Robust analysis of energy asset value and risk

There is no textbook which comprehensively covers the analysis of value and risk in energy markets. Energy companies are instead confronted with a diverse choice of analytical techniques, some specific to the energy industry and some adapted from the financial services industry. Choosing an appropriate analytical methodology is only the first step in an analytical exercise. Having a deep understanding of both implicit and explicit model assumptions, rigorously challenging these and seeking to benchmark results are of equal importance in obtaining a successful outcome…

September 12, 2011

There is no textbook which comprehensively covers the analysis of value and risk in energy markets. Energy companies are instead confronted with a diverse choice of analytical techniques, some specific to the energy industry (e.g. gas storage optimisation) and some adapted from the financial services industry (e.g. Value at Risk). Choosing an appropriate analytical methodology depends on the specific characteristics of the asset, contract or portfolio considered. It is also important to capture the way the asset/contract/portfolio will actually be optimised and hedged and the associated dynamics of underlying markets. But perhaps the most obvious but most neglected principle of effective analysis of value and risk is the rigorous challenge of analytical assumptions and results.

What is the problem?

Whether valuing a simple spot indexed gas contract or measuring the risk on a complex portfolio of flexible generation assets, it is good discipline to take a step back and deconstruct the problem to be analysed before deciding on the most appropriate analytical approach to apply. 

Some useful questions applicable to both valuation and risk measurement are:

  • What risk factors should be considered and how can uncertainty around the evolution of these risk factors be captured?
  • What exposures are present and how will these be optimised and hedged?
  • Of the analytical approaches available, what is the trade off between effectiveness, complexity and transparency?
  • What could go wrong with the analysis? What assumptions are implicit in the analysis and are they valid?  What is the potential impact of these assumptions breaking down?
  • How can analytical results be verified by applying an alternative approach or looking at external benchmarks?

Considering these questions in a structured fashion not only helps define an appropriate analytical methodology but makes it much easier to effectively apply and communicate the results.

Capturing uncertainty

Uncertainty is an inconvenient reality of operating in energy markets. Basing decisions on simple scenario forecasts of the future does not do justice to this reality. The greatest threats from uncertainty typically come from commodity prices and asset availability. The most effective way to capture the impact of uncertainty on asset value and risk is to apply a combination of stochastic modelling techniques and well constructed stress tests. We have explored stochastic modelling of commodity prices in a recent post. We plan to explore the issues associated with the impact of uncertainty on specific assets (e.g. power plants, interconnectors & gas storage assets) in a series of articles to follow.

Capturing asset management strategy

Every valuation and risk measurement methodology has an implicit hedging and optimisation strategy embedded within it. Often over looked it is important to explicitly consider this strategy. There is little point in conducting complex analysis of how an asset could theoretically be managed or hedged if in practice it will be managed differently.

For example, a gas and power origination team tasked with building up a portfolio of structured assets may be supported by an analytics function capable of building complex models that implicitly assume delta hedging strategies. But if the trading desk chooses to optimise and hedge the asset optionality on a more simple rolling intrinsic basis there is a clear divergence between the basis on which the contracts are being valued and managed. This gives rise to the risk that the value implicit in the execution price for the asset will never be realised.

Table 1 below describes the hedging and optimisation strategies implicit across a range of common valuation and risk methodologies and the pricing models required to support analysis.

Table 1: Implicit hedging and optimisation strategies


It can be seen from Table 1 that it is important to capture the evolution of forward price dynamics if hedging strategies impact asset value or risk. For example consistent modelling of spot and forward prices is required to effectively measure Earnings at Risk on a portfolio with a rolling hedge profile. In the case of delta hedging optionality embedded in a flexible contract or asset, only spot prices need to be considered as all exercise decisions will be made in the spot market. It should be noted that the output of these models should allow delta positions to be created to estimate the levels of forward hedges to remain delta neutral.  We plan to come back and look at hedging strategies for asset and complex contract optionality in a series of future articles.

Testing assumptions

When tackling an analytical problem it is easy to allocate most of the available time to devising a modelling approach, building the model and running it to get a set of results. But the first question a commercial decision maker is likely to probe before relying on the modelling is the sensitivity of the results to input assumptions. Sensitivity analysis can be conducted to test both the model and the behaviour of asset value and risk. However, what is typically overlooked is the task of challenging the implicit assumptions that may exist within a methodology, checking their validity and assessing what might happen if they break down.

This is best demonstrated by an example.  A common approach for valuing gas or power interconnector capacity is to use a spread option, capturing the right to receive the difference between the price of the two connected markets. So what implicit assumptions are there in this approach? 

  • If the markets are in different currencies most closed form spread option models would overlook the exchange rate risk factor.
  • If the actual flow decision is made on a daily basis, the option may be specified as a strip of daily options.  However, the market contract granularity that is actually available for hedging value will be much less granular.
  • Models typically assume that volumes can be immediately hedged at no liquidity risk or bid/offer cost whereas in reality these factors are likely to have a significant impact on value.

Benchmarking results

Benchmarking results against alternative methodologies or benchmarks is another important way to build confidence in results.  Most valuation and risk analysis in the energy industry is based by necessity on detailed ‘bottom-up’ modelling.  Alternative approaches help to bound the problem and inform decision makers of the potential range of outcomes.  For example, if the gross margin calculations for a gas fired power station investment are based on running thousands of power, gas and carbon prices simulations through a detailed plant dispatch algorithm, a quick and simple alternative benchmark can be provided using a simple spread option valuation approach. 

Some useful principles in benchmarking results are:

  • Wherever possible, use market benchmarks.  These could be implying asset values from share prices or any available asset or contract price information.
  • Use top down or “back-of–envelope” calculations that commercial decision makers can more readily associate with.
  • Try creating scenarios of extreme pessimism and extreme optimism.  Judging where results lie between these extremes is likely to be informative.

Pragmatism and commercial knowledge are as important as PhDs

The complexity of some of the analytical challenges presented by the energy industry is not far off those of nuclear physics and rocket science. As a result it is easy to forget that analysis is a means to an end not an end in itself. It is convenient to try and adapt the characteristics of an asset, market or hedging strategy to suit the constraints of an analytical methodology. But effective analysis is built on a foundation of recognising the complexity of a problem and understanding and challenging the limitations of analytical techniques. Most importantly, theory and methodology are only powerful tools if they are coupled with a strong commercial understanding of the assets and markets to which they are applied.

Other relevant articles:

Stochastic Pricing Models: Science, Art or Voodoo?