How traditional lenders can improve credit risk management to lend more to SMEs


May 12, 2022

When it comes to commercial lending, a strong approach to credit risk management is essential. Accurately and efficiently determining the creditworthiness of new or returning borrowers will help both traditional and newer lenders increase the amount they lend to businesses, in addition to also reducing the risk of late payments or defaults.

In our experience, most banks are limited to using traditional data sets, such as historical financials, personal experience of a particular sub-sector and, critically, no real-time data to provide forward looking insight. However, with more and better data, commercial lenders can find the ‘right path to yes’, expanding the range of businesses they lend to and uncovering new opportunities with their existing clients. With an increased level of understanding, relationship managers and analysts working at the banks will no doubt gain the confidence to lend into unfamiliar industries, supported by industry insights and peer comparisons.


 Take a forward look view

A typical high-growth business grows at 20% year on year, so the difference between its last 12 months and its next 12 months, is c.40%. So, if a lender is only lending to this business based on its historic data / past performance, they’re not lending to it based on what’s needed to support its current or future growth trajectory. This is akin to driving by only relying on what you can see in the rear-view mirror as opposed to driving looking ahead through the windshield.

It’s so important for lenders to use data and insights to build a clear picture of a business’ future growth potential and where it’s going, rather than simply relying on where it’s been. Unfortunately, so many lenders still see high-growth as high-risk, and because they’re unable to develop a reliable forward-look view of a business, they’re unable to get comfortable lending to them.


Data and analytical capabilities

At OakNorth, we’re firm believers that data and credit science can be leveraged by traditional lenders to enhance humans, rather than replace them. This hybrid approach is a pragmatic compromise where computers perform various tasks to allow the credit analyst to be more efficient, but the analyst remains in the driving seat and is able to train the models and direct and shape the final outputs to ensure they are coherent and understandable.

Due to the complexity of the space, we don't believe full automation is a desirable end goal, and aim instead to achieve c.80% automation, with a human analyst always involved in the process. This critically allows human judgement to always have an influence on the outcome and helps ensure understandability of outputs.

Furthermore, when it comes to commercial lending, traditional lenders need to assess the ability of a business to sustain a certain level of debt and repay loans. This is where data science comes in.

The only way they can effectively assess commercial credit risk is by using multiple data sources – including what may be unconventional or previously unavailable data – rather than just relying on what they’ve used in the past. At OakNorth, we apply data science techniques to create unique models that provide a granular level of analysis on each borrower. By combining borrower-provided data with our vast repository of external data, we are able to add depth to point-in-time analysis and monitoring.


Develop a granular, loan-level understanding of a business

Traditional lenders tend to lump all businesses into one of a dozen or so categories – for example, all restaurants, hotels, bars, leisure facilities, etc. fall under “hospitality and leisure”. There are a couple of issues with this approach – firstly, it ignores the unique differences between businesses within the same sector, and secondly, human bias can mean that certain relationship managers or teams are reluctant to lend to a specific business because it falls in a sector they have had a negative credit experience with in the past. This issue was thrown into the spotlight over the last two years with COVID as businesses within similar sectors or sub-sectors had vastly different experiences at different stages of the pandemic. Take a golf club and an indoor climbing centre in the same city for example – both would be classified as “physical leisure activities”, but their experiences over the last two years will have varied significantly. The golf club – which takes place outdoors, is played across huge social distances, and where players bring their own clubs – will likely have been able to re-open and begin trading much sooner than the indoor climbing centre which, being indoors, will have less fresh air, and which will see climbers touching the same climbing holds in quick succession.

Developing a granular, loan-level understanding of a business will therefore enable them to structure a facility that’s bespoke for that business’ needs.


Conduct events-based scenario analysis

As demonstrated by the COVID-19 pandemic, when it comes to adverse events, the traditional approach to commercial lending – using historical data, financial modelling of a base case, worst case and best-case scenario, and conducting annual reviews – is an approach that is not fit for purpose. In uneventful times, these models are fine. However, for unprecedented events such as the pandemic, the traditional models proved useless as historical correlations were broken; employing the traditional look-back approach was meaningless.

Commercial lenders need to be able to run a “bottoms-up” analysis of their loan books, assigning each business a vulnerability rating based on a subsector-specific, forward-looking credit scenario taking liquidity, debt capacity and profitability into account. This more dynamic view of risk is still valuable in a more stable economy because we can update risk inside a lender’s review cycles, allowing them to take a critical view of their loan book and maintain constant focus on the items of highest impact.

The realization that many industries experienced through COVID-19 is the same: we can’t predict the future but must be better prepared for the unknown and reduce risks across our businesses with an ability to adapt quickly with data-driven decision making. In doing so, banks will identify opportunities to lend faster, smarter and more to businesses.

About the author