If you torture the data long enough, it will confess to anything

The focal of today’s credit risk management awareness will centre on the importance to maintain data, uphold the accuracy of data, independent analysis of such data so as to ensure bias is non-existent and ensure results are confidently verified as valid so managerial decisions can be taken as a result of these data analysis.

Ronald H. Coase, a renowned British Economist is the author of the quote- “if you torture the data long enough, it will confess to anything”. He is a noble laureate awardee widely recognized for his seminal work on transaction costs, which reflects on some of the most fundamental concerns of economists over the past two centuries. How do organisations/ CEOs/ Credit Risk Managers decide what questions to address and how to choose their theories? How do they tackle the problems of the external environment happenings and give advice on public policy? With these broad questions, we will consider the work (quote) of Ronald Coase to encourage organisations in the positives associated with accurate data analysis in the Credit Risk Management processes.

According to Wikipedia, a theory is a group of linked ideas intended to explain something. They can be tested to provide support for or challenge the theory. The word ‘theory’ has several meanings: a guess or speculation. A law about things which cannot be seen directly, such as electrons or evolution. Theories are based on general principles independent of the thing to be explained.

Organizations have certain theories they uphold or beliefs in which by a large extent direct their processes, thoughts, mission and vision. The humans of organisations have their theories which in a biased way (if not intertwined with the company’s theory) interfere with their work process.

We are not interested simply in the accuracy of its predictions. A theory also serves as a base for thinking. It helps us to understand what is going on by enabling us to organize our thoughts. Faced with a choice between a theory which predicts well but gives us little insight into how the system works and one which gives us this insight but predicts badly, I would choose the latter, and I am inclined to think that most economists would do the same.

For the sake of sound credit delivery, decision making is not based on organisation theories, educational theories or the theory of the Head of Credit of the organisation. This is so because we are interested in both insights and predictions accuracy. As such we have just one fallback- DATA

Data analytics does not just give an accurate prediction into a matter but provides an insight into the problem in question. Data Analysis can be an extreme work for organisations with a lot of data and complex processes. An estimate of 75 percent of organisations in Nigeria lack the technology or strategy to effectively use their data (refer to the article on Risk Mgt and ICT).

Modern risk management is near impossible without gathering, storing and analysing data. Data analysis can help predict the outcome of a situation or allow you to protect your organisation against risk. It also has other benefits including mitigating repetitive losses, lowering insurance premiums, and more.

Analytics turn your data from useful to extremely effective information and allow you to make changes that will benefit your organisation as a whole. A survey by Deloitte found that 55% of organisations believe that analysis improves the organisation’s competitive position and 96% agree that it will continue to become more important over the next three years. Risk managers should utilize data analytics as they allow you to:

Avert reoccurring losses- Analytics helps identify red flags and trends that could be an issue and result to money loss. Identification helps initiate strategies and implementations for mutations and you can also detect if a certain area, department, or season has a particularly high claim occurrence and run a root-cause analysis to understand what’s going wrong and how you can fix it in the future.

Improve insurance pricing- Insurance companies are in business to make profits just as the insured company focuses on the same. As such, they strive to do business with risk-conscious organisations that are termed “good risk”: those that are likely to pay more in premiums than they require for loss coverage. A readied data analysis and mitigation strategies can be presented to the insurance company. They will be ready to do business with you and offer a more competitive rate leading to lower premiums.

Top-notch reporting- Accurate data analytics helps give in-depth knowledge and communication of updates in your organisation and industry. With analytics, issues in your organisation are diagnosed and fixed. Data will become actionable, understandable and support any business idea or strategy for risk mitigations.

Performance monitoring- Consistency in your analytics work will ensure you have actual picture/ patterns of how your credit book looks. Risk Managers will be able to hold units or departments accountable for exceeding or failing to meet goals, recognise red flags that may indicate something needs to be changed, or discover why a business strategy isn’t working out as well as planned. With this in-depth understanding, growth will be enabled while meeting goals and avoiding overly risky scenarios. An individual department will be more likely to work on issues if they are shown to be underperforming.

Forecasting and decision making- Without analytics, it’s difficult for risk managers to learn from the past or prepare for the future. They make it simple to improve the efficiency and effectiveness of any business. Understanding what happened in the past prepares you for likely incident in the future. A thorough risk plan based on data analysis will have you ready for almost anything. With analytics, you can track growth and performance which is key to subsequent decision making for achieving an organisation’s goal and objectives.

In conclusion, it is not only necessary to keep and analyse data, it is required that data interferences is not applied in your processes. This is defined as “data- torturing” in the words of Ronald H. Coase. Let it measure what is required of it and provide the result therein and not the result you wish to see.“Numbers are like people, if you torture them long enough, they’ll say whatever you want them to.”

Timothy Akinyomi an expert and well-trained individual in the field of credit risk management. He is the CEO of Tatoni Consults