The Energy Markets Use and interpretation of medium to extended range products ECMWF, Reading, 14 th of November 2005 drs. Stefan Meulemans, MSc Sempra Energy Europe Ltd. London smeulemans@sempracommodities.com
Major global player in metals, polymers, gas, oil and electricity markets Offices in New York, London, Houston, San Diego, Singapore, Geneva, Calgary and Toronto
Sempra Energy Europe Ltd. Speculative trading on gas, oil and electricity markets Using information to anticipate market moves Weather major fundamental
Energy markets Assessing the supply and demand of the main energy sources Determining a price based on all available fundamentals Often volatile markets compared with traditional markets
Futures and forwards Day ahead (EC) Week ahead (EC) Month Ahead (EC) Cal ahead [Professional] QENOYc1, Last Trade, Candle 06/06/2005 31.90 32.25 31.80 32.23 Daily QENOYc1 [Candle] 08/03/2005-07/06/2005 (GMT) Price EUR 33 32.7 32.4 32.1 31.8 31.5 31.2 30.9 30.6 30.3 30 14 21 31 07 14 21 28 06 13 24 31 07 Mar 05 Apr 05 May 05 29.7 29.4 29.1 28.8 28.5 28.2 27.9 27.6 27.3 27
Supply and demand Supply and demand determine the forward electricity prices Traded contracts are day-ahead, weekahead, month-ahead etc. Supply: wind and hydro Demand: temperature important
Nordic markets Weather crucial due to very high hydro capacity Market very dependent on ECMWF model
Hydro ECMWF forecasts have to be transposed into a precipitation energy forecast We also need an estimate of the snow cover and the expected melt, based on the temperature Alps, Pyrenees and Scandinavia
Hydro Reservoirs Scandinavia
Example Q106 Nord Pool [Professional] Daily QENOQc1 [Candle, MA 14, WMA 14, EMA 14, RSI 14, RSIMA 14,14] 13/06/2005-14/11/2005 (GMT) Q ENO Q c1, Last Trade, C andle 10/11/2005 39.00 39.15 38.70 39.10 Q ENO Q c1, C lose(last Trade), MA 14 10/11/2005 38.84 Q ENO Q c1, C lose(last Trade), W MA 14 10/11/2005 38.65 Q ENO Q c1, C lose(last Trade), EMA 14 10/11/2005 38.91 Price EUR 45 44 43 42 41 40 39 38 17 27 04 11 18 25 01 08 15 22 29 05 12 19 26 03 10 17 24 31 07 14 Jun 05 Jul 05 Aug 05 Sep 05 Oct 05 Nov 05 37
Dry weather in Nordic Cal 03 in 2002 400.00 350.00 300.00 250.00 200.00 150.00 100.00 50.00 0.00 02/01/2002 02/02/2002 02/03/2002 02/04/2002 02/05/2002 02/06/2002 02/07/2002 02/08/2002 02/09/2002 02/10/2002 02/11/2002 02/12/2002
290.00 270.00 250.00 230.00 210.00 190.00 170.00 150.00 Wet weather in Nordic Cal 05 in 2004 02/01/2004 02/02/2004 02/03/2004 02/04/2004 02/05/2004 02/06/2004 02/07/2004 02/08/2004 02/09/2004 02/10/2004 02/11/2004 02/12/2004
ECMWF Nord Pool Data is directly given in GWh. One run moves the market. Ensemble very important. Countries around Pyrenees and Alps only influenced by major hydro events, as wind and non-renewable power production is more widely used
Example Point Carbon Precipitation energy and date Histogram precipitation energy total
NAO outlook Development of Nordic NAO ECMWF data day 10 or beyond to calculate the NAO 5 4 3 2 1 0-1 -2 Nordic NAO day 10 based on EC ensemble (12Z and 00Z) Current NAO value known, so not of interest 19/10/2005 22/10/2005 25/10/2005 28/10/2005 31/10/2005 03/11/2005 06/11/2005 09/11/2005
ECMWF cluster outlook
Wind Short range deterministic storms! Mid range ensemble ECMWF Installed wind capacity
Wind production forecasts Short term models for day-ahead wind production. Error should be low. First days deterministic ECMWF model Further ahead ensemble ECMWF Probabilistic wind forecast? Germany, Denmark and Spain Unit MWh/h
Example Point Carbon Wind production in MWh and date
Temperature Very well correlated to heating and cooling demand Water temperature forecasts with ECMWF data Development of demand outlooks Point Carbon example
Cold spell and NBP gas Daily QGA2NB [Line, EMA 14, MA 14, WMA 14, RSI 14, RSIMA 14,14] [Professional] 21/06/2004-21/11/2005 (GMT) Q GA 2NB, C lose(last Trade), Line 10/11/2005 46.00 Q GA 2NB, C lose(last Trade), EMA 14 10/11/2005 39.97 Q GA 2NB, C lose(last Trade), MA 14 10/11/2005 37.67 Q GA 2NB, C lose(last Trade), W MA 14 10/11/2005 40.66 Jul Sep Nov Jan Mar May Jul Sep Nov 2004 2005 Price GBp 110 105 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20
Day ahead German electricity [Professional] Price EUR 100 QDE-B-1D=RWEE, Close(Last Quote), Line 06/06/2005 42.00 Daily QDE-B-1D=RWEE [Line] 21/12/2004-08/06/2005 (GMT) 80 60 40 20 0 03 17 31 14 28 14 28 11 25 09 23 06 Dec 04 Jan 05 Feb 05 Mar 05 Apr 05 May 05
Demand models Main variable temperature Demand industry and basic load Day of the week and holidays In case of electricity hourly load calculation
W W W DN W W DN W -2-2 -2-2 -2 n 4Ju n 3Ju n 2Ju n 1Ju n 0Ju n 9Ju n 18 0 19 8. 1 1. 5 8 19 7. 4 8 4 0 7. 6 7. 5 20 20 0 21 4 6. 3 0 22 3. 5 9 0. 59 9. 8 22 7. 3 0 1. 0 22 0. 03 22 22 6. 6 21 21 220 W DN W W DN W W DN W -1 8Ju 7Ju n 6Ju n n 190 W E -1-1 -1 5Ju 200 W E BO BO -1 n n 210 DA -J u -J u n n 5 23 4. 80 240 14 13 -J u -J u 230 12 11 Sempra s gas model 250 60 40 20 0-20 180 170-40
Oil markets ECMWF tropical model. New hurricane track forecast immediately moves the market. Weekly inventories oil and natural gas make the market move. Weather data important.
