Edward Cuoco
Director of Utilities & Energy Markets
Martin Dawes Analytics (Boston, Mass. USA)

Victor Milligan
Chief Strategy & Marketing Officer
Martin Dawes Analytics (Boston, Mass. USA)

As the illustration above shows, positions are generated daily and intra-day, and these reports leverage estimates of generation and off-take data. While the actual data (particularly generation data) is often available, neither the existing tools nor processes used by energy trading companies allow them to be leveraged in this reporting cycle. Consequently, these data elements end up as part of manual portfolio optimization, occurring days or even weeks after the initial position was reported. It is possible, however, to enhance risk reporting with tools found in other industries, allowing for faster integration of detailed supply and demand data.

Companies that look outside of traditional trading tools will be able to rapidly assess, evaluate and react to this critical data and put themselves in a stronger position to make more timely and informed decisions about future trades and contracts.

The value of integrating supply and demand data at intraday increments to intra-day risk reporting lies in reduced risk in the portfolio, lower fines and/or penalties for being out of balance and ultimately, in the improved optimization of trading strategy (and by extension, a more competitive advantage in the market). Operationally, integrated data improves the firm’s ability to adapt trading strategies to account for unexpected supply changes more quickly. Furthermore, timelier accounting for supply and demand data will reduce the value at risk (VAR) of short- and longterm contracts and provide improved insight into counterparty and contract profitability.

Yet few organizations are actually using either supply or demand data on an intra-day basis despite the obvious benefits. Processes for position reporting are organized around intra-day schedules, but supply and demand data are often unaccounted for in these processes.

In each case, trading desks are forced to perform a “true up” by means of manual reconciliation of the data to position reporting days or weeks after the fact, sacrificing margin or even moving from profit to loss due to “unexpected” changes. Therefore, the choice facing trading organizations today is between an analytic capability that allows one to look backward and report what happened, and a capability that allows an organization to understand what is happening now and adapt and react to it in real time.

Current Systems Are Not Sufficient
The difficulty trading organizations face when assessing the data in a more timely fashion stems from a perceived inability to apply the data to risk reporting more frequently, while ensuring that the data is granular enough to revise the position in quarter hour increments. Attempts at systemic solutions to this problem end up in one of two untenable outcomes; that is, either the information can be processed intra-day, but at insufficient granularity which limits the solution’s ability to remove the need for postday adjustments or, the data is granular but the time to acquire, process and analyze the data is too long to allow it to be used same-day.

For most trading organizations, it has not been possible to create a systemic solution for these problems within a time and price range that reflects value. Further, attempts at integrating this data within the risk reporting process have taken more time and cost more money than has been gained in profitability. For those that have been completed, many other attempts have failed or simply been prematurely abandoned due to either the exorbitant expense or lack of demonstrable success.

The myth prevalent within many trading organizations is that these issues cannot be resolved without a massive IT investment. This myth leads trading organizations to believe there are only two possibilities; either repeated and increasingly expensive attempts to create tools to integrate this data or, alternately, to give up altogether and implement work-arounds or reduce risk limits to allow the enterprise to simply live with the existing problem.

In reality, companies are simply trying to solve the problem with the wrong tools; focusing on using the typical tools of the trading world which inevitably leads to a dead end and include implementing enterprise systems such as Commodity Trading & Risk Management (CTRM) tools or back-office financial systems. These options, besides being large and complex, also require a significant investment of time and money and cannot be quickly adapted. While Excel-based tools are flexible, they lack both the ability to handle larger volumes of daily incoming data and the transparency and auditing ability needed to provide confidence at scale.

Casting Light on Suitable Solutions
In order to leverage the full range of available supply and demand data on an intra-day basis, trading organizations must look outside of typical CTRM or BI (Business Intelligence) tools. As an example, industries such as telecommunications and manufacturing make good use of process-driven analytic tools in order to address analogous issues in their spaces. These process-driven tools have many of the key elements needed to ensure that detailed data is quickly and granularly analyzed. Specifically, they are data-architecture independent, they allow business users to create and change analytics quickly and without a major IT engagement, and they allow both logic and data to be modeled in the same tool.

