ap-4-10.dvi Acta Polytechnica Vol. 50 No. 4/2010 Option Derivatives in Electricity Hedging P. Pavlatka Abstract Despite the high volatility of electricity prices, there is still little demand for electricity power options, and the liquidity on the power exchanges of these power derivatives is quite low. One of the reasons is the uncertainty about how to evaluate these electricity options and about finding the right fair value of this product. Hedging of electricity is associated mainly with products such as futures and forwards. However, due to new trends in electricity trading andhedging, it is also useful to thinkmore about options and the principles for workingwith them in hedging various portfolio positions and counterparties. We can quite often encounter a situation when we need to have a perfect hedge for our customer’s (end user consuming electricity) portfolio, or we have to evaluate the volumetric risk (inability of a customer to predict consumption, which is very similar to selling options. Now comes themoment to compare the effects of using options or futures to hedge these open positions. From a practical viewpoint, the Black-Scholes prices appear to be the best available and the simplest method for evaluating option premiums, but there are some limitations that we have to consider. Keywords: option derivatives, electricity hedging, evaluation models, electricity prices. 1 Key features of electricity prices Some years ago, the electricity market was vertically integrated and the prices of this commoditywere fully regulated by state-owned authority managed world- wide. These regulated prices had to reflect the cost of electricity generation, transmission and distribution. The prices were therefore determined by well-known factors, and changed only rarely. Since deregulation of the electricity market, prices have been determined according to the economic rule of supply and demand. Many countries have settled electricity pools, where bidsof electricity sellersarematchedwith thepurchase orders of end users. These pools trade with long-term products and also with short-term products. The dif- ferences areonly in liquidity andvolatility. This dereg- ulation has fully supported trading activities on the derivatives markets, which allow trading with finan- cial electricity contracts as derivatives, where electric- ity is the underlying asset. The relatively high volatil- ity of electric power and important specifics of this commodity have forced many market players to man- age price risk professionally. Hedgingmarket risks is a well-knownway to eliminate the risk of price changes, but there are also weak points, which are associated with specific features of electricity. Due to the obvious specific features of electricity, e.g. its unique nonstora- bility, electricity prices are more likely to be driven by spot supply and demand, which is inelastic. Any shock in consumption or production may give rise to price jumps [4]. 2 Hedging As the electricity market becomes deregulated and more competitive, changes in supply and demand are increasingly translated into price volatility and fluc- tuations. Another very important driver has been fi- nancial crises, which have shown us the impact of fi- nancial derivatives traded also on behalf of electricity contracts. Most of the volatility of the fluctuations in supply and demand is visible on the daily spot mar- ket, where the price is mainly influenced by inelastic demand and short-term supply. Figure 1 shows the increasing volatility of spot prices in recent years. Fig. 1: Spot prices of electricity in �/MWh (source http://www.eex.com) Mostderivatives, however, arenot typicallyused to hedge risks connectedwith daily price volatility. They 65 Acta Polytechnica Vol. 50 No. 4/2010 are used to hedge risks associated with trend fluctua- tions and seasonalprice volatility. Marketparticipants therefore often use annual andmonthly derivatives. In a competitive electricity market, daily fluctuations in electricity prices will therefore be the most dramatic driver of price volatility. There are two different ap- proaches depending on the type ofmarket participant. Generators, as entities owning power plants, have a natural “long” electricity position, and the value of this position increases and decreases with the price of power. When power prices increase, the value of the electricity produced increases, and when power prices decrease, the value of the produced commod- ity decreases. An electricity consumer is naturally “short” and, in the opposite way, consumers benefit when prices go down and have to suffer loss when prices increase. Price volatility introduces new risks for generators, consumers and traders (brokers). In a competitive electricity market, generators will have to sell some of the electric power that they produce in volatile spot markets, and will bear the risk if the spot prices are lower than the generation costs. In the case of consumers, we have to consider