Conducting compliance in today’s complex and fast changing energy markets is a daunting task. Relying on your intuition and experience together with random data will not be enough anymore. Your company certainly owns an overwhelming amount of data that you could use in your compliance process. Humans and models are usually not good at handling too much information. Machines that can learn from massive amounts of data can help you make faster, better, and cheaper decisions and detect abnormal and susceptive behavior in an early phase.

Machine learning brings together computer science and statistics to quickly gain insights and making predictions that are based on algorithms that can learn from so-called Big Data without relying on rules-based programming. Big Data itself is not so exciting. However, what you could do with it for instance lower your risk for being fined by the regulator and save costs, is.

Making the shift from traditional analysis to data analytics by machine learning will allow you to get crucial insights from huge datasets such as to detect rogue trading and identify and manage risks.

Below we will describe a practical case we came across during a compliance consulting project for one of our clients, of how predictive analytics and machine learning detected rogue trading practices that were not detected before by the `old` traditional compliance approach the energy company was using. The traditional approach most compliance teams are still using to review the trading activities certainly is having its limitations.

To name the most important ones:

A significant manual effort is required to pre-process, cleanse and analyze data. Given the number of traders and increasing volumes of trades this makes scalability of the process challenging, The old framework relied on selected data sources and provided only a partial view of actual behaviors. Thus, it was not possible to holistically monitor and detect suspicious activities. Due to the sheer size of the data, small parts of the data were randomly selected for analysis, leading to higher risk of missing suspicious activities. Besides, the framework is not adaptive to changing business situations. Finally, we can conclude that aside from these blind spots, the traditional compliance system is inconclusive and often more useful for reconstructing incidents that were already detected.

In contrast, a new system based on machine learning would essentially shift the paradigm away from a risk-auditing methodology based on backward looking sampling to a more comprehensive and continuous monitoring.

By using the machine learning tool your compliance team has the ability to process massive amounts of structured and unstructured data from multiple sources to reveal trends and detect deviations from expected behavior, incorporating data-driven rules that learn and adapt to changes in the environment. This solution includes extensive business logic to review multiple trading activities. It also mines and analyses recorded telephone calls, emails, chat-logs and news.

The machine learning approach would provide several advantages over the traditional compliance approach.

1. The approach is more efficient and allows the energy company to do more with less manpower.
2. It is more effective. The fact that incidents can be detected earlier allows the energy company to prevent them from spiraling out of control. For example, many rogue traders follow a “doubling up” strategy of risking an increasing amount of capital. Stopping the spiral early enough can prevent cases such as the collapse of Barings Bank from happening again.
3. The system is adaptive. Humans have a great capacity to adapt to controls imposed on them. In contrast, policies adapt at a much slower rate to changes in practices and business conditions. The new system’s learning capability helps address this problem. This creates a positive impact on organizational culture by reducing the bureaucratic burden created by meaningless controls and by protecting social norms through the detection of early deviations.

It would be our pleasure to discuss a workshop for you and your people that will help you identify what machine learning could do for your business.

Just call or send me an email, and I will arrange a no obligation discussion.

Kasper Walet

Last changed: 02/07/2018 at 12:06 am