Equally important is that process-driven solutions support an analytic methodology where discovery and analysis happen simultaneously – occurring when business users create analytic tools collaboratively even as they investigate the data itself. This combination of technology and methodology enables energy companies to enhance their risk reporting and fulfill their core need for integrating supply and demand data intra-day.

When looking to acquire this type of technology, organizations should consider the following critical criteria and select a solution that:

1. Enables the organization to access and apply data quickly, including acquiring and analyzing data in near real-time
2. Allows the organization to maintain sufficient data granularity in order to improve position reporting at quarter-hour increments
3. Is able to analyze supply/demand data in the context of trading and contract logic
4. Supports an agile analytic methodology, allowing business teams to adapt and tweak analytics and explore new data sources quickly and easily
5. Creates output that can be audited and tracked – providing confidence in output and reducing the likelihood of needing revisions
6. Delivers value within 3 to 6 months and be able to consistently adapt itself to new data inputs and analytics within days or weeks.

By leveraging this class of tool, it is possible to create a solution that is managed and maintained within the process analytics team or perhaps in the mid-office. This eliminates the risk and expense of a large-scale IT implementation but does so without sacrificing the ability to convert these analytics into standing operational controls and maintains transparency and audit capabilities.

Illuminating Profits and Reducing Risks
It is easy to ignore solutions from other industries when attempting to expand the analytic capability for energy trading. Indeed, both heads of trading and IT management often raise objections to such an approach, believing that trading is unique or too mathematically complex to find guidance in other industries, or that only companies steeped in trading and risk management expertise can possibly provide solutions in a timely, cost-effective manner. However, these objections do not stand up to investigation. In telecommunications, these solutions are already in use, having been proven in maximizing revenue for billion dollar industries and working with millions of records across multiple systems.

Simply put, experience in other verticals has shown that complex analytics can be implemented quickly and efficiently by leveraging solid technology, and proven best practices and expertise in process, logic and data.

In order to successfully adapt solutions from other vertical industries into trading, organizations should take an approach based on quick timelines and minimal risk. Big-bang solutions should be avoided and more attention should be focused on small systems that address core parts of the risk reporting process such as the integration of generation data, analytic, etc. Companies should begin by working with proofs of concept as a way to confirm a technology’s ability to perform a function and through proofs of value to test logic, data and analytics.

By keeping initial timelines and investments short, organizations will maintain more flexibility to mix and match technologies and avoid becoming trapped in a substandard solution due to large investments of time and money. Finally, along with these new tools, existing trading and risk management technologies also have a role to play. When investigating technologies, organizations should pay attention to the ease in which they can be integrated into existing architectures and data models. After all, even the most simple, elegant tool can result in a blown budget caused by complex and expensive integration tasks.

Market, industry and organizational pressure all suggest that organizations able to successfully improve the accuracy and timeliness of their risk reporting stand to benefit from improved margin, reduced controls and the ability to more quickly move in the market. New tools from outside the traditional trading or utilities space can provide the needed functionality to quickly and cost-effectively create this expanded analytic capability for any trading desk. With a little technological planning, the future can be very bright, not only in terms of profits, but also in the form of reduced risk.

Conclusion
An analytic approach to issues of risk management is inherently familiar to energy trading primarily because risk management, pricing and supply teams perform complex analytics on business critical data every day. Improved analytic tools from outside the typical set used by merchant energy and utilities allow this approach to be extended to the integration of supply and demand data to intra-day position reporting.

These tools make this extension possible by combining an enhanced set of analytic functionality with a design and architecture that allows organizations to “de-risk” upfront implementations. Further, this design allows Utility companies to integrate these technologies into the existing suite of trading and operational tools in short, inexpensive phases that reduce the impact on IT budgets and reduce both the likelihood and costs of potential over-runs. It is this combination of functional robustness and operational flexibility which provide real value to utilities without the attendant risks or large-scale system changes.

About the Authors
Edward Cuoco is the Director of Utilities and Energy Markets and Victor Milligan is Chief Strategy and Marketing Officer at Martin Dawes Analytics, a leading global, data-driven process analytics software provider. For more information, contact Ed at ecuoco@mda-data.com; Victor at vmilligan@mda-data.com.