The hidden cost of over-caution in commercial lending

When I speak with relationship managers and credit risk teams at UK banks, a few themes come up time and again. One, more than any other, is: the loans they wanted to do but couldn’t get past the risk committee.

I understand the instinct. The last five years have been genuinely turbulent, and the hangover from COVID is still shaping how banks price and assess SME risk. But there’s a cost to that caution that rarely shows up in the numbers boards look at, and it’s bigger than most people are comfortable admitting.

According to the British Business Bank’s Small Business Finance Markets Report, the loan success rate for SMEs applying to the seven largest UK banks fell to 45% in Q2 2023. A decade ago, that figure sat consistently above 80%.

Boards track defaults. They track impairments and provisions. What they rarely quantify is the cost of the loans that were never made, the relationships that walked out the door, the revenue that went somewhere else. That invisible cost is what I want to talk about.

The over-correction that outlasted its justification

The caution made sense at the time. COVID scrambled the financial profiles of hundreds of thousands of UK businesses almost overnight. Furlough, Bounce Back Loans, and state aid all created noise in the data that made it genuinely hard to tell which businesses were fundamentally sound and which were being kept alive by emergency support. Risk teams tightened their criteria and applied more conservative overlays – a very reasonable response to an abnormal situation.

The problem is that many of those overlays are still in place, years after the conditions that justified them have passed. The emergency posture has become the default posture. And the businesses bearing the cost are the ones that banks should most want to lend to: the SME and mid-market segment, which represents the largest share of UK companies and, in most cases, the biggest growth opportunity sitting in a relationship manager’s book.

And that business doesn’t disappear. It goes elsewhere.

Where the business is going

This is the part that should concern commercial banking leaders. Challenger and specialist banks now account for 59% of total gross SME lending in the UK, exceeding the big five for the third consecutive year. That is not a coincidence. Alternative lenders have moved deliberately into the space that traditional banks have vacated. They offer speed, flexibility, and are increasingly sophisticated in how they assess credit risk for smaller borrowers.

Peer-to-peer platforms, private debt funds, invoice finance providers: all of them have built products specifically for the SME that a traditional credit model struggles to accommodate. Every declined application that finds a home with one of these providers is a data point in a pattern that, over time, reshapes who owns the SME relationship in UK commercial banking.

The cost of over-caution isn’t just the interest income foregone on one facility. It’s a slow, largely invisible erosion of market position in a segment that banks cannot afford to cede.

The data problem behind the decision

Here is the part I find most important, because it changes the conversation from one about appetite to one about information.

A lot of the caution I see in SME lending decisions is rooted in a genuine limitation: the data available on smaller businesses is patchy. The incumbent risk providers that most banks rely on were built primarily to serve large corporate credit decisions. Their coverage of SMEs, particularly micro and small businesses with shorter filing histories or abbreviated accounts, is thinner than many relationship managers realise.

When an underwriter runs a check and gets back limited data, it’s a tougher task to write the risk. But limited data is not the same as a bad risk. It’s a gap in information, and that distinction matters enormously.

What’s needed are tools that can estimate financial health where hard data is thin. That means working with abbreviated accounts, identifying proxy indicators of profitability, and updating assessments in something closer to real time rather than waiting for the next annual filing cycle.

What better intelligence actually makes possible

This is where I’d point to what we’ve built at Company Watch, because it speaks directly to these gaps.

Our H-Score® was designed with the SME problem in mind. It forecasts the probability of financial distress even when filings are incomplete, delayed, or abbreviated, precisely the scenario that trips up many of the traditional scoring models. Where other tools return a blank or a low-confidence output when data is sparse, H-Score® estimates a company’s underlying financial position based on what is available. For a relationship manager trying to quickly assess a new client’s financial health, or for an underwriter needing a credible starting point before a credit decision, that difference is material.

The other challenge I hear about consistently is speed. Onboarding in commercial banking can be painfully slow, a friction point that costs deals and strains relationships before they’ve even started. Part of the problem is that risk assessments are running on static data: a score generated at the point of application that doesn’t reflect what happened to the business last month or last quarter.

Our API and Cloud Data Access platform lets banks embed real-time financial data, risk scores, and business intelligence directly into their existing workflows and systems. For compliance teams needing comprehensive UK company due diligence reports at scale, or relationship managers monitoring a large book, live integration of this kind is what separates fast, confident decisions from slow, uncertain ones.

Spot financial weakness before it hits.

Spot hidden risk early. Our flagship H-Score® gives you a predictive, data-driven business credit rating check that flags financial weakness before it turns into loss.

Discover H-Score® now

The framing that changes the conversation

I want to be clear about what I’m not saying. The answer to declining approval rates is not lower credit standards. That’s not a position I’d advocate, and it wouldn’t serve anyone in the long run.

What I’m suggesting is that the quality of a risk decision is only as good as the quality of the information behind it. If the tools being used to assess SME borrowers weren’t built for SME borrowers, the decisions coming out of that process will be systematically skewed toward false negatives: good businesses declined, relationships lost, revenue left on the table. None of it showing up on the board’s impairment report, but all of it showing up, eventually, in market share.

Better financial intelligence doesn’t just protect against bad loans. It gives relationship managers the confidence to approve good ones faster. That’s a different way to frame the investment case for better data. It’s not just a risk management cost. It’s a revenue enabler.

The banks that address the data quality problem first will be the ones that retain and grow the SME relationships that matter over the next decade. The ones that don’t will keep watching that business land on someone else’s balance sheet, one declined application at a time.