Energy Pricing Techniques in the Electricity Market

1
Energy Pricing Techniques in the Electricity
Market

• Part 4
• Application of Weather Derivatives
• Dr Harvey Stern,
• Bureau of Meteorology, Australia

2
Outline of Presentation

• Background.
• An Historical Note.
• Weather-related Risk.
• The Growing Interest.
• Asia-Pacific Region.
• Some Statistics.
• Weather Derivatives Explained.
• Examples of Applications.
• Concluding Remarks.

3
Background

• Weather risk is one of the biggest uncertainties
  facing business.
• We get droughts, floods, fire, cyclones
  (hurricanes), snow ice.
• Nevertheless, economic adversity is not
  restricted to disaster conditions.
• A mild winter ruins a skiing season, dry weather
  reduces crop yields, rain shuts-down
  entertainment construction.

4
Background (cont.)

• The increasing interest may be explained in terms
  of
• A desire to meet client needs.
• A need to reduce the cost of capital.
• Cross-fertilization between various fields.
• Entry of new participants.
• Growing responsibilities of Company Directors.
• Source Prof. John Hewsons presentation to the
  Weather Risk Management Association

5
An Historical Note An Early Example

• In 1992, the present author explored a
  methodology to assess the risk of climate
  change.
• Option pricing theory was used to value
  instruments that might apply to temperature
  fluctuations and long-term trends.
• The methodology provided a tool to cost the risk
  faced (both risk on a global scale, and risk on a
  company specific scale).
• Such securities could be used to help firms hedge
  against risk related to climate change.

6
Foundation of the Weather Market

• The foundation of todays financial weather
  contracts is in the US power market
• For the weather-sensitive end-user, not to hedge
  is to gamble on the weather.
• Robert S. Dischell

7
Weather-related Industry Risk

• "Shares in Harvey Norman fell almost 4 per
  cent yesterday as a cool summer and a warm start
  to winter cut into sales growth at the furniture
  and electrical retailer's outlets Investors
  were expecting better and marked the shares down
  3.8 per cent to a low of 3.55
• Sales at Harvey Norman were hit on two
  fronts. Firstly, air conditioning sales were
  weak because of the cool summer, and a warmer
  than usual start to winter had dampened demand
  for heating appliances.
• Source The Australian of 18 April, 2002

8
Weather-related Agricultural Risk

• The Australian sugar industry is facing its
  fifth difficult year in a row with a drought
  dashing hopes of an improved crop in Queensland,
  where 95 of Australia's sugar is grown...
• Whilst dry weather during the May-December
  harvest period is ideal for cane, wet weather
  during this time causes the mature cane to
  produce more shoots and leaves, reducing its
  overall sugar content.
• (Australian Financial Review of 8 May, 2002)

9
Channels for Weather Risk Transfer

• ART (Alternative Risk Transfer) is a generic
  phrase used to denote various non-traditional
  forms of re/insurance and techniques where risk
  is transferred to the capital markets.
• Source http//www.artemis.bm

10
The Growing Interest.

• 3,937 contracts transacted in last 12 months (up
  43 compared to previous year).
• Notional value of over 4.3 billion dollars (up
  72).
• Market dominated by US (2,712 contracts), but
  growth in the past year is especially so in
  Europe and Asia.
• Australian market accounts for 15 contracts worth
  over 25 million (6 contracts worth over 2
  million, previously).
• Source Weather Risk Management Association
  Annual Survey (2002)

11
The Diversification.

• Another significant development is the
  diversification of the types of contracts that
  were transacted.
• Temperature-related protection (for heat and
  cold) continues to be the most prevalent, making
  up over 82 percent of all contracts (92 last
  year)
• Rain-related contracts account for 6.9 (1.6
  last year), snow for 2.2 (0.6 last year) and
  wind for 0.4 (0.3 last year).
• Source Weather Risk Management Association
  Annual Survey (2002)

12
Why has ART grown?

• Risk management moving up the agenda
• A need to manage uninsurable liabilities
• A need to protect against irregular income
  spikes
• Source Modern ART practice (Gerling Global
  Financial Products)

13
What is the future of ART?

• The term Alternative Risk Transfer (ART) will
  soon be a misnomer. ART is fast becoming an
  essential risk management tool for primary
  insurers, reinsurers and non-insurance
  corporations.
• Source Modern ART practice (Gerling Global
  Financial Products)

14
The Asia-Pacific Region

• Interest in weather risk management has grown
  in the Asia-Pacific Region (covering electricity,
  gas, agriculture). Countries involved include
• Japan
• Korea and,
• Australia/New Zealand.
• Source Weather Risk Management Association.

