THEORETICAL BACKGROUND 3

Fuel Price

Uncertainty and

the Optimal

Generation

Portfolio

According to Markowitz (1952), investors in financial markets have always had to deal with exposure to risk. Modern Portfolio Theory (MPT) has been developed to help manage this uncertainty based on the concept of an “efficient portfolio” which has the smallest attainable risk for a given level of expected return. In an uncertain world this approach provides a very useful format for considering the trade off between minimising the expected price of electricity and minimising the risk inherent in any particular choice of fuel mix. It reflects the fact that forecasts are always uncertain and what may be forecast to be the best option assuming a particular price scenario may well turn out to be much more expensive than expected in the long run. This approach assumes that, as well as knowing something about the likely future growth in the prices of energy, policymakers have information on the volatility in energy prices and how the volatility in the price of individual fuels is correlated. In the case of energy prices there is good reason to believe that gas and oil prices will continue to be quite highly correlated in the future, as in the past.

The theory focuses on identifying either the return-maximising or the risk-minimising bundle of assets. The methodology uses a mean- variance analysis, based on empirical prices, and compares the average returns of different bundles of assets (portfolios of

generation plant) with the associated level of risk. In any economy there are two types of risks that an investor faces, systematic and unsystematic. The former has a similar affect on each player, and is associated with the market environment generally and is, therefore, non-diversifiable. An example of such a risk would be the increasing global dependence on gas. Although some countries will be more exposed than others to gas price increases, each country will suffer a comparable effect in the price of gas from a gas price shock. Unsystematic risk refers to the specific risks that each firm faces. It is up to the individual firm to try to hedge against these risks as best it can. This exposure to price fluctuations is avoidable through portfolio diversification and can be best solved through the use of Modern Portfolio Theory. Adding generation capacity to the portfolio whose fuel prices are uncorrelated with the current mix, acts as a hedging mechanism against future price uncertainty. To the extent that these additional types of generation capacity have higher costs (lower returns) than the ones that they replace, the risk reduction comes at a cost in terms of expected future price (return). In relation to the example given, this could mean using fuels other than gas in electricity generation to reduce risk, even though gas is expected to be the cheapest fossil fuel for electricity generation.

This methodology takes account of expected cost and expected risk, calculated as the weighted average of all possible outcomes. MPT puts an economic value on price stability. An example of the trade-off between risk and price is the choice by many homebuyers to adopt fixed rate instead of variable rate mortgages. The extra interest that is incurred by the guarantee of fixed repayments, is the price of certainty for each customer. Adapting this methodology to the choice of the optimal electricity generation portfolio is potentially useful in identifying the price to be paid for greater price certainty. For example, it permits the comparison on economic grounds of the cost of a wind turbine – characterised by a high capital expenditure, very low operational costs and no fuel cost – with the CER Best New Entrant (BNE) plant, a gas-fired CCGT that incurs relatively lower costs of capital per MW of capacity, but higher operation and maintenance costs and a substantial and uncertain fuel cost element. A well-diversified portfolio of generating assets will include a mix of technologies whose underlying fuel prices are uncorrelated with each other. This is the best means of hedging future uncertainty and similarly ensures that any increase in cost is due to market risk and is unavoidable. However, such hedging behaviour comes at some cost in terms of higher expected prices.

A generating technology with costs that are statistically negatively correlated to the rest of the portfolio can help mitigate portfolio cost swings… Current evaluation of energy mixes understates the cost of fossil fuels and value of renewables (Awerbuch, 2004).

Most fossil fuel prices, with the exception of peat prices, are to some degree positively correlated, and the volatility that these prices exhibit often has a negative impact on economic activity.

Consequently, increasing the amount of fixed cost generation, such as wind, in the portfolio, whose price is uncorrelated with any other fuel, will add price stability for the future. Greater certainty regarding costs emerges as a reduction in risk in the portfolio analysis, albeit at the cost of an increase in the expected price.

Figure 3.1: Illustrative Portfolio

0 10 20 30 40 50 0 0.5 1 1.5 Volatility Expected Cost 100% Fossil Fossil &Renewables Fossil & Ren (initial risk)

In Figure 3.1, we show a stylised example of three different portfolios of fossil fuel and renewable generation: all fossil fuel, and two different mixes of fossil fuel and renewables. Beginning with the fossil fuel and renewables (initial risk) mix we can see that in this stylised example it has an expected cost of around 40 and volatility of 0.5. It is significantly lower in volatility (lower risk) than the all fossil fuel portfolio, though having a significantly higher price. However, the alternative mixed portfolio could produce electricity at a lower cost (20) than the fossil fuel only option, while being equally risky as the fossil fuel only portfolio. In other words, for the same risk profile it may be possible to choose an alternative mix of generation that reduces the expected future cost of generation.

The lessons to be drawn from this approach suggest that an optimal portfolio of generation technologies should give an additional premium to technologies where the fuel price is fixed (renewable) or has a negative (peat) or low positive correlation (coal) with the price of gas and oil. If there is concern to reduce the risk of the portfolio there should be more wind, coal and peat than would be implied by a simple analysis based on a single forecast for future energy where prices are assumed to be certain. Obviously the weights in the desired portfolio will also be affected by the forecast movement in prices. For example, if the price of peat is expected to rise rapidly because of its high carbon content, this could offset its advantage in having a fairly stable basic price when carbon emissions are excluded.