A report by Thomson-Reuters titled The Rising Costs of Non-Compliance, states that Standard Chartered was fined $340 million for anti-money laundering failings.
The report also notes that a parallel enforcement action against Deloitte (citing the inadequacy of its consulting work on the bank’s AML issues) led to a $10 million fine and a one-year ban on all consulting work at NYDFS-regulated firms.
In Australia, the maximum fine for an individual contravention of the AML Act is $18 million, with government watchdog AUSTRAC set up to regulate AML and Counter-Terrorism Financing (CTF).
Perhaps the most high-profile recent case involves the civil action brought against Tabcorp, with the betting agency noting that civil proceedings initiated by AUSTRAC cost it $13.6 million last year.
As well as the financial burden of AML compliance fines, no institution wants the bad publicity for supporting money laundering or financing to terrorist or other criminal organisations.
Tabcorp’s full-year profits fell by 49 per cent in the last financial period, partly due to ongoing concerns over 236 investigations currently being carried out by the regulatory body.
One answer to the problem of non-compliance lies in big data analytics, with AML solutions powered by an analytical business intelligence program offering banks a clear path to fast, effective and cost-efficient compliance that can scale and adapt as requirements change.
However, the challenge does not stop there.
Streamlining compliance
Anti-money laundering regulations have evolved and become more complex, costly and difficult to comply with. For example, at a bare minimum, a financial institution’s AML processes and systems must support your customer (KYC) activities, including deeper Customer Due Diligence (CDD) and transaction monitoring.
In many cases, the detection of bribery, corruption and tax evasion is also required as part of an AML plan.
Suspicious activity reports (SARs) must now be filed within 60 days rather than 90 days, a big shift for already overtaxed staff.
No longer content with check-the-box compliance, regulators now expect banks to proactively seek out and catch perpetrators, with heavy fines being imposed when banks fail to meet expected outcomes, as mentioned above.
The move to outcomes-based compliance has been driven, in part, by the fact that criminal operatives are avoiding detection by strategically ‘following the rules’.
For example, with currency transaction reporting required for all transactions above $10,000, perpetrators try to stay under the radar by limiting their transactions to just below that threshold. This tactic, known as smurfing, illustrates why banks need to go beyond traditional rules-based detection to proactively identify patterns indicating when customers circumvent the rules.
The tendency within banks is to ‘throw more bodies at the problem’. However, this approach just drives up costs and leaves too much room for error.
Given the ever-growing scope of AML compliance and the massive volumes of data that must be analysed to detect bad actors, it’s simply not feasible to use brute force solutions.
For example, it would take an army of analysts to manually cross-reference large customer lists against sanctioned party lists or trawl through large volumes of data to identify suspicious activity and report on it.
The analytic challenges
To meet modern AML requirements, most banks face a series of analytic challenges. AML teams typically have outdated analytic infrastructure, as many financial institutions have been unwilling to invest heavily in this area until very recently.
With the increasing risk of large fines though, it is incumbent upon financial institutions to protect themselves by properly arming their teams.
According to Ovum’s annual ICT Enterprise Insights survey, 55 per cent of retail banking respondents expected AML-related IT budgets to grow in 2016.
One area that can vastly improve outcomes and cap expenditure is to invest in a big data Business Intelligence (BI) plan. Investment in big data analytic platforms dramatically increases the efficiency and effectiveness of existing AML staff, and as a result, they eliminate the need to throw more people at the problem, even as staff work with larger and more complex data sets.
AML rules demand analysis of a wide variety of sources and types of data that encompass both public and private data sets, in a variety of formats – structured, semi-structured or unstructured.
Examples include the OFAC (Office of Foreign Assets Control) sanctions lists of Specially Designated Nationals (SDNs), Politically Exposed Persons (PEPs), sanctions programs and countries and others.
However, conventional data warehouse and business intelligence tools simply can’t deliver the flexibility, speed and processing of big data needed to prepare and analyse it for regulatory demands.
AML analysts are already spending 80 per cent of their time preparing and analysing data, leaving no time for higher-order investigative work.
For example, in order to track transactions and determine if they were completed by known, high-risk individuals or non-cooperative jurisdictions, banks need to enrich transaction data by joining it with client/legal entity data (including names, addresses and other identifiers) and publicly available OFAC lists in order.
To cast a wider net, they need to enrich this data even further with verbal and written communications information and additional systems information.
Joining these large, complex data sets manually would be both error-prone and extremely laborious.
Business intelligence solutions
The solution lies in harnessing the power of Hadoop (an open-source framework for processing very large data frameworks) and therefore moving from a static pool of data stored in a warehouse to a situation where data is fluid and actionable in real time.
Hadoop technology has been around for more than ten years now, but has recently started expanding very rapidly in Australia as pools of data increase, and businesses realise the need for a faster, more efficient way of actioning it.
Data is stored in ‘data warehouses’, since no organisation really has the hardware to keep it on-premises these days.
Hadoop allows an organisation to store either very large files or very large amounts of data, then access that data much faster than was traditionally possible.
Rather than moving files over the network in order to process them, Hadoop brings software to the files themselves, using an application called MapReduce.
Therefore, the files are already locatable and actionable, and using the data is greatly streamlined.
This enables a much more efficient ingestion, enrichment, analysis, and visualisation of large, diverse and constantly changing data sets so they can be harnessed strategically in the fight against AML.
By investing in the right big data platforms and analytical tools, banks can drastically lower AML compliance costs and satisfy the escalating levels of due diligence required by regulatory agencies.
Modern, big data analytic platforms can manage and analyse extreme data volumes far more effectively and at a fraction of the cost of traditional approaches.
They can easily integrate multiple, diverse data sources and analyse large volumes of data in minutes rather than months, dramatically reducing compliance analytic cycle times.
Equally important, big data analytics can also perform types of analyses that were previously impossible due to the sheer volume and diversity of the data and the complexity of the analysis involved.
Stuart Rees is Datameer’s regional sales director for Australia and New Zealand.
