In a world of ever increasing globalisation, markets which were once discrete have become increasingly interlinked and none more so than the wholesale energy markets within Europe. In short, market abuse in one Member State potentially also affects the price of energy in other Member States – yet until recently, there was little definition on what was deemed as market abuse. As a result, the European Union judged it essential to set up a dedicated market integrity and transparency framework for the gas and electricity markets through an EU-wide monitoring scheme. This is how the Regulation on Wholesale Energy Markets Integrity and Transparency (REMIT) was conceived.
REMIT attempts to align the definition of market abuse across Europe and explicitly prohibits it, and has introduced the mandatory disclosure of inside information. The third main obligation is around trade reporting of wholesale energy market contracts – i.e. physical and financial contracts relating to both the supply and transportation of electricity and natural gas within the EU. Unlike EMIR and its similar trade reporting obligations, the Agency for the Cooperation of Energy Regulators (ACER, the REMIT regulator) is obligated to monitor the trades that are reported, and to investigate any cases of potential market abuse.
With the implementation of REMIT reporting obligations fast approaching (Phase 1 of the reporting obligation is currently expected in October 2015), Europe once again turns its attention to the world of market abuse and transaction surveillance – this time, with a focus on the energy markets. The calm and ease with which energy companies have historically operated from a regulatory perspective is officially over. This brings significant implications for compliance functions and aligns energy companies with players operating in the financial markets.
What should be done?
If we follow recent trends, the depth of oversight and inspection that regulators exert over energy firms will only increase. Earlier this year, the US regulators; the Commodities Futures Trading Commission (CFTC) and the Federal Energy Regulatory Commission (FERC) announced a partnership in data sharing under the CFTC-FERC Memorandum of Understanding (MOU), with the joint objective of using analytical tools to protect market integrity. In Europe, ACER have announced that they are implementing a transaction surveillance solution – the ACER REMIT Information System (ARIS), which is currently undergoing the second phase of the prototype development cycle. The main component of ARIS is the NASDAQ ‘SMARTS’ system, which has been adapted for surveillance of the European energy markets. However, REMIT does not explicitly state that firms must have a transaction monitoring capability of their own.
So that leaves energy companies in an awkward predicament – in a world of ever increasing scrutiny, what is an appropriate level of response to the increasing levels of oversight that they will experience? Are preventative measures such as training and embedding the right values and cultures sufficient? Is a reactive approach where investigations are done when requested by the Regulator adequate, or should a proactive monitoring approach be adopted? Are companies comfortable that the regulators know more about their own data and activities than they do?
In the past, fines have settled at hundreds of millions of dollars, so the financial penalties can be very severe. Recent FX market manipulation incidents have settled in the billions. Reputational loss is harder to quantify – though if the banking incidents are anything to go by, then it can only fuel the fire when energy firms are already subject to public backlash after increases in retail electricity and gas prices. With the REMIT mandated reporting of trades, it means that activities today can be viewed forensically by regulators for years to come. By the time a potential market abuse inquiry starts, a ‘rogue trader’ may have already received their bonus and moved on, leaving the company to bear the inconvenience and liability of the investigation. The risk of this scenario in itself may be the trigger point for firms to implement their own internal surveillance capabilities.
Some additional benefits that a transaction surveillance solution could bring are that it demonstrates to regulators, customers and counterparties a commitment to regulatory best practice by the organisation. This would support reputational image if the company was ever pulled into court. It can also act as an internal deterrent to any would-be rogue traders. Armed with the right tools, Compliance Officers can raise their overall credibility within the firm through obtaining new visibilities and insights that were previously absent. Traders will find it harder to disregard any potential regulation breaches if the evidence is presented to them in an effective way. This could even lead to behaviour amendments going forwards.
Finally, transaction surveillance tools hold an abundance of data about patterns and trends in the traded markets, both internally and externally. With this solution in place, this extensive source of market intelligence can be further exploited across the organisation, particularly within the commercial function.
So what to do next?
The first step for every company looking to implement a transaction surveillance capability is to conduct an upfront assessment on the compliance risk on the portfolio. This process categorises trading activity into distinct product classes for review. Products which are perceived as high risk, such as those where there are high volumes of trading activity, elements of speculative trading activity or where the firm holds a dominant market position should be identified and evaluated against the risk of each type of market abuse.
