our Categorisation Engine.
Affordability requirements are the most recent incarnation of regulatory guidance to come out of the Financial Conduct Authority (FCA)1. Britain has a growing debt problem at levels not seen since the global financial crisis almost a decade ago2. The lack of strong affordability checks arguably led to the crash and their purpose is to prevent such a disaster from recurring.
The FCA’s guidance is clear: it will not be sufficient for lenders to rely on a statement of income without independent evidence when reviewing applications. The challenge and opportunity now is for lenders and brokers to stand out and offer their customers the most streamlined digital journey. This will not only protect their own bottom line but also the welfare of their customers.
We see affordability as a holistic view into the financial health of an individual. It is more than just income and expenditure analysis; it’s the integration of financial data into a more complete view of an individual applicant. According to the FCA, affordability is to “consider any likely material reduction in income” and “establish or estimate non-discretionary expenditure (i.e. disposable income)”3. Understanding these two pieces in particular is not a trivial undertaking, in that:
Most aggregation solutions on the market offer between two and three months of historical data for an individual. This is insufficient to identify income accurately. We’ve been generating observations on 12 months of data for each applicant for some time. Having built our models on such a large dataset, we’re better positioned than most to transition to the same quantity of data available through Open Banking APIs.
Greater quantities of data collected leads to more accurately categorised transactions. It is less about algorithmic quality than it is about data quantity. Yet, even once categorised, transactions must be put in relation to others. By building a categorisation framework on common standards, rigorously developed by the Office for National Statistics (ONS) and championed by the Money Advice Service, we have a framework designed with lending in mind and a standard for understanding the affordability of any individual.
It isn’t enough to just consider if [the borrower] will repay, we must also consider whether this will put them in distress — for example by deprioritising important household bills to cover repayments.
Taking a simplistic view of underwriting, we seek to answer two key questions:
Traditionally, lenders have sought to answer these questions using a combination of borrower-contributed information and a credit report. Increasingly however, credit reports lack sufficient information to make informed decisions, particularly when serving a near-prime market. Account transaction data, like that exposed by Open Banking through Credit Kudos Connect, can provide far more granularity than previously available. The challenge to the lender is how to make sense of this influx of new information whilst maintaining quality and efficiency of service.
In order to understand the influence financial activity has on propensity to repay, Credit Kudos has developed Financial Behaviour Score (FBS). The FBS uses Supervised Machine Learning (ML) to marry financial activity (“Observations”) with lending performance (“Labels”) as shown below.
Providing the best possible experience to our customers whilst satisfying legal, fraud, risk and compliance requirements presents a unique challenge. The opportunity to quickly and effectively gather identity, income, expenditure from a certified source without disrupting conversion is exciting indeed.