Fraud Dynamics: The smart approach to fighting financial crime – Sjoerd Slot
Criminal activity such as fraud, money laundering and terrorist finance poses a threat to the integrity of the international financial system, and customers and regulators want to see the industry do all it can to keep that system clean. Future Banking asked the founder of financial analytics company Fraud Dynamics, Sjoerd Slot, how banks can improve performance without spending millions on a complete technology overhaul.
When it comes to combatting fraud, money laundering and other abuses of the financial services infrastructure, the banking industry can claim a good track record, although major incidents can and do occur. The problem today is that the challenge of maintaining this level of success is becoming much harder.
The nature of financial crime is constantly changing and is growing increasingly complex. This difficulty the industry faces in keeping track of this moving target is compounded by the ongoing changes in the banking business, where financial institutions are constantly altering their structure through mergers and divestments. At the same time, the demands of regulators and customers to stay in control of the financial services infrastructure are increasing and banks are expected to be more proactive than ever in combatting crime. Regulators in the US and Europe want more guarantees of integrity and are increasingly prepared to hand out fines and jail time to those responsible for failures in the system.
"When there is money around people will want to get it. The world is getting smaller and more responsive, so the threats are truly global and dynamic," says Sjoerd Slot, the founder of Fraud Dynamics. "There are hundreds of thousands of attempts every year to commit fraud and launder money, and the industry does not know who the criminals are, when they will strike or what their strategy will be. There is a large variety of risks, so it is very complicated. Viewed over a long time frame, the industry's track record is decent, but it is getting tougher and more expensive to fight the criminals.
"In the online world, the cost of doing business for criminals is falling. They can easily resource people to write software or harvest money, so the barriers to committing attacks are getting lower. For banks, the problem is that they need more resources to combat these attacks, so their teams are growing massively. At the same time, many large banks are refocusing on Europe or the US and moving away from exotic jurisdictions by selling their operations so they don't have hands on the risk there."
Build strong foundations
Slot started Fraud Dynamics in 2013 as a result of a collaborative effort between international experts focused on the future of fraud management. Its sole aim is to address the growing threat and changing nature of financial crime with results-driven solutions based on top-class analytics, agile platforms and the first-hand knowledge of experienced people in the area of fraud management.
"With money laundering, for instance, there is a big societal impact and public perception can be damaged. It is not an attack on the bank itself, but banks are part of the infrastructure for money flows, so they bear a regulatory burden. When it comes to vulnerabilities, changes in the composition of the banking industry through mergers and acquisitions - and through the development of fintech that makes certain processes easier and more customer-oriented - make it easier for fraudsters to exploit weaknesses in the system. The challenge with M&A is that IT schedules are packed and complexity is doubled, hence making it more difficult to respond quickly to new modus operandi," Slot observes.
As a specialist in financial crime control - encompassing fraud, anti-money laundering, KYC and counterterrorist finance processes - Fraud Dynamics has already seen keen interest in its solutions from banks.
"Banks have welcomed our models because they perform very well and they work with the systems that are already in place. Larger banks are jumping in as they now see the performance we've delivered in different situations," Slot remarks.
The future Slot envisages is not one where the problem is tackled with more and larger systems. Instead he proposes a more cost-effective way to improve performance. Banks already collect vast amounts of data and the volume will only keep growing, but Fraud Dynamics gives banks the ability to find what matters within that data and focus more accurately on the anomalies that are significant.
"Fishermen with large nets catch everything but it is costly. For banks, there is the huge cost of the team to process more suspicious activity. For instance, every transaction over €2,000 to a specific suspicious country could be flagged up. So we need an intelligent way to massively decrease the amount of alerts you need to process. When there are fewer false positives, the process becomes much more efficient and the cost component is hugely important for banks," says Slot.
"To get ahead of criminals, you have to look at what is a stable factor. Rather than trying to predict fraud, you try to establish what is legitimate. Most customer behaviour is stable, though there will be the usual changes and peaks, so our security measures mainly work on predicting legitimate activity. If you force criminals to get closer to what appears to be normal customer behaviour then their costs rise because they can't do everything in just a few big attempts. For example, the coordination required to take small amounts from many accounts over a long period costs more, but even that becomes structural and structural anomalies again make detection easier," he adds.
Make the call
So, how does Fraud Dynamics work to differentiate between legitimate and suspicious behaviour? The answer lies in its analytics approach, which focuses on two key areas. The first is anomaly detection and, as Slot explains, this predicts normal behaviour rather than specific patterns of fraudulent activity. Previously, approaches have focused on identifying a new type of fraud and creating a specific detection rule set, but by the time the model is in place often the fraudsters had already adapted to it. Furthermore, these models can be typically inefficient by generating a lot of false positives.
"We use the triangle of a deep understanding of the analytics techniques - we work with econ PhDs from the best universities - understanding the battleground through input from top fraud and AML experts, and adapting the model to the data and systems at hand," Slot remarks.
The second key aspect of the solution is flexible and integrated case management, which enables banks to change their processes in seconds rather than months and to automate as much as possible. Human input is reserved to intuitive and contextual judgment, as well as client interaction. A simple and intuitive workflow allows most typical fraud processes to be addressed in just a couple of clicks, and ensures that all processes are auditable, which is a must from a regulatory perspective.
Not bigger, just smarter
The combination of these factors not only reduces costs, but also enables banks to maintain a client-oriented business model while staying firmly in control. Furthermore, there is minimal change to the existing IT infrastructure.
"Our solution tackles serious problems without huge cost. It requires no large IT investment and is quick to implement. It is integrated with and operates within core systems that would be very costly to replace. It can also adapt to systems as IT infrastructure changes over time. Smart analytics works even in older systems as the rule set can be very simple. We are replacing the driver, not the car," Slot explains.
In practical demonstrations, the advantages Fraud Dynamics offers soon become clear and so far Slot's message about working smarter rather than harder has been well received across the globe.
"To demonstrate what we can do, we apply our model to existing client data at the client's site, so they don't have to hand over any data to us. We show the performance they could have had over a historical period. We show scientifically the figures that would have come out of the system using real data, not hypothetical data, and our results have generated global interest. Our clients include banks and payments processors in the Netherlands, Germany and Scandinavia, and there are discussions happening now in the UK, US and Latin America," Slot explains.
"Make no mistake, this is the future," he stresses. "A good system is invisible and banks may not be able to talk about it too much but they will stay out of the headlines because they will have no money-laundering scandals or major incidents of fraud. Damage to reputation lasts a long time in this industry."