With stricter banking regulations since 2008, Brexit uncertainty and sparse data available on small-medium enterprises (SMEs, mainstream lending sources are not sufficiently servicing the credit appetite of UK businesses. The size of this ‘SME funding gap’ is estimated to increase to £22 billion by 2017. This is a major factor in the growth of the alternative finance (alt-fi) industry, with many alternative business funders focusing specifically on SMEs.
At this nascent stage in the industry’s development, it is vital for these lenders to develop healthy portfolios, not just for their investors, but to build the reputation of the alt-fi industry as a whole. Peer-to-peer lending accounted for 13.9% of new bank loans to SMEs in 2015. This leaves plenty of room for further growth; however, without reliable data, it will be a difficult trajectory.
Without the resources and name recognition of the leading banks, alt-fi lenders are continuously innovating to gain market share – there has been an explosion of credit risk modelling innovation in the SME finance market segment. To complement this, Commercial Credit Data Sharing (CCDS) is expected to kick off later this year; which will require nine major banks to share credit information on all their (willing) SME clients. This influx of new current account and credit information will help the alt-fi industry to continue serving a host of currently overlooked smaller businesses.
It is, of course, the SMEs with good prospects and in urgent need of finance to whom we most want to provide lending capital. Data scientists can identify and further refine this segment of SMEs, to pinpoint those that are currently overlooked by the banking industry. ThinCats is launching its first targeting model, whose algorithm is aimed at identifying these UK SMEs. By focusing origination efforts on this segment, ThinCats can grow many more mutually beneficial arrangements with SMEs.
The algorithm is an ensemble of two models: one that assesses a company’s need for borrowing, the other assessing their risk profile. Combining these ensures origination effort is focused on appropriate companies. The algorithm is trained, using publicly available data, on a UK SME ‘universe’ of borrowing and non-borrowing enterprises. The borrowing model considers a firm’s past behaviour and access to other funding sources to predict their desire to borrow over the next 12 months.
The risk model initially examines balance sheet characteristics but, crucially, ThinCats differentiates itself by including proprietary risk metrics that are not used by other lenders. These can enhance the view of firms undervalued by mainstream banking. For example, our model will highlight the rating of firms that demonstrate adaptability to market fluctuations. The intended result is a cohort of healthy SMEs that are poised for successful growth, yet still might not tick the boxes of a banking credit model.
This comprehensive modelling approach is part of ThinCats’ effort to provide the UK’s best SMEs with the finance they need to flourish. Since 2008, banking finance has been difficult to obtain and often slow. ThinCats offers an open door and faster service to firms that don’t ‘fit the mould’ of mainstream banking, by thinking laterally and picking up the SME vibrations that no one else is hearing.
Vyas Adhikari, Credit Analyst ThinCats