Katrina and the Gulf Depending on track, certain oil product are influenced Probability outlook very important
Brent oil price since April QLCOZ5, Last Trade, Bar 10/11/2005 5,651 5,675 5,566 5,583 QLCOZ5, Close(Last Trade), EMA 14 10/11/2005 5,832 QLCOZ5, Close(Last Trade), MA 14 10/11/2005 5,851 QLCOZ5, Close(Last Trade), WMA 14 10/11/2005 5,813 21 28 06 13 20 27 06 13 20 27 04 11 18 25 01 08 15 22 30 06 13 20 27 04 11 18 25 01 08 15 Apr 05 May 05 Jun 05 Jul 05 Aug 05 Sep 05 Oct 05 Nov 05 Price USc Bbl 6800 6750 6700 6650 6600 6550 6500 6450 6400 6350 6300 6250 6200 6150 6100 6050 6000 5950 5900 5850 5800 5750 5700 5650 5600 5550 5500 5450 5400 5350 5300 5250 5200 5150 5100 5050 5000 4950 4900
European gasoline contract QPU-1M-ARA, Last Quote, Bar 08/11/2005 518.25 523.50 518.25 523.50 QPU-1M-ARA, Close(Last Quote), EMA 14 08/11/2005 547.06 QPU-1M-ARA, Close(Last Quote), MA 14 08/11/2005 541.98 QPU-1M-ARA, Close(Last Quote), WMA 14 08/11/2005 538.22 Price USD T 810 800 790 780 770 760 750 740 730 720 710 700 690 680 670 660 650 640 630 620 610 600 590 580 570 560 550 540 530 520 510 17 21 25 29 03 07 11 15 19 23 27 31 04 08 12 16 20 24 28 01 05 09 13 17 21 25 29 03 07 11 15 19 23 27 31 04 08 Jun 05 Jul 05 Aug 05 Sep 05 Oct 05 Nov 05 500
Emissions market Emission markets originated from Kyoto agreement Each country has a maximum allowance to emit CO2 If country emits more than allowed, it has to buy allowances, and vice versa
Emissions market Weather also important to anticipate emissions output ECMWF data can be translated in expected C02 output from power stations
Emissions market [Professional] QCFIZ5, Close(Last Trade), Line 10/11/2005 2,265 QCFIZ5, Close(Last Trade), EMA 14 10/11/2005 2,202 QCFIZ5, Close(Last Trade), MA 14 10/11/2005 2,178 QCFIZ5, Close(Last Trade), WMA 14 10/11/2005 2,189 Daily QCFIZ5 [Line, EMA 14, MA 14, WMA 14, RSI 14, RSIMA 14,14] 27/04/2005-14/11/2005 (GMT) Price EUc 04 11 18 25 02 09 16 23 30 07 14 21 28 04 11 18 25 02 09 16 23 30 07 14 21 28 04 11 May 05 Jun 05 Jul 05 Aug 05 Sep 05 Oct 05 Nov 05 3100 3050 3000 2950 2900 2850 2800 2750 2700 2650 2600 2550 2500 2450 2400 2350 2300 2250 2200 2150 2100 2050 2000 1950 1900 1850 1800 1750 1700 1650 1600 1550 1500 1450 1400 1350 1300
Weather derivatives Directly linked to weather outlook. Trading HDD, CAT and precipitation indices. Weather markets about to develop significantly Weather derivatives also important to hedge other energy contracts
Converting data in HDD Sempra directly converts the model data in different possible scenario s for HDD and CAT. Mid- and long-range forecasts all needed
Risk management Use of long-term probabilistic outlooks Assess risk of certain energy scenario s Try to have view on weeks- and monthsahead
Probabilistic long-range outlook Reduced risk on mid-range positions Now particularly for temperature and precipitation, but wind would be interesting too When signal is strong, it makes sense to take significant long or short positions
Summary Mid- to long-range model data very relevant to different energy commodities Importance of also considering the energy industry when models would be enhanced. E.g. probabilistic data Development of 30-days ensembles would interesting