higher sea- sonal and hourly price variability. It is obvious that this uncertainty could make it more difficult to assess andmanageacustomers’s long-termfinancialposition. Electricity futures and other power derivatives help electricity generators and end consumer to hedge price risks (market risk) in a competitive electricitymarket. Futures contracts are legally binding and negotiable contracts that call for future delivery of electricity. In many cases, physical delivery does not take place, and the futures contract is closed by buying or selling a futures contract on the delivery date. Other power derivatives include options, price swaps, andOTC for- ward contracts. Power derivatives, like futures and options, are traded on an exchangewhere participants are required to deposit margins to cover all potential losses due to the credit risk of the counterparty and the market risk of an open position. Other hedging instruments, such as forwards, are traded bilaterally “over-the-counter”. Recently we have seen how ex- pensive membership of a power exchange is, and we can compare the costs of financial capital for margins with paid option premiums. 2.1 Short-term or long-term hedging It is relative difficult to define a strict boundary be- tween long-term hedging and short-term hedging. We could define short-term hedging in terms of the most distant maturity traded month contract on the power exchange. This could be between 6 and 12 months. The main important risks associated with short-term hedging are cash-flow problems. In the first case, a cash-flow problem emerges from an insufficient initial and variation margin for MtM (“Mark-to-Market”). The result is that the intendedhedging transactionbe- comes to speculative position after a margin call that we are not able to pay for. In the second case, we can consider an unhedged price risk, which results from inadequate hedging of open positions. This case very often occurs and is associated with volumetric risk. Most electricity consumption depends on short-term conditions, and there are not enough strict plans or “take or pay” contracts, which will motivate the end customer to consume in orderwith the contracted vol- ume. Gains and losses from hedging activities that occur in the futures market when a hedge is under- taken must be viewed as part of the electricity price that the market participant provides to its customers. The same approach has to be undertaken in the case of option premiums. Sometimes amarket player takes a profit in the fu- tures market and loses in the spot market, and some- times the reverse situation occurs. It is clear that hedging profits and losses must be treated simply as part of the cost of purchasing energy. With an im- perfect hedge, the market player could earn less on his futures position than he loses between his fixed price contract and the spot market, or he could earn more. There is no clear line to distinguish long-term and short-term hedging. The cash flow risk increases exponentially due to margin calls as the maturity of the long-term hedge increases. We consider that the increase in risk is faster than linear, for two reasons. Firstly, the price volatility increases approximately in proportion to the square root of the length of the hedge, and secondly, the amount being hedged is gen- erally proportional to the length of the hedge, because the market player will be hedging an constant volume over the time. The primary risk associated with long- termhedging isagainassociatedwithmargincalls risk. Nowwe can compare forwardand futures contracts. A key difference between forward and futures contracts is in cash settlement, which is performed by a clear- ing bank in the case of futures. A buyer or seller of a futures contract will have to realize short term losses or gains as the futures price changes. This cash settle- ment is performed daily. In the case of a forward con- tract, profit and loss is realized only at maturity and there is no cash flow problem due to the payment of a variation margin. There is another more important specific consideration which could make forward deal- ing less interesting for smaller business units, and that is the credit risk exposure of an electricity seller. In the case of futures, this credit risk and also themarket risk is solvedbyMtM (daily cash settlement) clearing. It is obvious that the money lost on the future is en- tirely regained from the addedprofit on the fixed price contract that was sold at the start of this example. If the loss is quite large, it may be impossible for the hedging market participants to raise the cash margins necessary to meet the variation margin requirement. 66 Acta Polytechnica Vol. 50 No. 4/2010 In this case, the clearing bank has the right to liqui- date all open positions of the counterparties. Hedging over longer periods puts traders at risk for extremely large margin calls. The consequence is that long-term hedging requires significantfinancial resources tomeet variation margin requirements. 