PRICING ASPECTS OF FORWARD LOCATIONAL PRICE DIFFERENTIAL PRODUCTS

Transcription

1 PRICING ASPECTS OF FORWARD LOCATIONAL PRICE DIFFERENTIAL PRODUCTS Tarjei Kristiansen Norwegian University of Science and Technology and Norsk Hydro ASA Oslo, Norway Abstract This paper gives an analysis of how Contracts for Differences (CfDs) are priced at Nord Pool, the Nordic electricity exchange. These contracts were introduced at the exchange the 17 th November 2, and there is now available data for the first trading periods. The traded volume is still small, but the spectrum of products is increasing. CfDs are used in adjusting and hedging portfolios as to price differences between the System Price and the price in the specific areas. We do statistical analysis of the contract prices and the underlying asset, and calculate the correlation between the prices. The paper shows that the market still is immature, and that it is hard to find any consistent relationship between the underlying asset and the contract prices. This may be due to the lack of data, since it was only available information for two seasons (per December 21). This paper is therefore only a first look on the prices. Keywords: Contracts for Difference, contract pricing, statistical analysis 1 INTRODUCTION In Scandinavia Nord Pool organises two different markets for electricity, Elspot and Eltermin. In the Elspot market buyers and sellers trade in a daily spot market concluded at the day-ahead stage. Traders can submit offers to sell or bids to buy the physical electricity they expect to produce or consume (the contracts in the Eltermin are purely financial commitments). The spot price (System Price) is set equal to the marginal offer to supply (or bid to buy) accepted by the exchange. The Eltermin is a futures market for cash settlement of a specified volume of power at an agreed upon price, date and period. The market participants may trade in the futures market up to three years in advance. Futures are used for trading and risk management. Contracts traded on Eltermin are defined with a weekly resolution. Contracts due more than 5-8 weeks ahead are grouped in blocks of 4 weeks each. Contracts that are due for delivery more than one year ahead are combined in seasonal contracts: Winter 1 (weeks 1-16), Summer (weeks 17-4) and Winter 2 (weeks 41-52/53). The spot price is used as a reference price for the futures in the Eltermin market. Statnett (the transmission System Operator (TSO) and the main network owner) operates the regulating market to secure real time balance between generation and load. Electricity traded in this market is called regulating power. Pricing in the regulating market is as follows; if total demand exceeds total demand in the spot market, the price for regulating power is set equal to the highest accepted offer to supply. If total demand in the regulating market is less than total demand in the spot market, the price for regulating power is set equal to the lowest accepted bid to buy. Besides the Nord Pool there exists a bilateral market for over-the counter (OTC) contracts. In this market the most common contracts types are forward contracts with different (fixed) load profiles, options and forward contracts with flexibility in the load profile (load factor contracts). The main difference between the futures and forwards is the daily marking to market and settlement of futures. Forwards are settled when the contract is due to delivery. Transmission constraints within Norway and between Norway and the other Scandinavian countries are handled by splitting the market into price areas. Statnett (the Norwegian System Operator) defines the fixed price areas in Norway according to its information on the likely pattern of flows on the system for a certain period of time. The price areas are named as NO1 and NO2. Sometimes NO3 is used when it is necessary with three price areas. Electricity traders typically buy and sell energy under contracts at particular locations. The only pay spot prices for differences between their contracted and actual deliveries. Contracts allow traders to hedge against the risk associated with spot price variations in one location. However, when congestion is present and the traders face different locational spot prices the must pay a congestion fee (according to the market splitting model used in Norway). This fee is equal to the difference between the Area Prices multiplied by the amount transported across the bottleneck. When traders are hedging the System Price by use of financial market electricity contracts they remain exposed to the possibility that the locational spot price and the System Price will differ. 1.1 Forward locational price differential products During the last years problems with congestion and different Area Prices have been of growing concern in the Nordic area. The 17th November 2 Contracts for Differences (CfDs) were introduced at Nord Pool as a result of growing market demand for hedging instruments related to local Elspot Area Prices. The CfD is a forward market product with reference to the difference between the Area Price and System Price. Nord Pool

