Why bespoke funding needs a custom-built credit model

The ability to assess credit risk is fundamental for all lenders. It is an intricate mix of art and science, whereby each lender has its own view when determining their appetite to lend and at what price. While there are some commonalities between each lender’s credit risk models, what is deemed a great business by one may appear less attractive to another. The enduring search for new factors to incorporate within a lender’s risk assessment process not only drives potential competitive advantage, but also provides the opportunity for businesses previously starved of growth capital to secure valuable funding.

The insight produced by ThinCats’ credit model, PRISM Risk, is the culmination of bringing together both financial and non-financial information, using advanced predictive modelling techniques. We search exhaustively for businesses that we determine to be healthy but may not be recognised as such by other lenders. Our overall credit philosophy, which draws on the analysis produced by our credit model, is to take an holistic view of a business as a long-term investor rather than as a transactional-only lender.

Limitations of traditional credit models

When establishing our in-house underwriting function, we researched a number of “off-the-shelf” credit scoring solutions from some of the leading credit agencies. None of them matched all of our requirements typically failing on one or more of the following:

  • Not specific to our market - Most third party credit scoring solutions cover the universe of all limited companies from sole traders up to those listed on the stock market. We wanted a more precise tool that could better serve mid-market SMEs.
  • Static view - Businesses go through natural lifecycles, but the credit models that we considered could not capture this to our satisfaction.
  • Lack of Integration with proprietary non-financial information – We focus on a wide range of proprietary non-financial metrics which provides an extra level of insight on a business’s creditworthiness. We needed a solution that could maximise the benefit of this data.

Having assessed the market for third-party credit scoring models, we were left with one choice – build it ourselves.

PRISM Risk credit model

The risk model we have created is designed to credit score mid-market SMEs with assets between £0.5m and £40m covering all sectors and regions of the UK.

To build the PRISM Risk model we analysed data from more than 700,000 SMEs over the last 12 years incorporating more than 200 different variables. Using machine-learning techniques, the model predicts the likelihood of a business becoming insolvent. An iterative modelling approach also allows new learnings to be implemented quickly. For example, a new enhancement will analyse a business’s financial strength by comparing it against a peer group of relevant competitors.

The non-financial data that feeds into PRISM Risk includes metrics such as the age of the business, its invoice payment performance and even the churn rate of its directors.

The model also takes account of a company’s lifecycle by assessing its growth profile ranging from very strong growth to very strong decline. By incorporating this additional information within PRISM Risk, we find that growing companies that struggle to secure funding from traditional sources, often exhibit much lower levels of risk than indicated by traditional lending models.

Designed for the new economy

PRISM Risk has a unique way of gauging the strength of security that sits behind a loan. This is particularly relevant for the cashflow secured loans that we fund where businesses won’t have significant physical assets; an increasingly common occurrence as the UK continues its transition to a more service-based, digital economy.

Using our 1 to 5 padlock security grading process (5 padlocks represents the highest strength of security) PRISM Risk scores businesses on factors such as customer concentration and the certainty of recurring income from existing contracts. These can markedly increase a business’s security strength score, enabling us to lend either at a lower rate of interest or often when others can’t lend at all.

Which businesses best suit the ThinCats credit risk model?

We built PRISM Risk to do a better job of understanding mid-market SMEs. By focusing purely on companies that make up this segment, not only do we support those businesses served by traditional lenders, but we also support those types of business that traditional lenders find much harder to fund such as:

  • Growing companies – We call these businesses the deserving underserved. By including a range of non-financial data, we can often better support growth plans that traditional lenders may not.
  • Asset-light businesses – The changing nature of the UK economy means that more and more companies simply don’t have physical assets to use as collateral, therefore need lenders with a credit assessment policy that has evolved to assess future cash flow alongside assets.
  • Businesses in transition – Event-driven deals such as acquisitions, MBOs and MBIs require a flexible and holistic risk-assessment process not well suited to previous one-size-fits-all credit models.

ThinCats came into existence because of the lack of funding being provided to the UK’s mid-market SME sector, which accounts for approximately a third of the UK’s economic activity. Our ambition is to help these businesses become as successful as they possibly can be through sharing our business insight and providing the most relevant funding solutions. The creation and ongoing investment in sophisticated, data-led credit models such as PRISM Risk is essential in helping us achieve this.