Executives don’t trust new data techniques
A study conducted by Deloitte Financial Advisory Services polling more than 2600 executives found that less than a third of respondents were deploying the kind of data analytics tools that can detect fraud or waste by vendors and or employees. Another 13% had the necessary tools but were still learning to use them, while 22% had no data analytics of any kind. The study drew upon conclusions that many managers do not understand “the power of analytics and many companies harbour an attitude of ‘it can’t happen here” (Bowman, 2014).
In an increasingly competitive world, where more services are offered via the Internet, the instances of fraud and opportunities for fraud are reported to be increasing dramatically. Under these cost-focused and tight-margin business conditions, managing the costs of fraud can be the difference between profit and loss. The organisations that reduce the risk of fraud gain an important competitive advantage over those that don’t.
Surprisingly, the study found that data analysis is not used very much to detect corporate fraud. Considering the reams of data that is now commonly available from data and technology companies such as credit bureaus, you would believe that adopting new data techniques to detect and prevent fraud would be a slam dunk for any organisation operating in South Africa in the wake of recent corporate fraud and State capture scandals.
Only one third of companies polled were deploying the kind of data analytics tools that can detect fraud or waste by vendors and or employees. Another 13% had the necessary tools but were still learning to use them, while 22% had no data analytics of any kind (Bowman, 2014).
In South Africa however, the number of organisations that use new data techniques to detect and prevent fraud suggest that many companies are not sold on the idea. Some do not understand what analytics can do for them. Others balk at the expense. And still, others feel they do not need it unless a major fraud actually occurs. If a detection program is going to succeed, it must have access to reliable data and be trusted to perform according to expectations. Executives must have confidence that the analytics will work.
Fraud among employees, suppliers and their subcontractors can take many forms. Those with the largest financial impact involves collusion between a company employee, usually in a procurement role and an outside party and or an unreported or undetected conflict of interest.
How can technology improve the efficiency of the fraud detection process?
The availability of data from credit bureaus, social media platforms and publicly available data sources provides companies with the ideal opportunity to build lifestyle assessment solutions that continually detect and prevent employee fraud and collusion. This can be done by identifying lifestyle behaviours that cannot be supported by the individual or household earnings. By implementing this type of technology, you will be able to make risk-based decisions and enjoy the 360-degree view of fraud or corruption risk that threatens your organisation.
Some of the areas where data exists and technology can enable organisations to undertake automated lifestyle assessments to detect and prevent fraud include the following:
Identity documents: to ensure that the identity number which employees present to their employer is valid, has been issued to them and does not belong to someone else or to someone who is deceased.
Financial distress: to determine whether the employee experiences financial distress by considering their financial status, levels of expenditure or whether they have a financial judgment against them.
Financial irregularities: The movement of money due to corrupt activities is one of the key aspects which investigators look for when tracing corrupt practices. People involved in corrupt practices do their level best to shield their transactions from scrutiny.
Income that exceeds earnings: By detecting and identifying income that is materially in excess of salaried earnings that has not been declared to your organisation, could be a source of risk.
Spousal earnings: Financial risk indicators should be considered in conjunction with family earnings to ensure that you are not detecting false positives or genuine sources of additional income, such as from investments or independent directors’ fees. Equally, there are cases that have been discussed where business relationships and income were routed through family members or spouses. This can be an important area of consideration when you are looking for elements of collusion.
Cellphones: Flag employees with unusually large numbers of cell- phones connected to their names. Determine whether any of those cellphones have been blacklisted or blocked by the mobile networks or whether the number given is non-existent.
Politically Exposed Persons: In the wake of the Guptas and State capture, Politically Exposed Persons (PEPs) have greater prominence in the South African business landscape. Although not all persons who are PEPs pose a risk to business, not knowing a person is a PEP could expose your organisation to risk or disqualify you from trading with British or US companies.
Social media activities: These days, social media activity is pervasive. Significant numbers of employees have caused damage to their employers’ reputation, sometimes resulting in crisis and financial loss as a result of social media postings.
Undeclared commercial interests: Undeclared commercial interests that result in a conflict of interest are considered leading indicators of risk. You should ensure that you compare these results with your declarations of interests and remove any results for deregistered companies etc, other than companies that were liquidated through a judicial order etc.
By using technology to monitor the risks you have identified on a continuous basis, you can use the data and the results to make better risk decisions and to protect the reputation of the company and ultimately, generate greater returns for shareholders and the community.
New data techniques make fraud detection more reliable
Your company should not have a one-sized fits all approach to fraud detection and the use of technology. Being overly reliant upon traditional methods or being solely tied to “in-house” systems are some of the factors that we identified as a weakness in companies we interviewed.
Here are five benefits we have found that are realised from using new data techniques to detect and prevent fraud:
- Increasing productivity and profitability
One of the most significant advantages of using decision-making models is to exponentially improve productivity and profitability. Processing hundreds of thousands of applications in an hour is nothing for a decision-making model, whereas only a handful of applications could be properly processed by a human being.
- Decisions are more consistent
By automating the decision-making process, you can be sure that the same methodology is used each time an application is processed. The decision-making model can even consider factors such as your definition of an acceptable criminal past, if any (for example non-violent misdemeanours involving speeding), the position (accounting department, factory floor or elder/childcare), the place you are considering placing a person (locations that have been identified as risk hotspots etc), or even the types of claims that have been filed against an individual (eviction notices, tax liens, bankruptcies).
- Impartial decision-making
Decision-making becomes largely objective and impartial when using a decision-making model. Unintended discrimination based on subjective or unwanted factors such as race, ethnicity, age or gender can be reduced to significantly or even be wholly eliminated. We do not collect or report in our solution upon race, ethnicity, religious or political affiliation for example.
- Ability to customise your decision-making model
While scoring models are certainly not unique (we are all accustomed to looking at our credit score to ascertain whether we can afford a new house), client customisable decision-making models are certainly less discussed and less understood.
- Quality reporting and improved governance reporting
The final valuable differentiator is the way that we report the results to you. You do not simply see one final “numeric score” and recommendation. Rather, you will see the results of each input that goes into the decision so that you can understand what gave rise to the output of the decision-making model – the score and the recommendation. This empowers your organisation to explain the results to the executive team or even to contextualise the results where your business model allows for it.
Harness the power of automated technology
Corporate Insights have developed a one-of-a-kind modular system that combines TransUnion’s big data universe with our own artificial intelligence and smart logic algorithms. It enables you to continually monitor, detect, act on and prevent critical risks, both internally and externally.
The Corporate Insights system will allow you to protect your business from succumbing to the typical pitfalls that lead to corruption. It also comes with a host of additional benefits to ensure your company continues to operate optimally, free of the threat of corruption.
Bowman, R. 2014. “Companies are failing to detect financial fraud in supply chains: Deloitte.” 16 April.