Resource Planning After the Crash

How to update yesterday's IRP model to account for tomorrow's risk profile.

The process we know today as integrated resource planning (IRP) got its start back in the 1980s, when regulators first came to grips with nuclear plant cost overruns and urged utilities in effect to hedge that risk-to give equal weight to conservation, "negawatts," and demand-side management (DSM) as sources of new electric capacity.

Times Have Changed, Even If IRP Hasn't

Today's utilities face a bevy of new risks in this post-Enron, post-crash, post-blackout world. This litany of new dangers includes such risks as: (1) gas price spikes; (2) thin and illiquid trading; (3) suppliers with poor credit; (4) boom-and-bust cycles; and (5) merchant power bankruptcies. And this new set of risks demands a fresh look at traditional IRP.

In the standard IRP approach, utilities first develop a long-range load forecast. Then they evaluate their existing supply-side resources and prospects for DSM to determine uncommitted needs. Lastly, they select among competing supply- and demand-side programs that appear most viable and economic, develop alternative portfolios, and test those portfolios against selected scenarios to see which portfolio will best satisfy a set of economic, regulatory, and environmental mandates.

The tools most typically used in this traditional IRP involve forecasting, shaping, and characterization of load, plus price forecasting for power and fuels. Other tools are used to model power plant dispatch.

However, the standard IRP process was not designed to efficiently address a host of new factors, such as:

Market volatility and the correlation of markets for power, demand, and fuels;

Market liquidity and poor market information;

Risks associated with poor and declining supplier credit;

The greater number of supply options (build, buy, toll, financial and physical hedges); and

Utilities' risk tolerance.

Integrating Resource Planning With Risk Management

Managing in today's energy environment requires the proper identification, measurement and mitigation of risks associated with the extreme volatility in fuel and power markets, as well as regulatory issues, counter-party credit risk, operational risk, and others. Risk management tools that provide executives the feedback required to mitigate market risk typically are focused on an operating horizon of one to five years. Those same tools may also play an effective role in long-term resource planning by adopting a new integrated risk perspective.

The current tools used by utilities for resource planning were developed in the 1980s, but today's industry is characterized by immense volatility. The deterministic resource modeling of the regulated era needs to be upgraded. Instead, a probabilistic framework should be deployed to identify not only the expected outcomes for rates, supply costs and earnings, but also a distribution of potential outcomes consistent with observed characteristics of any significant volatile elements. This stochastic approach will help utilities support their decision-making process more effectively to mitigate the potential for intolerable outcomes.

For short-term exposures, Value at Risk (VaR) is a valuable tool. At specified intervals, VaR can identify the risk of imminent migration of commodity prices (fleeting hedging opportunities) as a function of market volatility. Price variability can be estimated directly from market quotes, or as implied in option prices. In its more comprehensive applications, VaR may serve as a foundational element in assessing earnings risk or the risk of customer bill increases. For the operational horizon, VaR and its offshoots are metrics that allow utilities to make short-term hedging decisions.

When assessing alternative resource plans, the analytical underpinnings of VaR may be extended to assess risks for longer horizons. This may be done with a Monte Carlo analysis where the objective is to assess states of the market that might exist as potential outliers in a future probability distribution. To do so requires quantitative observations of market characteristics and related uncertainties; necessary inputs include today's forward market prices for power and fuel, price volatilities, and the characteristic variability of observed volatilities, correlations, mean reversion characteristics, etc. A typical result of this step might conclude with a table of potential market environments as seen in Table 1.

This stochastic representation of market values, which may include further assumptions regarding regulatory outcomes or other items, is then used to drive generation dispatch simulations for the range of market environments.

Other factors such as credit risk should also be integrated. Note, for example, that abrogation risk might be considerable if long-term contracts are signed with a single counterparty that ultimately may be counted on to deliver $5 fixed-price gas in an $8 market.

In the RIRP™ process (), we do not weigh the outcomes by their probabilities. A 2 percent probability of incurring a $300 million problem is not just $6 million on a weighted basis; it is dealt with as a discrete issue because the $300 million problem could bankrupt the company.

On reviewing results, management is faced with metrics that are very similar to those deployed in managing risks for the shorter operating horizon. Benchmarks for evaluation are similar as well; some typical questions and objectives might be: 1) What is a tolerable range for earnings or rate variability for Earnings at Risk (EaR), and Rates at Risk (RaR)? 2) What long-term exposures need to be protected so that the economics of investment decisions are looked after?