2 picked this difference instead of the spread between the Area Prices, which is greater. They mean this will increase the liquidity of the products. The term Contract for Differences (CfDs) has another meaning here than the corresponding term used in the English market. In Nord Pool CfDs are used to hedge against the difference between the two uncertain prices - Area Price and System Price, not as in the English case, against the difference between the spot price and a pre-defined reference price or price profile. The Nordic CfD is a locational swap, while the English CfD is settled based on the difference between the spot price and the reference price. The traded volume is still small, but the 15-th of June 21 yearly contracts were introduced. Table 1 shows the OTC volume traded for the two first trading periods. The number in the brackets is the percentage of the total volume traded. The volume traded is as expected highest for the Summer 21 contracts which has a longer trading period and, with the majority of the contracts traded referred to Helsinki and Oslo. For the Winter 1 contracts most of them traded, are referred to Stockholm. Volumes for the CfDs at Nord Pool were not reported at the home page, but this volume is usually lower than the OTC volume (approximately 1/3 of the total volume is traded at Nord Pool). Including the power cleared via Nord Pool, a total power volume of TWh was handled by the exchange in 2. With the limited information available, it looks like the traded volume of CfD is increasing with time, but more data is needed to make a final conclusion. Reference area Volumes OTC (GWh) Winter 1 21 Summer 21 Århus (4.5%) (DK1) Helsinki 43.2 (7.3%) 918. (25.7%) (FI) Oslo (NO1) 28.8 (4.9%) (23.7%) Stockholm (87.8%) (46.1%) (SE) Total Number of trades 7 48 Table 1: Traded volumes of CfDs. The pay-off from the CfD is defined as in equation (1). CfD = average during settlement period of (daily Area Price daily System Price) (1) From this formula we see that every time the Area Price is higher than the System Price the holder gets a rebate equal to difference between the two prices. Otherwise the trader has to pay the difference in prices. The price of these contracts will be settled by the supply/demand drives. There are many possible hedging and trading strategies that can be used for CfDs. A CfD used in combination with a forward contract referred to the System Price will give the following payoff: CfD + forward contract = average of the daily Area Price during settlement period p CfD - p forward This combination should be used when a trader wants to hedge the forward Area Price in the actual season. The total effect of these two contracts is to create a forward Area Price contract, which is currently not available at Nord Pool. The CfD makes it possible to hedge the price differential Area Price System Price. If we want to hedge directly against the Area Price differences between two areas A and B, we buy one CfD for area B and sell one for area A and get the following pay-off: CfD B CfD A = average of (daily Area Price B daily Area Price A) during settlement period (p BCfD - p ACfD ) If there were forward markets for each area an arbitrage strategy would be to buy and sell locational CfDs and trade these against the respective forward Area Prices. An introduction of TCCs at Nord Pool together with CfDs would make it possible to exploit profits if the market was not well-functioning. 2 PRICING ANALYSIS In this paper we analyse the prices of the contracts with the technical information in Table 2. All contracts were traded in the periods and The main difference between these contracts is that they are referred to two different forward seasons Winter 1 and Summer 21, and that their trading periods differ. The market value of a CfD during the trading period reflects the market s prediction of the price difference, Area Price System Price, during the delivery (settlement) period. Based on the available information from the Nord Pool we did an analysis of how the CfDs were priced in the respective periods. The results are given in Table 3 and Table 4. As Table 3 and Table 4 show, the average trading prices are highest for the contracts referred to Århus and Helsinki for the CfD Winter 1 21 and Summer The price for the contract referred to Oslo is negative, which means that the traders expect the Area Price to be less than the System Price.