15
Weather-linked Securities

• Weather-linked securities have prices which are
  linked to the historical weather in a region.
• They provide returns related to weather observed
  in the region subsequent to their purchase.
• They therefore may be used to help firms hedge
  against weather related risk.
• They also may be used to help speculators
  monetise their view of likely weather patterns.

16
Securitisation

• The reinsurance industry experienced several
  catastrophic events during the late 1980s early
  1990s.
• The ensuing industry restructuring saw the
  creation of new risk-management tools.
• These tools included securitisation of insurance
  risks (including weather-related risks).
• Weather securitisation may be defined as the
  conversion of the abstract concept of weather
  risk into packages of securities.
• These may be sold as income-yielding structured
  products.

17
Catastrophe Bonds

• A catastrophe (cat) bond is an exchange of
  principal for periodic coupon payments wherein
  the payment of the coupon and/or the return of
  the principal of the bond is linked to the
  occurrence of a specified catastrophic event.
• The coupon is given to the investor upfront, who
  posts the notional amount of the bond in an
  account.
• If there is an event, investors may lose a
  portion of (or their entire) principal.
• If there is no event, investors preserve their
  principal and earn the coupon.
• Source Canter Cole at http//www.cnare.com

18
Catastrophe Swaps

• A catastrophe (cat) swap is an alternative
  structure, but returns are still linked to the
  occurrence of an event.
• However, with swaps, there is no exchange of
  principal.
• The coupon is still given to the investor
  upfront, but the structure enables investors to
  invest the notional amount of the bond in a
  manner of his own choosing.
• Source Canter Cole at http//www.cnare.com

19
Weather Derivatives Explained Clewlow et al.
(2000) describe a derivative as "a financial
product that derives its value from other more
basic variables". These products include futures,
forwards, call options, put options, and swaps.
They describe weather derivatives as being
similar "to conventional financial derivatives,
the basic difference coming from the underlying
variables that determine the payoffs", such as
temperature, precipitation, wind, Heating Degree
Days (HDDs), and Cooling Degree Days (CDDs).
20

• Pricing Derivatives
• There are three approaches that may be applied to
  the pricing of derivatives.
• These are
• Historical simulation (applying "burn analysis")
  
• Direct modelling of the underlying variables
  distribution (assuming, for example, that the
  variable's distribution is normal) and,
• Indirect modelling of the underlying variables
  distribution (via a Monte Carlo technique).
• Direct modelling is chosen for the current
  exercise, the distribution of forecast errors
  being assumed to be normal.

21
Returning to the Cane Grower

• Suppose that our cane grower has experienced an
  extended period of drought.
• Suppose that if rain doesn't fall next month, a
  substantial financial loss will be suffered.
• How might our cane grower protect against
  exceptionally dry weather during the coming month?

22
One Approach

• One approach could be to purchase a Monthly
  Rainfall Decile 4 Put Option.
• Assume that our cane grower decides only to take
  this action when there is already a risk of a dry
  month.
• That is, when the current month's Southern
  Oscillation Index (SOI) is substantially
  negative.
• So, the example is applied only to the cases when
  the current month's Southern Oscillation Index
  (SOI) is in the lowest 5 of possible values,
  that is, below -16.4.

23
Specifying the Decile 4 Put Option

• Strike Decile 4.
• Notional 100 per Decile (lt Decile 4).
• If, at expiry, the Decile is lt Decile 4, the
  seller of the option pays the buyer 100 for each
  Decile lt Decile 4.

24
Pricing Methodologies

• Historical simulation.
• Direct modeling of the underlying variables
  distribution.
• Indirect modeling of the underlying variables
  distribution (via a Monte Carlo technique).

25
Payoff Chart for Decile 4 Put Option
26
Outcomes for Decile 4 Put Option
27
Evaluating the Decile 4 Put Option

• 14.2 cases of Decile 1 yields (.142)x(4-1)x100
  42.60
• 13.2 cases of Decile 2 yields (.132)x(4-2)x100
  26.40
• 8.4 cases of Decile 3 yields (.084)x(4-3)x1008
  .40
• The other 25 cases (Decile 4 or above) yield
  nothing.
• leading to a total of 77.40, which is the price
  of our put option.

28
Should Companies Worry?

• In the good years, companies make big profits.
• In the bad years, companies make losses.
• - Doesnt it all balance out?
• - No. it doesnt.
• Companies whose earnings fluctuate wildly receive
  unsympathetic hearings from banks and potential
  investors.

29
Another Example

• Another example of a weather linked option is the
  Cooling Degree Day (CDD) Call Option.
• Total CDDs is defined as the accumulated number
  of degrees the daily mean temperature is above a
  base figure.
• This is a measure of the requirement for cooling.
• If accumulated CDDs exceed the strike, the
  seller pays the buyer a certain amount for each
  CDD above the strike.

