From Accumulators to Default Rates: Decision-Making Under Uncertainty 

The biggest lesson I learned from the sports betting industry wasn’t about sport. It was about making decisions under uncertainty.  

As my career moved from sports betting risk into SME lending, I realised both industries wrestle with a similar problem. How do you make a good decision when you don’t know what the future holds? 

One of the biggest things sports betting taught me is that you’re not trying to be right all the time. You’re trying to be right most of the time, while taking uncertainty into account when pricing risk. No punter, model or bookmaker gets every decision right. What matters is understanding the probabilities well enough that over time the good decisions outweigh the bad ones. That is where value is found. 

How does this translate across to SME lending. Although the tolerance for loss is a lot lower because the upside – the spread premium for covering losses – is a very low % on each deal, you’re still trying to make the right decisions within an uncertain environment. We assess businesses using the information available through financial statements, management information and industry conditions but ultimately, we’re still making a judgement about the future. We can build a picture of the risks, but we can’t predict the outcome with complete certainty. With a large portfolio it’s unrealistic to expect complete elimination of risk or to assume, even with the greatest prudence, that every bad outcome is avoided. It’s to assess risk systematically and accurately enough that the return justifies the risk being taken. From time to time that means being wrong but being wrong for the right reasons.  

Working in both industries has taught me some valuable lessons about risk, decision making and why process matters more than individual outcomes. 

Good outcomes don’t always mean good decisions

In sports betting, there were plenty of occasions when the result went against us, but we were still comfortable with the decision we’d made. We’d see competitors move their lines in the market and discuss whether we should react and move with them. Sometimes we’d adjust our price, other times the traders would look at the information available and decide they were happy where we were. The outcome might prove us wrong on occasion, but that didn’t necessarily mean the pricing decision was poor, we trusted our methodology and reasoning believed that long term it will provide success. 

I’ve seen similar situations in SME lending, although sometimes in reverse. A customer might look strong on paper, make every repayment on time and ultimately become a successful lease. However, during the term we might observe signs of distress that suggest the business was not as solid as we originally assessed. The positive outcome doesn’t necessarily mean the original decision was flawless. It just means that there may still be lessons to take from the weaknesses that only became apparent later and an opportunity to improve our process. 

In both cases, focusing solely on outcomes to assess decision quality can be lazy and misleading. A poor outcome doesn’t automatically mean the decision was wrong, just as a good outcome doesn’t always mean it was right. What’s more important is whether the decision was made using the right information, sound reasoning and a consistent process. Even when things go our way, there is often something to learn that can improve future decision-making. 

Pricing matters as much as the decision

Sports betting markets are built around probabilities. Margin is then added, which is effectively the reward for taking on uncertainty and risk. The objective isn’t simply to predict the winner, it’s to ensure the price makes sense given the likelihood of different outcomes. For example, in obscure sports markets such as table tennis, the margin added on top of the predicted fair value would be a higher than that of a match result in a Premier League game. This reflects the greater uncertainty and reduced availability of information. In other words, the bookmaker is compensated for taking on additional risk.  

I found something similar with SME lending. When we assess a business, we’re making a judgement about the probability of repayment, and ultimately calculating expected loss (EL) net of equipment and security recovery and the risks involved. This can be dependent on what industry our customers are in, if they seem to have cashflow issues or if they don’t have a long trading history, just to name a few. This rests on asset recovery economics and strength of additional security. This is then pricing.

At the end of the day both industries ask themselves the same important question. Is the risk worth the reward?

Systematic decision making 

One of the biggest lessons I took from sports betting is that long-term success comes from consistency over brilliance. 

Nobody in the sports betting industry is successful by relying on gut feel or trying to land the occasional big call. The objective is to make good decisions repeatedly over many events. Over time, success comes from pricing markets as accurately and consistently as possible, while relying on the fundamentals of statistics and probability rather than emotion. Being occasionally brilliant is far less valuable than being consistently good. We weren’t worried about customers that had that big once-off win, but more so the customers that had consistently found value across a lifetime of punts.

Lending has traditionally involved a degree of discretion. Alongside financial analysis and risk assessments, decisions can be influenced by subjective factors such as the “smell of a deal” or whether a director comes across well. One area where traditional asset finance differs from sports betting is in the mass use of information. Betting markets consume enormous amounts of data, incorporating millions of different datapoints into models and pricing decisions. Information beyond the game stats themselves, such as injury news, managerial changes or if someone made front page of the tabloids at the weekend, is rapidly reflected in prices. SME lending, often involves dealing with incomplete information. Financial statements can be outdated, management accounts may be limited or inaccurate and many of the factors that determine a business’s future success are difficult to quantify. However, that doesn’t necessarily mean information is unavailable. There is often a wider range of data that could be used to support decision-making other than the traditional financial statements alone. Sports betting has shown the value of systematically gathering, processing and acting on information.

At equipal we adopt similar values as we build on the practices of traditional SME lending. We are naturally curious about alternative data sources that can bridge the information gap, creating a more accurate evaluation of our customers, and in turn being able to make better credit decisions. Our proprietary models have been built on both open accounting and open banking, providing an up to date and comprehensive view of our customers’ financial position. Building on this, equipal is currently exploring the use of social media data as an additional source of insight. Through ‘Project Fanfare’, we will analyse an array of signals from social media and other online channels. This is Big Data in action, helping to drive more efficient (i.e. lower!) pricing for Customers.

At equipal, we believe the future of SME lending lies in combining traditional credit expertise with a broader range of data sources and a more systematic approach to decision-making. The goal is to guide judgement with a better information, leading to more consistent and accurate credit outcomes. By building a more complete picture of our customers we can make more informed, consistent and accurate lending decisions.

The challenge isn’t finding perfect information, it’s making the best possible decision with the information available. The more we can quantify, measure and assess consistently, the more likely we are to make better credit decisions over the long run. The challenge is finding the right balance. Data and models can provide consistency and reduce bias, while experienced credit judgement can capture nuances that are difficult to measure. For now, neither are sufficient on their own. Long-term success is likely to come from finding the right balance between both within a robust and repeatable decision-making process. Ultimately, helping SMEs to grow around the UK. 

Jordan Hoang has been on an internship at equipal in Credit and Data since March 2026. He has a degree in Statistics from UCD and a Masters in Computational Finance from UCL. Amongst other claims to fame, he holds Ireland’s U20 national record for the triple jump.