3 Product- Series Århus (DK 1) Winter 1, 21 Helsinki (FI) Winter 1, 21 Oslo (NO) Winter 1, 21 Stockholm (SE) Winter 1, 21 Århus (DK1) Helsinki (FI) Oslo (NO) Stockholm (SE) (DK2) 1 Contract Trading hours Period Settlement period Table 2: Product specification for the CfDs at base load. Table 3 shows the average price and standard deviation for the contracts traded in the period The standard deviation for both of these contracts is in the same order of magnitude as the price of ordinary forward contracts. Closer analysis of the prices (Figure 1) showed that the prices of these contracts decreased in the beginning of this year and stabilised at a new level. This is the main cause for the high relative standard deviation. If we divided the period into two, at the point where the change took place the standard deviation amounted to about 1 per cent in the first period and to 5-1 per cent in the second period. Reference place (2879h) price (Euro/MWh) St. dev. Århus (DK1) Helsinki (FI) Oslo (NO1) Stockholm (SE) Table 3: Data concerning the prices of CfD Winter 1 The standard deviation of the price differential Area Price minus System Price is also relatively high. In all cases and for both the trading and the settlement period the standard deviation is several magnitudes of order higher than the price differential. This means that the prices are very volatile. Since the underlying is expected to be uncertain, the traded forward contract may have this uncertainty incorporated. In addition some of the parameters used for calculating the security require- ments were changed in January 21 because they generally were too high. This affected the contracts referred to Helsinki, Stockholm, Oslo and Århus. Reference place (3672h) price (Euro/MWh) St. dev. Århus (DK1) (DK2) Helsinki (FI) Oslo (NO1) Stockholm (SE) Table 4: Data concerning the prices of CfD Summer The prices of the CfD Winter 1 21 contracts are not plotted here, but they were slightly increasing during most of the trading period, contrary to the Summer 21 contracts which were decreasing during most of the trading period. This may be understood from the fact that the trading period for CfD Winter 1 21 contracts is the first part of the trading period for the Summer 21 contracts, and the prices for these contracts were relatively stable during this period. The contracts also refer to different seasons and there may be some seasonal trend which is incorporated in the prices. Euro/MWh 6, 5, 4, 3, 2, 1,, -1, 1 Prices of CfD Summer DK1 DK2 FI NO1 SE Figure 1: Prices of the CfD Summer 21 contracts traded in the period We also calculated the average of the underlying asset in the same periods, Table 5 and Table 6, and plotted it in Figure 2. The trading price should reflect the market s prediction of the price difference as defined in equation in (1) during the delivery period. From Table 5 we see that the prices are varying from the underlying asset. The result is not very surprising because the unforeseen shocks during the settlement period (e.g. unexpected constraints due to plant, and line outages as well as relative demand in each region) are bound to occur. 1 The CfD referred to Eastern Denmark was introduced the 23-th of March of

4 Reference place (3672h) price (Euro /MWh) Area price System Price Århus (DK1) Helsinki (FI) Oslo (NO1) Stockholm (SE) Table 5: Data concerning the prices of CfD Winter 1 The difference between the Area Price and the System Price is negative in all cases and it is only the Oslo contract which has a coinciding sign with the average price. The contracts referred to Helsinki, Århus and Stockholm are over-priced relative to the average value of the underlying asset. The Oslo contract is on average under-priced and it is possible to earn.57 Euro/MWh ( ) by buying it. Reference place (3672h) price (Euro /MWh) Area price System Price Århus (DK1) (DK2) Helsinki (FI) Oslo (NO1) Stockholm (SE) Table 6: Data concerning the prices of CfD Summer For the contracts traded in the period (Table 6) the results are like the preceding case. First of all we see that the sign of the average prices are not coinciding, except for the contract referred to Århus. Secondly, for both types of contracts the magnitudes of the prices are quite deviating from the underlying asset. For instance the contract referred to Stockholm (Winter 1 21) has an average traded price of 1.44 Euro/MWh and an average value of the Area Price minus the System Price of -.24 Euro/MWh. This is a price about forty-seven times higher than the underlying asset. All the Summer 21 contracts are on average over-priced, except for the Oslo contract, which it is possible to earn.292 Euro/MWh ( ) by buying it. One reason for the deviating results can be that the prices of these contracts were generally set to high in winter 2 season (January to April) due to some of the parameters used to calculate the security limits. The prices were changed in January It is hard to see any systematic relationship between the trading prices and the average of difference between the Area Price and the System Price. Figure 2: The difference between the Area Prices and System Price (1 Euro ~ 7.71 NOK) in the period We see that Århus, Helsinki and Stockholm show the greatest spread between the Area Price and the System Price. The System Price is on average higher than the Area Prices for the Winter 1 21 contracts. The willingness to pay is highest for Århus and Helsinki in both cases. The next step would be to analyze the forward prices the respective areas for the seasons Winter 1 21 and Summer These prices are not currently available from Nord Pool, but we would expect better accordance between these prices and the CfD prices. In Figure 2 we have plotted the difference between the respective Area Prices and the System Price. We see that there are several peaks showing up, which implies periods with congestion or outages. 3 PRICE DISTRIBUTION AND CORRELATIONS To study the distribution of the prices we made two histograms, one for each series. In Figure 3 we have plotted the distribution for the prices of the CfD Winter 1 The distributions show concentrated peaks and seem to have some degree of normal distribution. We also plotted the distribution of the contract price for the CfD Summer 21 in Figure 4 and 5. frequency Difference between area prices and systemprice , -1, , 17, 26, 35, Århus Helsinki Oslo SE NO1 FI DK2 DK1 Stockholm Figure 3: The distribution of the prices of CfD Winter 1 21 (1 Euro ~ 7.71 NOK).

