ABC Analytics: Bridging the Gap in Anti-Fraud and AML Compliance: Anti-Bribery and Corruption Analytics Are Tools That Go beyond Traditional Rules-Based Tests

Article excerpt


In the current regulatory environment, global financial companies face new challenges in their anti-fraud and anti-money-laundering (AML) compliance efforts. New forensic data tools--known as anti-bribery and corruption analytics, or "ABC" analytics--are helping companies cost-effectively and efficiently bridge their existing anti-fraud and AML compliance.

Combining anti-fraud and AML disciplines and technologies, or FRAML, by incorporating ABC analytics is an emerging model that aims to increase the parity and transparency of data sources and information.

The benefits gained from integrating ABC analytics--increased fraud detection and reduced duplication of effort-make a strong case for FRAML. There are many challenges in merging anti-fraud and AML programs, however, as they are often in separate departments within an organization. Regardless, the application of ABC analytics in a FRAML context can drive significant efficiencies when the current anti-fraud and AML programs are separate.

The concept of FRAML allows investigators or analysts to use their experiential-based judgment along with anti-fraud tools to analyze transactions and unusual patterns, in addition to following written procedures. There are instances where judgment calls have to be made and additional forensic analytics can supplement the research required by AML procedures, as well as provide targeted data to support disposition analysis. This is where ABC analytics can play a key role in identifying potentially improper transactions, especially in a financial crimes context.

ABC Analytics: Beyond Traditional Rules-based Tests

When monitoring financial data, institutions often find it difficult to detect corrupt schemes and potentially suspicious transactions using the rules-based data tests usually associated with traditional anti-fraud or AML monitoring. The reason is that funds and fraud techniques often circumvent existing rules-based and red-flag scenarios.

The rules-based test examples used in an AML context may include, for example, the analysis of high-frequency, low-dollar-amount transactions or the search for large round payments.

The traditional accounting and transactional algorithms also frequently yield many "false positives" that slow down the transaction review process and create inefficiencies. As a result, traditional rules-based tests often require significant time, and perhaps luck, to uncover financial fraud or potentially suspicious activity that can make the integration of a FRAML program difficult.

By contrast, ABC analytics specifically monitor potentially suspicious transactions and fraud schemes, taking advantage of advanced data-mining techniques. As a result, ABC analytics can be far more effective. Unlike traditional tests, ABC analytics integrate text mining and statistical and data visualization techniques that allow the data to define or describe itself, without reliance on a predefined query or rule. ABC analytics also help executives search for large and unusual transactions by analyzing several data attributes at the same time, instead of just one variable.

Specifically, ABC analytics integrate both structured (transactional) data and text-based, unstructured data to enable increased transparency in payment and wire-transfer information. This is especially important because free-form text embedded in check memos, wire details, accounting systems (such as Oracle or SAP), and other banking systems may be overlooked using traditional monitoring approaches.

For years, analysts have focused only on the numbers within transactional data, but a wealth of context, insights, and anomalies are embedded in free-text fields that provide transparency into what might be a potentially suspicious or fraudulent transaction. When combined with statistical and data visualization tools, where the analyst can drill down into the data rather than have to look at rows and columns in a spreadsheet, anomalies become much easier to identify. …