This analysis phase is a critical step to choosing a solution which is fit for purpose, proportionate and appropriate for the business needs. It should also provide a strong steer regarding the priority in which to address the different products, which will in turn shape the surveillance roadmap.
Surveillance capabilities can range from the rudimentary (i.e. basic excel-run macros) to highly sophisticated. The sophisticated solutions typically comprise of algorithms configured with user defined thresholds and boundaries, which automatically run against sets of trade, order and market price data. Any unusual patterns in trading behaviour are identified and raised as an alert in a workflow management system for investigation and escalation. There can be multiple algorithms, or scenarios installed within a solution. Each scenario will detect a specific pattern of trading which indicates potential market abuse. For example, a scenario may look specifically for a high number of orders followed by a high number of cancellations, which is an indicator of spoofing and market manipulation. This type of technology has been common in the financial services sector for many years, with solutions spanning the equities, fixed income, forex and the derivatives space.
More recently, there has been a surge in vendors claiming to provide packaged solutions for surveillance in the energy markets. Most will have well-proven solutions in the financial services sector, but given the complexity of the power and gas contracts, it should be noted that there will be challenges in porting these applications across into the energy space. Generally, the coverage of these vendor solutions will be focused on the liquid markets trading standardised products, such as energy derivatives and futures. The data required here tends to be simpler, readily available and cleaner – the algorithms themselves are also comparable to the ones utilised in the financial services sector, enabling lessons learned from those implementations.
In the energy sector, however, potential abuse can be more complex to spot and may impact across multiple markets due to the way in which the markets are structured and the underlying dependencies on the physical commodities. For example, a party could deliberately exert a ‘squeeze’ through restricting physical supply and then benefit through increased prices on related derivative markets. Generally, existing vendor offerings will rarely be all-encompassing within the energy space, especially in the subject of complex multi-market scenarios. In these cases, the unusual activity can typically be detected in at least one of the markets – then using prior knowledge and experience, the Compliance Officer can look at the connected markets to assess the related activity. This may be the most pragmatic approach in detecting these types of incidents when you weigh up the cost of scenario development against the ever changing trends and market evolution.
In our experience, these vendor solutions can be highly effective for those liquid markets as described above, where there are reliable benchmarks allowing configuration of thresholds with confidence. This allows much of the brute-force search activity to be automated, reducing the manual burden on the Compliance Team. However, for the less liquid Over the Counter (OTC) and physical markets, the thresholds may be hard to define, resulting in either missed incidents or an excessive number of false positive results.
An alternative to a vendor option is to create a custom built solution – this can be achieved by taking a similar algorithmic approach as the common vendor packages with an associated basic workflow capability. This might take more effort to establish initially, but may be more flexible to enhancements and modifications as markets evolve in the future. Alternatively, a less sophisticated solution would be a report or a data-mining based solution built on top of internal data warehouses.
The best solution for each energy firm will depend on a variety of factors, including the compliance risk on the portfolio, size and skillset of the compliance team as well as the firm’s compliance risk appetite. It is likely that a hybrid solution is the preferred outcome – combining the automated tools for the liquid markets with some bespoke data mining capabilities or analytical tools for the analysis of more illiquid scenarios.
Requirements for success
Irrespective of the technical solution selected, there are some basic capabilities that are required for a surveillance capability to be a success.
From a people perspective, highly analytical staff are required for ongoing business as usual monitoring activities, who understand trading patterns and norms. The extent in which they will also have to look at raw data and manually spot patterns is wholly dependent on the level of automation of the selected solution. The number of people required to ensure an effective surveillance function is also dependent on this.
Resources will also be required for the delivery phase – getting people with suitable knowledge may be a challenge due to the availability of specialists who have experience in both compliance and the energy sector. Seconding a trader onto the Compliance team is one approach that could be taken – not only do they bring relevant trading insights during the implementation phase, they can also boost the credibility of the team throughout this transition period from a low capability to a high one.