2.2 Options In 1996, NYMEX introduced options for electricity. There are two types of options for electricity: a put option (“floor”) and a call option (“cap”). In the first case, the buyer of an electricity put option pays a premium for the right, but not the obligation, to sell electricity at a specified price, the strike price or exercise price, at a specified exercise time. End users use call options to place a maximum cap price that they will pay for the commodity at a specified exer- cise time. Market participants often use combinations of calls and puts to ensure a particular price range. Generators often use put options to guarantee a min- imum price of the produced electricity in conjunction with the physical sale of electricity. By this product, a generator could benefit from increases in commod- ity prices, but would avoid the risk of lower prices. Consider that the futures contract price is �43/MWh and the generator of electricitywould like to receive at least this amountdue toprofitanalysis. Therefore, the generator has to purchase a put option, for�2/MWh, which thegeneratorwill pay for. If thepriceof electric- ity increases, the generatorwill sell electricity into the spotmarket and receive the higher spot price (see Fig- ure). If the price of electricity falls, the generator will sell electricity to the option holder for �43/MWh, or hewill sell his option at its exercise value,�43/MWh, on or before its expiration date. Fig. 2: Put option of an electricity producer A consumer, end user, deals with the opposite problem. In the case of hedging, he would utilize a call option to avoid the risk of higher prices, while re- taining the ability to participate in potentially lower prices. Let us assume that the futures contract price is �40/MWh and the consumer would like to pay no more than this price. In this case, the customer will buy a call option; say for �2/MWh, which the end users have to pay in advice. If the price of electric- ity falls, the consumer will buy electricity in the spot market. If theprice goesup, the enduserwill buy elec- tricity from the option holder for �40/MWh or will sell his call option for its exercise value, �40/MWh, on or before its expiration date. 3 Option evaluation Market models for evaluating derivatives and options work mainly with storable commodities. The non- storability of electricity implies a breakdown of the relationship between the spot price and the forward price (Eydeland and Geman 1988). The second prob- lem is with convenience yield, which is important in many cases of commodities. The convenience yield is a numeric adjustment to the cost of carry in thenonar- bitrage pricing formula for forward prices in markets. In the case of the money market we can consider the following situation. Let Ft,T be the forward price of an assetwith initial price St andmaturity T . We sup- pose that r is the continuously compounded interest risk free rate. Then, the non-arbitrage pricing formula of the forward price is: Ft,T = Ste r(T −t) (1) However, this relationship does not exist in most commodity markets, partly because of the inability of investorsandspeculators to short theunderlyingasset. Instead, there is a correction to the forward pricing formula given by the convenience yield c. Hence Ft,T = Ste (r−c)(T −t) (2) The convenience yield exists because owners of the assetmayobtain a benefit fromphysically holding this asset as inventory to maturity. These benefits include the ability to profit from temporary shortages. Any- one who owns inventory has the choice between con- sumption today versus investment for the future. A rational investor will choose the outcome that is best. When inventories are high, this suggests anticipated relatively low scarcity of the commodity today versus sometime in the future. Otherwise, the investorwould sell his stocks and the forward prices Ft,T of the asset should be higher than the current spot price St. This tells us that r − c > 0. The reasoning become inter- esting in the case of low inventories, when we expect that the scarcity now is greater than it will be in the future. Therefore, the investorwants to borrow inven- tory from the future but is unable. We expect future prices to be lower than today and hence Ft,T < St. This implies that r − c < 0. The concept of convenience yield was introduced by Kaldor(1939) and Working(1949) for agricultural commodities. This concept represented the benefit from holding the commodity as opposed to a forward contract. The concept of convenience yield does not 67 Acta Polytechnica Vol. 50 No. 4/2010 make sense in the caseof electricity, because there is no availablemethod to store electric power, and therefore we cannot consider the benefit from storing the com- modity versus storing costs. The first, very important, characteristic of electricity prices is a mean reversion toward a level representing the marginal cost of elec- tricity production, which can be constant, periodic