30
Pay-off Chart for a CDDCall Option
31
Cooling Degree Days (1855-2000)(and climate
change)

• The chart shows frequency distribution of annual
  Cooling Degree Days at Melbourne using all data

32
Cooling Degree Days (1971-2000)

• The chart shows frequency distribution of annual
  Cooling Degree Days at Melbourne using only
  recent data

33
Weather Climate Forecasts

• Daily weather forecasts may be used to manage
  short-term risk (e.g. pouring concrete).
• Seasonal climate forecasts may be used to manage
  risk associated with long-term activities (e.g.
  sowing crops).
• Forecasts are based on a combination of solutions
  to the equations of physics, and some
  statistical techniques.
• With the focus upon managing risk, the forecasts
  are increasingly being couched in probabilistic
  terms.

34
An Illustration of theImpact of Forecasts

• When very high temperatures are forecast, there
  may be a rise in electricity prices.
• The electricity retailer then needs to purchase
  electricity (albeit at a high price).
• This is because, if the forecast proves to be
  correct, prices may spike to extremely high
  (almost unaffordable) levels.

35
Impact of Forecast Accuracy

• If the forecast proves to be an over-estimate,
  however, prices will fall back.
• For this reason, it is important to take into
  account forecast accuracy data in determining the
  risk.

36
Forecast Accuracy Data The Australian Bureau of
Meteorology's Melbourne office possesses data
about the accuracy of its temperature forecasts
stretching back over 40 years. Customers
receiving weather forecasts have, recently,
become increasingly interested in the quality of
the service provided. This reflects an overall
trend in business towards implementing risk
management strategies. These strategies include
managing weather related risk. Indeed, the US
Company Aquila developed a web site that presents
several illustrations of the concept http//www.g
uaranteedweather.com
37
Using Forecast Accuracy Data

• Suppose we define a 38 deg C call option
  (assuming a temperature of at least 38 deg C has
  been forecast).
• Location Melbourne.
• Strike 38 deg C.
• Notional 100 per deg C (above 38 deg C).
• If, at expiry (tomorrow), the maximum temperature
  is greater than 38 deg C, the seller of the
  option pays the buyer 100 for each 1 deg C above
  38 deg C.

38
Pay-off Chart 38 deg C Call Option
39
Determining the Price of the38 deg C Call Option

• Between 1960 and 2000, there were 114 forecasts
  of at least 38 deg C.
• The historical distribution of the outcomes are
  examined.

40
Historical Distribution of Outcomes
41
Evaluating the 38 deg C Call Option (Part 1)

• 1 case of 44 deg C yields (44-38)x1x100600
• 2 cases of 43 deg C yields (43-38)x2x1001000
• 6 cases of 42 deg C yields (42-38)x6x1002400
• 13 cases of 41 deg C yields (41-38)x13x1003900
• 15 cases of 40 deg C yields (40-38)x15x1003000
• 16 cases of 39 deg C yields (39-38)x16x1001600
• cont.

42
Evaluating the 38 deg C Call Option (Part 2)

• The other 61 cases, associated with a temperature
  of 38 deg C or below, yield nothing.
• So, the total is 12500.
• This represents an average contribution of 110
  per case, which is the price of our option.

43
A Financial Guarantee The guarantee described is
that the forecast will be in error by no more
than 3C. The terms of the guarantee are that the
seller of the guarantee will pay the buyer
100.00 for each 0.1C greater than 3C that the
forecast is in error. It is the purpose of the
paper to develop an approach to pricing such a
financial guarantee, and to provide it as a
technique that is available on the web.
(after Stern Dawkins, 2003)
44
The Instrument The instrument is made up of a
combination of a call option and a put option
about the next day's maximum temperature at
Melbourne, the "strikes" being set respectively
3C above and below the forecast temperature.
The taker of this option combination receives
100 for each 0.1C that the observed temperature
is above or below the respective strikes. (after
Stern Dawkins, 2003)
45
Forecast Errors Dawkins and Stern (2003) show
that the magnitude of the forecast errors is
largely a function of season and synoptic
pattern. Dahni (2003) describes an automated
technique for "typing" synoptic patterns.
(after Stern Dawkins, 2003)
46
Forecast Errors as a Function of Season (after
Dawkins Stern, 2003)
47
Forecast Errors as a Function of Synoptic
Pattern (after Dawkins Stern, 2003)
48
The Approach Used

• The approach used is as follows
• The forecast verification data is stratified
  according to month, and also according to the
  nature of the prevailing atmospheric circulation
  - cyclonicity, direction and strength of the
  surface flow.
• The distribution of the magnitude of forecast
  errors for each month (and also for each synoptic
  pattern type) is noted this distribution is
  adjusted in order to take into account a
  long-term downward trend in the magnitude of the
  errors
• The distribution of forecast errors is assumed to
  be normal for each data subset, and a "fair
  value" price for the option combination for each
  month and each circulation type is then obtained.
  