From a data perspective, the resulting outputs of transaction surveillance will only be as good as what you put in. This means high quality data is required, with unique IDs to allow tracking back to source data if needed. For more complex models, other types of data such as orders, market data, Energy Trading and Risk Management (ETRM) system generated metrics and newsfeeds may also be brought in to establish norms and patterns. With the rigorous reporting requirements established by EMIR and REMIT, there is an opportunity to recapitalise on the investments through reusing the clean and normalised data repositories as a data source for surveillance.
From a governance perspective, the REMIT regulation also introduces strict rules on how to raise a potential market abuse incident, named the ‘Suspicious Transaction Reporting’ process. This suggests that robust processes and/or “case management” tools should be established to ensure that any reportable incidents are locked down, and the right level of access and control is in place.
The magnitude and approach with which these three capabilities should be set up will somewhat depend on the technical solution selected.
Trade surveillance is likely to be a new capability within a compliance function. Furthermore, the complexities within the energy markets mean that it would be beneficial to take a phased approach to implementation.
As with any other system delivery, a build or buy decision should be made against the high level requirements. If it is decided to proceed down the buy route, then a vendor selection phase should be conducted to ensure that the packaged solution is fit for purpose. A Proof of Concept (PoC) phase should then be conducted, demonstrating the effectiveness of a subset of scenarios against sample firm data which will be manually imported.
With expectations in place, the scenarios should be prioritised for delivery against both the compliance portfolio risk (as per the analysis phase described previously), and the ease in which the scenario can be implemented. It is advised to start with the simpler straightforward scenarios, where only a single data source is required (e.g. detection of wash trades on exchanges). The second release could incorporate an additional data source, such as order data, for example, and associating scenarios which require order data. The third data source could be market data, newsfeed data, or ETRM-generated metrics and so on.
This approach will minimise delivery risks, in terms of the number of scenarios and interfaces that need to go live on day one. It will also allow the compliance team to learn how to use the tool effectively and enable the gradual building of their markets and trading knowledge. It also provides an additional element of flexibility should market and regulatory trends evolve, or if there is a change in internal prioritisation.
For each implementation, there will be an ongoing tuning phase until an optimal configuration is achieved. This is often completed by the compliance team, with support from the project team. This ensures that the thresholds and boundaries are set so that the alerts generated are manageable quantity-wise, and that the potential incidents generated are useful and worthwhile examining. This activity will continue for the rest of the scenarios lifecycle – as markets evolve and the definitions of market abuse shift, the thresholds will need to be adapted to ensure relevance to the modern-day markets.
From observing comparable developments in the financial services sector, the general expectation is that energy firms will increase the size and capabilities of their trading compliance functions due to ever increasing regulatory oversight. Surveillance capabilities are one of these areas which is likely to develop extensively in line with technological trends.
The challenge for firms is in understanding what solutions (or hybrid solutions) are the most appropriate and suitable for their specific energy portfolio and associating risks. There is little desire to invest large amounts of time and money in developing a complex monitoring scenario, which may occur sporadically and which the traders can easily counteract. Furthermore, with some markets rapidly evolving and regulatory trends consequently reacting, it is perhaps prudent to consider other forms of surveillance which may be more manually intensive.
Next steps for transaction surveillance itself? With the wealth of data in this digital era, it would be natural to further incorporate additional sources of data into the surveillance capability. Some firms have already started looking at integrating business communication data (voice, email and IM) into their overall compliance frameworks. Tying these discrete pieces of data together with transactional data can be very powerful in understanding the larger picture. But why stop at that? Feeds from assets, scheduling systems, system operators, facility operators etc. can be further incorporated, leading to a holistic view of the entire company, the market and its activities. As mentioned earlier, companies can also ultimately look to use the data gathered for business insight and commercial benefit.
In conclusion, it is likely that transaction surveillance, in one form or another, will soon be a standard function to be found within the compliance operating models of energy firms. With the likes of ACER and the US regulators embarking on their own surveillance ventures, the profile and credibility of these tools have been raised, potentially prompting other regulatory authorities to do the same. ACER have also set expectations that they will make the REMIT reported data available to the market, providing a consolidated source of clean and robust data, perhaps advancing the speed at which these solutions are adopted and embraced. In a world of increasing data volumes and data analysis, along with intense regulatory pressure, serious consideration of transaction surveillance has become compelling.