Managing Transaction Data for Sarbanes-Oxley Compliance

6th Annual Energy Transaction Software Guide 2004
A Supplement to Hart Energy Markets, April 2004
By Peter Weigand, Chairman & CEO, Skipping Stone
and Jeff Felton, Principal, Skipping Stone

pweigand@skippingstone.com    jfelton@skippingstone.com

For energy merchant companies, the scramble to achieve Sarbanes-Oxley compliance has been in full swing for many months. Typically, a top-down approach drives the initial compliance planning, given that CEO’s and CFO’s are the people personally responsible, and literally liable for the outcome. Between the external auditors, CFO and CRO, high-level objectives are set, reporting requirements designed, and projects initiated to meet deadlines. While Sarbanes-Oxley covers a wide swath in reporting requirements, for energy companies one of toughest areas to reach compliance is energy transaction and risk management. 

To effectively reach appropriate confidence levels, the top-down mandate and report design approach must be augmented with a “foundation layer” improvement project. Data integrity must be addressed at the foundation layer to ensure that all report designs achieve the confidence level required for corporate officer sign-off. 

The Transaction Foundation Layer

The transaction foundation layer can best be defined as the creation point of the data that eventually rolls up into the reports that have been designed for compliance purposes. Each commodity transaction contains multiple data elements that make up that unique transaction. For example, a purchase of natural gas contains many data elements such as counterparty name, price, location, etc. 

To raise the level of confidence needed for Sarbanes-Oxley compliance, a thorough understanding and control of the underlying data must be achieved at detailed levels, which means at the point of creation of each data element. To effectively achieve signature level confidence, a transaction system and internal business processes must work together to achieve the results as defined in the top down mandate. Figure 1 provides an illustration of the layers that are needed for organizing the data.

Figure 1.  Transaction process layering.

The three primary elements of a foundation layer improvement project are business process mapping, data modeling, and a data dictionary. No single one of these elements can be applied in a vacuum, as they are dependent on each other.  When combined, the individual elements act as a check and balance for each other. Having observed companies that pursue only one element (most often just the business process mapping), many find themselves in an endless business process edit and technology implementation loop, which leads to higher and higher costs, with compliance deadlines missed. 

Most people are familiar with business process mapping entails and have seen the flow charts that result. The data model defines all of the data elements associated with each particular business process, and then models where the individual data elements come from and go to across the business process map. When overlaying the business process map and data model, data dependencies take shape that determine where data is needed along the chain, what the data looks like, and at what stage in the process is it required. Understanding data flow at the detail level serves to better define software requirements and supports reporting accuracy.

The data dictionary is critical because it provides a common definition and exact description of each data element. Without using a data dictionary similar looking information will be assumed to be the same or called the same, when in fact they are different. When this occurs, base line data assumptions across the map are not consistent, which leads to future errors and significant expense in reworking to find the underlying problem. This is further exacerbated when applying the maps and models to software without a data dictionary.

As the transaction foundation layer improvement project takes shape, one of the early findings will be a significant level of redundancy in data collection and processing. By discovering this early, technology tools and data processing can be reviewed along the way to reduce inefficiencies and improve accuracy. Both outcomes serve to lower future labor and software costs, and produce more reliable reports for compliance and other management purposes. An early benefit is the identification and reduction of the number of errors that result when technology and business processes are not in sync. 

A second discovery during the transaction foundation improvement process is the identification of gaps in data that are critical to reporting requirements. Most companies, at the operational and detail level, have confidence in their reporting if only because they are too close and too familiar with it to see potential shortcomings. Companies who exhibit these traits are easy to spot because there always seems to be some mysterious reason the forecast or financial numbers need to be modified after they have been published. Another foundation layer issue can be seen with internal disputes arise as departments compare reports on the same topic and yet have totally different results.

A transaction foundation layer improvement project has a number of benefits besides Sarbanes-Oxley compliance and has demonstrated a significant return on the investment. Among these other benefits is the ability to streamline and gain efficiencies operationally and better define software and technology requirements, thus eliminating significant process and technology overlaps.  

The following are examples of common data gaps or reasons for reporting problems that have been found during a number of Sarbanes-Oxley compliance engagements for energy commodity trading companies: 

1. Comprehensive Position Reporting. All energy-trading companies produce position reports. To produce the position report, the transaction processing system accounts for most deals, but often complex structured deals are processed outside the system and their results added to the position report later in the process or not at all. Without these transactions, it is not a comprehensive position report and thereby not Sarbanes-Oxley compliant.
   
   The additional fallout of an incomplete position report is the risk measurements and metrics that are derived from the position report. If the base information is not complete, then none of the other reports will be accurate. 
2. Counterparty Credit Worthiness. All energy companies assess the credit worthiness of their counterparties and place limits on the size of transactions or outstanding accounts-receivable (A/R) balances that can be carried for a given counterparty. However, many companies have different business units that may establish transactions with the same counterparty. While no single transaction or individual business unit’s activity may exceed the credit limit, the net credit position across the enterprise may exceed the credit limit. Failure to recognize this effectively circumvents the credit exposure management process and impacts the organization’s ability to achieve compliance.
   