• (after Stern Dawkins, 2003)

49

• Example
• The example we shall use to illustrate the
  methodology is a forecast produced during the
  month of January, associated with a synoptic type
  flow possessing the following characteristics
• weak strength
• cyclonic (curvature)
• from the north-north-west.
• Over the 40-year period (1961-2000), occurrences
  of such a flow across SE Australia (over all
  months of the year) have been temperature
  forecasts with an RMS error of 2.70C.
• (after Stern Dawkins, 2003)

50
Example (cont.) More recently (1991-2000), such a
flow has been accompanied by an RMS error of (a
much reduced) 2.26C. It is then assumed that the
forecast performance during the period 1991-2000
better represents what one might anticipate to be
the current level of performance, than does the
forecast performance over the 1961-2000
period. It is also assumed that the proportional
improvement in forecasting for each individual
month (January, February, March etc.) is the
same, that is, a proportional decrease in RMS
error of (2.26/2.70)(0.84) in the current
case. (after Stern Dawkins, 2003)
51
Example (cont.) The monthly RMS error calculated
over the 1961-2000 period for the current
synoptic type and the current month (3.32C in
this case) is then multiplied by the ratio (0.84)
in order to achieve an estimate of the likely RMS
error for the current forecast. So, the case of a
January cyclonic weak north-north-west synoptic
flow yields (0.84x3.32)2.79C for our estimated
RMS error. It is then assumed that the errors are
normally distributed and, utilising areas under
the standard normal curve, one calculates the
expected return on the guarantee to be 410. This
procedure is then repeated for all months and for
all synoptic patterns. (after Stern Dawkins,
2003)
52

• The WEB Site
• A web site is developed in order that
• potential "customers" may readily obtain a price
  for the instrument and,
• researchers may test its output.
• This may be viewed and tested at
• http//www.weather-climate.com/guarantee.html
• (after Stern Dawkins, 2003)

53
A View of the WEB Site
54
Testing the Instruments Validity
It was considered that if, over a large number of
cases, writers of the option combination do not
make either a significant profit or a significant
loss, the validity of the "fair value" price
would be demonstrated. The instrument's validity
was then tested by calculating the "fair value"
price on independent cases taken for the entire
year of 2001. However, from an analysis of all
of the year-2001 cases, it was determined that
writers of the option combination would have
received 75,574 over the year, while paying out
only 23,800. (after Stern Dawkins, 2003)
55
Testing (cont.)
Nevertheless, this substantial profit (over 200
return) is not necessarily suggesting a possible
flaw in the valuation technique. On the
contrary, it may be explained in terms of the
spectacular improvement in the accuracy of
forecasts achieved during 2001 (see next
slide). One may show that had the forecasts been
of similar skill to those of previous years, the
payout would have been much closer to the monies
received. The profit achieved by the option
writers can, therefore, be explained in terms of
that increased skill. (after Stern Dawkins,
2003)
56
Sharp Improvement in Forecast Accuracy in 2001
(after Dawkins Stern, 2003)
57

• Comments on the Financial Guarantee
• A methodology to price a financial guarantee
  about the accuracy of a forecast has been
  described and demonstrated with "real" data.
• It has been shown that had such a guarantee been
  applied to day-1 maximum temperature forecasts
  issued during 2001 for Melbourne, providers of
  the guarantee would have made a substantial
  profit
• on account of the increased skill displayed by
  the forecasts.
• (after Stern Dawkins, 2003)

58
Ensemble Forecasting(another approach to
measuring forecast uncertainty)

• Another approach to obtaining a measure of
  forecast uncertainty, is to use ensemble weather
  forecasts.
• The past decade has seen the implementation of
  these operational ensemble weather forecasts.
• Ensemble weather forecasts are derived by
  imposing a range of perturbations on the initial
  analysis.
• Uncertainty associated with the forecasts may be
  derived by analysing the probability
  distributions of the outcomes.

59
Some Important Issues

• Quality of weather and climate data.
• Changes in the characteristics of observation
  sites.
• Security of data collection processes.
• Privatisation of weather forecasting services.
• Value of data.
• Climate change.

60
Concluding Remarks

• The sophistication of weather-related risk
  management products is growing.
• In evaluating weather securities one needs to use
  historical weather data and forecast accuracy
  data, and also to take into account climate
  trends.
• Ensemble forecasting is a new approach to
  determining forecast uncertainty.