Implementation of a Value-at-Risk management system

at TransCanada Energy Management

The need for risk management systems within the energy commodity trading business has never been greater than in today’s environment.

Stew Fairbanks at TransCanada Energy discusses.

Financial instruments are becoming increasingly complex; globalization of the marketplace continues, the volume of transactions being performed has multiplied, and the magnitude of capital within these markets is increasing. Companies are focusing on protecting shareholder value which has created a requirement for the development, maintenance and enhancement of a solid risk management system.

A good risk management system is dynamic and consists of three key parts:
1. Identification of the risk
2. Quantification of the risk
3. Risk modification through the appropriate hedging strategy

TransCanada Energy Management (TCEM) has developed a very effective system which covers a wide variety of trading operations, This includes all energy commodities, from natural gas, to crude oil and NGL products. While still a relatively young player in the risk management game, compared to financial institutions, TCEM is a leader within the energy industry. As a firm, we are continually modifying and adjusting our risk management systems to better meet our stakeholder needs and to improve the efficiency in our operations.

The main focus of this paper is the quantification of risk and in particular the systems used to measure and report this risk- A number of drivers impact our current systems, including,

• Seasonal, often volatile, energy markets. Futures contracts that expire during the measurement period.
• Outright and basis (location grade, time, etc.) exposures for all l0 energy products.
• Positive and negative price co-movements (correlation coefficients)
• Long and short exposures
• Time, locations and product diversification.
• Aggregating > 1,200 Separate VAR calculations.

To quantify and report market risk across our trading books, we use a Value-At-Risk (VAR) calculation. VAR measures the worst expected loss over a one-day period under normal market conditions at a 95 percent level of confidence. This measure can be applied equally to all commodities. VAR is expressed as a $ amount that provides an aggregate measure of the market price risk that TCEM is exposed to in the conduct of its energy marketing activities. By quantifying our risk in this way, our senior management can compare and manage the market risks throughout TCEM's business.

The initial TCE system involved a manual process which was very time consuming. It consisted of a variety of books, loaded into an Excel spreadsheet, in which the calculations and reporting functions were performed. It took a great deal of time to consolidate and report this information. While the system was effective, it lacked efficiency. With the desire to view our exposures and risks from a number of levels, book, region, and portfolio, as well as senior management requesting the information on a more timely basis than this manual system, it was apparent that changes in efficiency were required. Enterprise-wide risk analysis is so complex, that it requires a flexible and simple-to-use risk management tool.

TCEM undertook a wide-range of research into various software providers. It was imperative that any new system had the capability to fit into their overall I/S strategy. TCEM had already made the decision to use a Sybase database to consolidate their risk information, and any provider needed the ability to effectively integrate with Sybase, one of the key drivers was a system which would reduce the manual process and increase accuracy. Two other critical factors in the decision were flexibility and speed. Flexibility in terms of the ability to customize and adapt to the changing environment.

TCEM's business changes almost daily and they really didn't want to get stuck with an inflexible system which would limit their risk management requirements.
Speed was also a major concern TCEM preferred to purchase a system that would take them 80 or 90 % of the way to a solution quickly, rather than design a huge system up front without knowing the time frame.
Based on meetings, discussions and systems validation procedures, including a week long series of live testing in Calgary with data collected from two of TCEM's Forward Books, Axiom Software Laboratories, it was determined could meet TCEM's needs.

Axiom Software Laboratories, Inc. (Axiom), a software services company based out of New York, was enlisted to help TCEM with their requirements to quantify market risk. Axiom had previously developed a number of key elements into an overall risk management system, primarily focused on financial institutions but which could be readily adapted to meet TCEM's unique requirements. Axiom had optimized the complex VAR calculations using sophisticated programming applications which resulted in significant reductions in processing time.

Risk Monitor® supported multiple risk methodologies, and was readily adaptable to TCEM's specific requirements. It integrated several approaches to VAR calculation, including Variance/Covariance, Monte Carlo, and Multi-factor Analysis, and provided the foundation to consistently measure risk across all product lines, FX, and Structured Products. A number of features of AXIOM-Risk Monitor were particularly useful:

• Mark-to-market VAR portfolio valuation
• Multi-dimensional VAR application scenarios
• Disaggregation capability of VaR portfolios to the transactional level

Risk Monitor® allowed TCEM to define portfolios in a variety of ways, thereby enabling market risk to be examined in aggregate form.

The AXIOM-Integration Server m provided TCEM with the flexibility to reassemble exposures, mark-to-market profits and VAR into format that are meaningful to each individual business unit's perspective, This was very important to meet the broad range of stakeholders who require the VAR information among the TCEM portfolio. Traders required the information on a transaction or book level, whereas senior management required this information to be aggregated into a region or across the total portfolio. This in turn has resulted in more cost-effective decision-making, more accurate reporting, and the ability to quickly support business process reengineering. From an analytical standpoint, TCEM used proprietary algorithms developed using Axiom's quantitative toolkit. The system's architecture allowed TCEM to plug in other analytic modules if so desired.

TCEM ran Axiom's financial data warehouse on a relational database populated with low-level transaction data such as deals, counterparties and positions. Market prices and factors were also fed daily into the warehouse. Axiom's program automatically pulled this information from a consolidated database and performed the required risk calculations. This was not a simple task as there were a large number of market factors (super large covariance matrix) to quantify. TCEEM also calculated VAR in multiple currencies. Axiom was able to make specific software modifications to meet TCEM', needs, primarily in terms of the unique characteristics of the energy commodity trading business. Axiom's software was installed and implemented in only two months. Speed was a critical clement in terms of the selection of a software provider. Axiom's software services contained TCEM's three critical criteria of systems development, increased accuracy in terms of quantifying market risk, flexibility in terms of segmenting and reporting the information to a broad range of stakeholders, and speed in terms of aggressive implementation deadlines being met.

Benefits to TCEM:
Ability to generate better information in terms of sensitivity analysis, stress testing, l-off projects, CVAR, Monte Carlo arid other sophisticated VAR methodologies with improved accuracy. Generate and report the information in a more timely manner (daily as opposed to weekly). Free-up senior staff to address the broader corporate-wide risk management issues.

Future:
The implementation of Axiom has significantly improved TCEM's overall risk management system, and they are currently working on a number of enhancements to gain further benefits. This includes a method to automatically override new exposures to ensure that statistical parameters (30 day price history) are not violated when a daily VAR is calculated. An enhancement to directly link exposures with deals will allow specific reporting requirements to be made directly frown the Axiom database and save a significant amount of overhead. The commodity trading business is changing very quickly, and it is flexible tools like Axiom which have helped TCEM to maintain their leading position within the energy industry in terms trading and risk management.

Stew Fairbanks is a financial analyst in the market risk management group at TransCanada Energy Limited. /p>

Reprinted with the permission from: Commodities Now, November 1997