   A further consideration is the possibility that one business unit may be having difficulty in collecting funds from the same counterparty to which another unit is authorizing payments. 
3. Optionality Accounting. There is a trend regarding entry into more complex deals for both customers and counterparties. Many of these deals involve “optionality.” For example, customers have the option of switching between fuels, distribution companies have swing options on the volumes they purchase, and ever more complex structured deals can be created that imply optionality at one or more levels of the deal. We often find that the optionality of these deals is not captured anywhere in the transaction system nor included in Position and Mark-to-Market reports. A set of assumptions is often embedded that serve to define the deal and this information becomes data that must be captured.
   
   If these deals are not entered into the system, they also bypass the middle office or independent risk monitoring operation, which eliminates the checks and balances as they were intended during the top down, Sarbanes-Oxley compliance process.  
4. Pipeline Capacity Allocation.  In US markets, pipeline capacity contracts allow the holder to dispatch volumes to a variety of secondary points within the footprint of the pipeline and within the contract terms associated with the capacity rights.  However, some companies do not take their forward month capacity planning to the detailed secondary point level.  This can lead to several problems on position and Mark to Market (MTM) reports, including:
   • Max rates, variable costs and discounts – These are typically applied at the secondary point level of detail.  Inability to plan forward months at this level leads to inaccurate basis cost calculations;
   • “Open” or unallocated capacity - This open capacity must be used to calculate cost deferrals to accurately accrue costs in proportion to volumes shipped, or to allocate costs for open capacity back to the operational regions in proportion to volumes shipped.  This is essential to support more accurate Position and MTM reporting by point or region.
   • Inefficient use of available capacity - Inefficient use of available capacity along segments of the contract, or worse yet, inability to prevent “double usage” of the space along one segment of the pipe by different schedulers moving gas to different secondary points.
5. Volume or Settlement Risk. At the end of each month, financial statements must be prepared and the books closed for that accounting period. While the books should be completely accurate for reporting purposes, the appropriate accrual and measurement of impacts from physical volume value at risk (Vvar), or settlement risk is at issue. In many physical commodity markets there is a perpetual process often known as “prior period adjustments,” which are corrections from estimates to actual for some period of time in the past. This is often an issue that cannot be resolved because it is dependent on some other company’s information system. While this is a convenient excuse, it does not make compliance possible.

These are just a few examples of the reporting and compliance issues that keep the Sarbanes-Oxley team up at night. When viewed in total, these and other problems appear daunting, which is where the foundation improvement project bears fruit. The issues discussed here have been experienced through the application of the three-part methodology of business process mapping, data modeling and data dictionary creation in our transaction foundation improvement project engagements.  While undertaking such a project appears challenging, there are some easy-to-use rules of thumb that make the process more manageable:

• Work backwards. Start with the Position and MTM reports and determine how each type of deal should appear to be most meaningful and accurate. For cases where there is no right or wrong answer (e.g., do I represent basis position at the receipt or delivery point) adopt a convention and stick with the same convention throughout the modeling of similar types of deals. This will support your data dictionary, which should be referred to often and modified only when needed. 

• Test with real data. After developing reporting rules above, select a handful of deals of each type (some simple, some complex) and manually work through the individual data elements, calculations and processes by which those would be reported. Adjust the reporting rules and conventions as necessary and be certain to list each data element to see where else it appears along the process chain.

• Map back from the report to the data source. Once the report rules are documented and tested, map the data back to the source system. It may be the one transaction in the source system that maps to multiple line items on the report, or vice versa. 

• Develop a standard set of roll-up and drill-down rules. To be most effective, all reports must be able to be rolled up for management presentation and then drilled down to get to the operational level of detail to manage the business on a day-to-day basis. Develop a standard roll-up/drill-down structure so that all levels of the report may be viewed as different views of the same report, as opposed to different reports. People viewing different reports on the same topic are the leading cause of reporting and compliance errors today.

• “Conservation of mass.” Inevitably, some transactions will not fit into some categories on some reports. Do not include transactions in one section of a report and exclude them in another. Rather, put transactions that do not apply (e.g., Henry Hub transactions on a basis report) into another reporting bucket called “other” or “NA” (not applicable).  Doing so will greatly improve the ease of investigating program and data aberrations. Once you have compiled an “NA” bucket, it will become apparent that many of the transactions in the bucket share data attributes.

• Align business processes along the way. Don’t wait until the end for the big bang implementation of information systems to set it straight. While this may appear to cause additional organizational upheaval, this is a reengineering of the workflow and significant change management exercise, and people respond better to gradual change than sudden change. 

In addition to the Sarbanes-Oxley benefit of having your transaction system be in sync with your business processes at the foundation level, management will truly understand the business at a level that improves strategic and tactical planning capabilities. For those commodity merchants that undertake a transaction foundation improvement project the results have proven to create a true competitive advantage. 

pweigand@skippingstone.com    jfelton@skippingstone.com