Finally, the resource plan that is selected will narrow the range of unfavorable outcomes that management might have to deal with in the future. The RIRP™ process begins with a risk management perspective to better optimize planning decisions, and then comes full circle in that the firm's ongoing risk management program will deal with the residual risks that could not be mitigated through the RIRP^TM.

Verify Program Objectives and Acceptable Risk Boundaries

The consolidated risk profile generates an expression of the utility's risk tolerance and forms tolerances of commodity risk exposure that the utility may want to manage. The primary considerations with respect to forming either short- or long-term risk boundaries include the following factors:

Net rate impact associated with uncovered commitments;

Net revenue risk of significant contingencies; and

Risk of appearing imprudent due to unfavorable current fixed positions.

It is important to clearly articulate objectives consistent with constraining risk exposures using the same metrics contained in the initial risk profile. Below are some examples of corporate risk objectives. Typical objectives might be, within a 97.5 percent confidence band, the following:

Manage the volatility to constrain the potential for unfavorable rate outcomes to no worse than a ___ increase;

Limit the credit exposure from any one supplier to ___;

Limit hedge positions to assure that the aggregate cost of fuel and purchased power will not diverge unfavorably from market by more than ___ per kWh; and

Limit forced outage exposure to ___ million for any cumulative short-term outages and ___ million for any 30-day or greater outage.

The next step guides the procurement and management decisions. The objectives will govern issues such as reliability, volatility, liquidity, and volume risks and will be related to, but managed separately from, the objectives defined for the operational risk management program. For example, the risk program typically is focused entirely on defensive risk management objectives. In addressing the overall objectives of the resource portfolio, tradeoffs between risk and other objectives are contemplated. Matching the long-range objectives to both the current and expected conditions over a longer time horizon in light of the prevailing market fundamentals will allow for the construction of the portfolio needs to be pursued in order to achieve its objectives.

Defining Resource Options

The next step is to institute a longer-term process that will build on the risk metrics and objectives developed above. Procurement options consist of a mix of a utility's own generation, plus forward contracts, gas tolling arrangements, unit contingent options, financial options, transmission services, congestion management, and capacity exchange options.

To address longer term contractual assessments, it is important to utilize market insights and procurement expertise to review the condition of the utility's existing physical energy portfolio and management strategy, identify gaps between the current and projected position condition and longer-term objectives, and structure initiatives and implementation steps to build a physical energy portfolio and management strategy that is consistent with the objectives and the risk parameters defined above.

This step includes a summary of historical transactions, and a description of term energy and capacity commitments. While reviewing the current condition, the approach would be to structure recommendations to the utility for analyzing and projecting market fundamentals relevant to the utility's physical procurement needs.

Once needs are assessed and objectives for the portfolio are clearly defined in the context of a longer-term risk profile, various supply options can be identified and analyzed. These supply options can consist of different mixes of plant builds, plant acquisitions, forwards and tolls, existing resources, options and financial hedges that can meet load and sales requirements as a function of market prices and their associated volatilities.

Market Volatility and Simulation

It is necessary to define different combinations of supply options and test them against a combination of load and sales configurations, and market and fuel price combinations with proper correlations of these conditions. The analytic approaches involve using both spreadsheet and Monte Carlo simulation tools in combination with dispatch algorithms as described below. Then we test our portfolios to determine whether there are significant exposures in the context of confidence intervals of likely load and price outcomes. This process is presented in Figure 2.

Credit Risk: Betting the Company

For each of these alternative supply portfolios, a list of creditworthy suppliers must be identified and credit risk evaluated. The lowest expected cost options may have a high degree of credit risk if supply is concentrated among suppliers with poor credit ratings. We explicitly consider how much the utility is willing to risk in its portfolio lower credit suppliers. Historical data on failures can be correlated with the credit ratings of the suppliers. In this way, the risk of contract abrogation can be evaluated for different supply portfolios.

All portfolio options are valued using the metrics defined in the objectives (earnings, costs or rates) and are characterized by a probability distribution. The ultimate portfolio provides the best outcome distribution over a wide range of outcomes and avoids the disastrous outcome.

The optimal reconfigured supply portfolio will both reduce the basis of the VaR for operational risk management over time, but also provide the best alternative across a range of potential market and regulatory outcomes. Figure 3 shows how these earnings or rates at risk might be represented graphically for each portfolio alternative.

The final step is to attain its future state portfolio and manage its long-term effectiveness in a dynamic marketplace. Through this process, we provide: (a) an outline of the data requirements necessary for monitoring the electric and gas markets, and a white paper assessment of the drivers that affect the initial energy portfolio; (b) an interim report on the tradeoffs of risk and market opportunities among the portfolio combinations; and (c) a final report that documents the desired portfolio reflecting both cost and risk tolerance parameters and a means of achieving the desired strategy. This last identifies system needs to manage the portfolio on an on-going basis.

