The data blows the whistle: Using analytics to reduce fraud in China’s payments ecosystem 

Stepping up the scope
Ask most banking executives and they will likely tell you that they already use analytics to spot fraud and other illicit activities. While this may be true, most only do so in a reactive way by reviewing suspicious activity or identifying anomalies in patterns – arguably once the fraudulent activity has already occurred.

Moreover, the use of analytics to detect fraud has, to date, largely been focused on the retail market where the risks are relatively well-known. Today, however, China’s banks are starting to recognise the real and material risk of fraud or rogue activity in their trading and investment management spaces as well. This is much more difficult to identify: insiders generally have a clear understanding of the systems and controls in place to identify fraud and are therefore uniquely positioned to circumvent them. As a result, anti-fraud measures in this space must be flexible, adaptable and conducted in real-time in order to identify and interrupt fraudsters’ new permutations and tricks.

This is where data analytics really proves its worth in eliminating fraud. Modern data analytics is not only capable of simultaneously conducting monitoring and alert processes in real-time across a massive volume of transactions. It can also adapt to changes in patterns and incorporate important external data. For example, today’s systems are able to monitor a trader’s communication across email, telephone, text, social media and messenger services to spot suspicious activity, and then correlate that against the trader’s transactional activity. It can also reach across publicly-available data from other payments parties to flag potential occurrences of inter-institutional or international fraud.

Turning data into action
It will, however, take much more than simply implementing the latest data analytics solutions to reduce or eliminate fraud; it will also require banks to develop appropriate processes and responses to allow them to proactively manage the resulting prompts. The activities of Nick Leeson in the 1990s and Jerome Kerviel in the early 2000s all went unnoticed causing significant exposure. More recent incidents have been identified earlier, but the reaction time taken to prevent rogue trading is still poor and timely responses do not occur. It is one thing to identify fraud; it is altogether another to be able to turn those prompts into immediate and effective action.

Moreover, these controls must be efficient enough to operate in real-time, particularly given the rise of SEPA in Europe and Singapore’s G3 payments mechanism, both of which have effectively eliminated the 24-hour turn-around time on transactions that had traditionally acted as a buffer for financial crime analysts and investigators.

The point here is that financial crime analytics is not just about data and information. It is also about speed, efficiency, effectiveness, and the ability to embed the system into operational risk and management processes. Consequently, given all the changes that are now underway in China’s payments ecosystem, this seems like the perfect opportunity for the country’s banks to build these robust, real-time financial crime responses into their new operating models.