5 frequency Figure 4: The distribution of the prices of CfD Summer 1 21 for Helsinki and Århus. The Helsinki and Århus contracts (Figure 4) have several peaks and have a non-normal distribution. For the, Stockholm and Oslo contracts (Figure 5) the distribution is more concentrated but still with several peaks. frequency , -5,, 5, 1, 15, 2, 25, 3, 35, 4, -1, -7, -4, -1, 2, 5, 8, Helsinki Århus Stockholm Oslo Figure 5: The distribution of the prices of CfD Summer 21 for Oslo, Stockholm and. We calculated the correlation between the series of prices for the CfD Winter 1 21 and CfD Summer The objective was to study if there was any relationship between the prices. As Table 7 shows there is low correlation between the prices for the Winter 1 21 contracts referred to the different areas. Corr DK1 FI NO1 SE DK FI NO SE Table 7: The correlation between the CfD Winter 1 Table 8 shows that there is high correlation between the prices of CfD Summer 21 DK1, FI and SE contracts. It is also high negative correlation between DK1, NO1 and SE. The reason why the Winter 1 21 and Summer 21 contracts have different correlation is not quite clear. Corr DK1 DK2 FI NO1 SE DK1 1 NA DK2 NA FI NO SE Table 8: The correlation between the CfD Summer 4 PRICING ISSUES Transmission congestion contracts (TCCs) are extensions of CfDs where it is possible to hedge directly against the differences in the Area Prices [1]. The pricing of the TCCs should correspond to the expected value of the congestion fee over the specified settlement period. The System Operator (who gets the congestion rent) 2 can provide these contracts for the lowest price because his risk is negatively correlated with the congestion rent [2]. If a third party provides these contracts and charges the same price as the System Operator he will on average get zero income. This is the case if both face the same set of scenarios (i.e. the same number of hours with congestion and Area Prices in a period). It is possible to reproduce the cash flow from a TCC by buying one CfD for the buyer s area and sell one CfD for the generator s area. As mentioned earlier a first approximation for the pricing of the CfDs would be to use the forward prices for the respective areas and the seasonally System Price forward curve. In addition to the prices charged for the contracts Nord Pool has established trading fees, which are related to the customer s buying pattern. 5 CONCLUSIONS This paper has showed how the CfDs for the two first trading periods have been priced at Nord Pool. Based on a comparison between the prices of the contract and the average of the Area Price minus the System Price, they seem to be over-priced on average. Most of the prices also show a non-normal distribution with several peaks. There is also high correlation between some of the prices for the CfD Summer 21 contracts. The next step would be to analyze the prices for rest of the available contracts with more advanced statistical methods. We also would like to have the forward Area Prices to compute the average of the Area Prices minus the System Price in the settlement period. Because there is a general lack of data, a rigorous analysis of the prices is 2 To what extent the System Operator keeps the congestion rent depends on the economic regulation of the grid company. With current regulation in Norway (revenue cap regulation) the congestion rent adds nothing to the net revenue of the System Operator

6 insufficient to judge whether the market for CfDs is working properly. The contracts not have been traded since the 17-th November 2 and every contract is referred to a season or a year. This involves that it will take some time to get enough data to use methods from finance to evaluate if the prices of the contracts are consistent under or over the value of the underlying asset. In conclusion the analysis indicates that Nord Pools CfD market still is immature. REFERENCES [1] W. W. Hogan, Flowgate Rights and Wrongs, Center for Business and Government, John F. Kennedy School of Government, Harvard University, August 2, 2. [2] T. Kristiansen, Cash Flow Analysis of Contract for Differences, RIMAPS 21 - EuroConference on Risk Management in Power System Planning and Operation in a Market Environment, Porto, Portugal, September [3] homepage for Nord Pool the Nordic Electricity Exchange.