Case Study: The Wires Utility and the Duty to Serve

Consider a typical utility today. Utility A sold off its generation several years ago, but it has a provider of last resort (POLR) requirement to fulfill for its residential and commercial customers. As it considers its options in a standard integrated resource planning framework, it must address the following complications:

Its load is not only a function of demand growth, but also the stickiness of its customers, and the possibility that if market prices exceed the price it can charge to its POLR customers, a large volume of customers will suddenly return to the POLR utility;

Predictions of gas prices are more uncertain than ever before. The range of market prices is from $2/MMBtu to $10/MMBtu and gas price volatility is likely to remain high. Figure 1 shows both power and gas prices over the past several years;

While forward power market prices exist, the lack of liquidity of these markets makes them more volatile than one might expect. A few deals can change long-term market perceptions dramatically;

A few years ago, merchant suppliers were considered reliable sources to supply load requirements, but now many of those suppliers are either bankrupt or their credit is close to junk bond status. Long-term contracting without owned generation is now considered risky and must be carefully managed;

The number of available hedging instruments has increased dramatically:

Physical hedges include the acquisition of a merchant plant to cover the POLR swing and market price hikes and credit issues or tolls to manage the fuel risk more directly; and

Financial hedges include annual, seasonal, or daily options, insurance to cover plant outage risk, etc.

Utilities are concerned about risk, but they may not have an accurate measure of earnings, revenues, costs, or rates at risk. Most utilities that do measure risk, measure only trading risk, and then only for short periods of time and not on an enterprisewide basis. For longer term decision-making, some utilities have no established risk measure at all.

Appendix: The Process

The RIRP^TM approach begins with a review of company objectives and perspective on the company's tolerance for risk. All utilities focus on managing costs and rates, while most IOU's also focus on maximizing shareholder value through earnings and dividends. With policy-level objectives as a backdrop, the RIRP™ program consists of the following steps:

Define market risk exposures;

Reach consensus regarding objectives and risk tolerance;

Contrast risk exposures with tolerance and objectives;

Screen viable resource options;

Conduct simulation of market prices and volatilities to assess a distribution of future states;

Evaluate performance against objectives and tolerance for each plan alternative;

Address credit risk issues; and

Rank plan alternatives by key metrics.

The first steps in this process deviate from the traditional IRP approach. The process begins with an assessment of risk exposures and building a consensus perspective among the management participants of that risk. It is important to fully articulate the utility's circumstances and objectives. This serves three purposes:

It offers a framework to view risk from a clear and manageable perspective;

It formalizes those corporate, operational, and philosophical views that help formulate a targeted risk approach; and

The process fosters better convergence among all management participants in the articulation of directional objectives.

With a better understanding of the corporate objectives and business philosophy, the next step is to analyze the utility's risk exposure. For each energy-related risk element, we quantify market price volatility within statistical confidence bands. Most of the risk effects can be quantified from the understandings derived from the initial consensus-building combined with selected quantitative data, such as a forecast of natural gas consumption, forecast of electric generation and load requirements, correlation between fuels, water flow and electric generation/consumption, fuel and power commitments, rate pass-through mechanisms, and hedged positions.

The current risk profile presents the utility's commodity risk profile in terms of VaR, with an appropriate focus on the potential financial impact of commodity price movements to the utility and its customers. For purposes of evaluating plan alternatives, this measure could be extended to look at the VaR of the overall portfolio over an extended period so that longer-term decisions may be analyzed.

The risk profile considers both open-position risk as well as fixed-position risk. Open position risk (VaR-OP) measures the financial impact to the utility from increasing commodity prices in the absence of hedging, while fixed position risk (VaR-FP) measures the financial impact of declining commodity prices, and how such a price decline might leave management looking imprudent for any hedges that they might have placed. The risk profile incorporates material volatile elements, and considers any natural mitigation effects implicit in the diversity of costs and revenue streams (i.e., covariance effects).

The next step is to produce a consolidated risk profile. This work product shows the critical risk elements, and primary exposures of the utility's natural positions on both a commodity specific and an aggregate basis. Further, it delineates actionable items, and provides a fundamental structure from which to progress a fully constructed Risk Management Program (the short term operational program that is linked but separate from the RIRP™ process).

Once this is accomplished, it defines the metrics that will be used to evaluate alternative portfolios. It is important to use dispatch and financial tools that can accommodate representations of market volatility and correlation among key variables. -GV, MG, AH, TLB, and RM

Articles found on this page are available to subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.