or periodic with some trend. In the case of electricity we have to expect the mean to revert to a deterministic periodical trend driven by seasonal effects. The sec- ond specific driver of electricity prices is the existence of temporary imbalances of supply and demand in the network which affects randomprice moves around the averagetrend. Wearenot able topredict this effect. A third feature is the jump character of electricity prices (spikes), because shocks in power supply and demand cannot be smoothed away by inventories. As wasmentioned above, the convenience yield at- tached to a commodity can be interpreted as a con- tinuous dividend payment made to the owner of the commodity. Then we could suppose that the price of the underlying asset is driven by a geometric Brown- ian motion and use Merton’s (1973) formula for pric- ing options (3). This formula provides the price of a plain vanilla call option written on a commodity with price S: C(t)= S(t)e−y(T −t)N(d1) − ke−r(T −t)N(d2), (3) where d1 = ln(S(t)e −Y (T −t) ke−r(T −t) )+ 12σ 2(T − t) σ √ T − t (4) d2 = d1 − σ √ T − t (5) As was mentioned above, the main difficulties in valuing power options are due to the fact that this commodity cannot be stored and the associated prob- lem with convenience yield. The results of using for- mula (3) are very similar to the prices at the EEX pool. All inputs and results are obvious from Tab. 1. Fig. 3: Market data of a traded option (call/put, source http://www.eex.com) Table 1: Inputsand results of a comparison ofEEXmarket data and an evaluation of formula (3) Market Data (EEX pricing) contract CAL2011 futures �/MWh 47.58 �/MWh premium �/MWh Strike price CALL Option PUT Option �/Mwh 49 2.47 3.88 50 2.11 4.51 51 1.80 5.19 53 1.30 6.67 55 0.92 8.28 57 0.65 9.99 58 0.55 10.87 59 0.46 11.77 64 0.18 16.45 Results of Black — Scholes formula (3), volatility approx. 16 % Strike price CALL Option PUT Option �/Mwh 49 2.12 3.51 50 1.76 4.13 51 1.45 4.80 53 0.96 6.28 55 0.62 7.90 57 0.39 9.63 58 0.30 10.53 59 0.24 11.44 64 0.06 16.17 Fig. 4: Comparison of market data(EEX) with result of the B-S formula (3) 68 Acta Polytechnica Vol. 50 No. 4/2010 The primary category of traded electricity options includes calendar year contracts and monthly physi- cal options, which are also traded at pool EEX. Call options allow the buyer to receive power at the strike price. These options are relatively liquid. See the next figure with pricing of a set of options divided by the strike price. The results of using the B-S formula (3) are relevant in this case and after a comparison (see Fig. 4) with market pricings of options, they are very similar. The second categories of options are daily options, or options for a block of hours. These (Asian type) options are specified for a given period of time and can be exercised every day during this period. It is obvious that daily options are very difficult to man- age and are not liquid. Therefore it is important to work better with pricingmodels associatedwith jump characteristics andmean reverting of electricity prices. 4 Conclusion Liberalization of the energy industry requires adapta- tion to risk-management techniques. Pricing or selling derivative products poses new challenges for market participants. The non–storability of electrical power and the inability to hold short positions in electricity spot prices has eliminated the utilization of techniques from financial mathematics. In the case of European type of derivatives in energy markets, namely options (call, put and collar), it turns out that these derivative prices are related to the standard Black-Scholes op- tion pricing formula. The main important parameter for calculating fair value is volatility. These findings are positive, since the standard tool for pricing option derivatives can also be applied in powermarketswith- out losing the influence of specific market features. Acknowledgement The research described in this paper was supervised by Prof. O. Starý, FEE CTU in Prague References [1] Borovkova, S., Geman, H.: Analysis andModeling of Electricity Futures Prices, Studies in Nonlinear Dynamics & Econometrics, Nonlinear Analysis of Electricity Prices, Vol. 10, Issue 3, Article 6, The Berkeley Electronic Press, 2006. [2] Hull, J.C.: Options, futures, and other derivatives. 6th edition, Prentice Hall, 2006, ISBN 0-13-149908-4. [3] Redl, C.: Modeling Electricity Futures, Energy Economics Group, Vienna University. [4] Stoft, S., Belden, T., Goldman, C., Pickle, S.: Primer on Electricity Futures and Other Deriva- tives, University of California, 1998. About the author Pavel PAVLATKA was born in Ceske Budejovice in 1982. He was awarded a master‘s degree in Febru- ary 2008. He is currently a doctoral student at the Department of Economics, Management and Human- ities, FEE, CTU in Prague. Pavel Pavlatka E-mail: Pavlap1@.fel.cvut.cz Dept. of Economics Management and Humanities Czech Technical University Zikova 4, 166 29 Praha 6, Czech Republic 69