Option Markets and Real options

Long Han
Constellation Energy
Agenda

• Why do options markets exist?

• Basic structures: Options
• natural options positions
• Basic valuation and risk
• The role of models
• Limitations of option valuation methods

2
Why do options markets exist?

• As with futures/forwards, to move risks from those who have them to

those who want them.
– Those “who want them” could be entities with naturally opposing risks or
speculators.
– Buyers of standard option products risk only the premium (analogous to an
insurance premium).
• The origin of these risks is often natural or institutional
– Producers of natural gas are naturally “long” gas. LDCs or gas-fired power
generators are naturally “short.”
– Mortgages typically have prepayment options by necessity/tradition.
• A very large source of interest-rate optionality
• A major driver of fixed-income options markets
– Gas storage facilities are the physical origin of gas price optionality
• Withdraw gas when the price is high; inject it when it is low.
– Power generation is a physical origin of optionality on the spread between power
prices and fuel prices:
• Generate when power prices exceed the cost of generating.

3
Basic Structures: Options

Options: Transactions for potential delivery/settlement at a future date.

• Example:
– You have the option to purchase 10,000 MMBtus of gas for delivery in Dec18 at
$3.10 / MMBtu.
– You must choose whether or not to purchase the gas on the exercise date which
is fixed in the contract.
– Clearly you will exercise this option only if the price of gas for Dec18 on the
exercise date exceeds $3.10.
• Terminology:
– Option types:
• Call option: is the option to purchase the underlying asset
• Put option: is the option to sell the underlying asset
– If you “exercise the option”: you have opted to purchase (call) or sell (put) the
asset.
– Exercise Type:
• European exercise: decision to exercise can be made on a single date by a
specified time of day. This date is the exercise date.
• American exercise: you can choose to exercise the option on any day prior to
the contracted exercise date.
– Strike: is the price at which you can purchase (call) or sell (put). Specified in
contract.

4
Basic Structures: Options

• To specify an option contract:

– Quantity (e.g. volume of gas)
– The strike price (e.g. $3.10)
– The exercise type (European)
– The option type: (Call or put)
– The exercise date
– Delivery date/terms
– Settlement: physical or financial
– Premium: The cost of the option
• Option payoff:
– Call: max( Underlying - Strike, 0 )
– Put: max( Strike - Underlying, 0 )

5
Basic Structures: Options

• Option Payoffs:
Exercise

Call
option
payoff
Asset Price at
Settlement
Strike price
No
exercise

• Note that the payoff is nonlinear.

– This makes valuation much more challenging than for forwards.

6
Basic Structures: Options

No
exercise
Put option
payoff
Exercise
Asset Price at
Settlement
Strike price

7
An important relationship: Put/Call Parity

• An innocuous mathematical equation

max( x, 0 ) = x + max( –x, 0 )
• Leads to a very important relationship amongst European options
Call = PV( Underlying – Strike ) + Put (assuming 0 dividends)
Proof: Two portfolios that with certainty have the same value at some known time in
the future must have the same value today.
• This is often written as

8
An important relationship: Put/Call Parity

• This relationship does not depend on the distribution of the

underlying.
• However it does depend on various, usually implicit, assumptions
– E.g. could be violated if C, P and F are with different counterparties of varying
credit.
• Aside:
– Put/Call Parity is/was a major source of profits for equity options market-makers,
even though those are American exercise.

9
Basic Structures: Options

Other Common Options Structures:

• Call Spread (aka Vertical Spread):
– Long a call at a low strike; short a call at a high strike
– Like a call but limited upside

Call spread
payoff

Asset Price at
Settlement
Low Strike High Strike

10
Basic Structures: Options

• Straddle/Strangle:
– Strangle: Long a put at a low strike; long a call at a high strike
– Straddle: A straddle where the low and high strikes are the same
– Often used by options traders as no immediate hedging is required

Straddle
payoff

Asset Price at
Settlement

Strike

11
Basic Structures: Options
• Collar:
– Short a put at a low strike; long a call at a high strike
– A producer (natural long) can acquire a put thereby protecting downside price risk,
funding this hedge by ceding potential upside price gains (selling a call).
– Strikes oftentimes chosen so that the premiums offset, resulting in a “Costless Collar”.

Collar payoff

Asset Price at
Settlement

12
Basic Structures: Markets

• What trades in the market?

• Liquidity and terms vary by region and commodity and instrument:
• Power:
– Greatest liquidity in PJM
– Not a whole lot in ERCOT
– Daily, monthly, and swaptions (aka “1x”) with both financial and physical
settlement. Also in the west, quarterly swaptions.

• Gas:
– A variety of options with all manner of exercise provisions
• Dominated by exchange options (exercising into an exchange futures
contract)

13
Natural Options Positions

Generation:
• Baseload Units: Nuclear/Coal
– Low-cost, relatively constant fuel price
– Can be viewed approximately as a fixed strike power option:
• The strike K is the cost of fuel per MWh of power

• Intermediate/Peaking Units: Combined cycle and peaking gas or oil fired

generation
– Approximately a spread option between power and gas
– Heat rate options
– Exercise can be viewed as hourly or daily depending on unit characteristics

• Key Point: The physical operational characteristics are complex and result
in optionality which is not identical (and is sometimes far removed) from
traded options contracts

14
Natural Options Positions

Full Requirements Load Obligations:

• Basic contract to serve wholesale load at a prescribed price.

• Two potential sources of embedded optionality:

– Fuel Adjustment Clauses: The contract price can be adjusted depending upon fuel
prices.
– Customers can be endowed with the flexibility to leave
• This is analogous to mortgage prepayment
• If prices drop well below contract prices customers (especially large ones) will
“attrit.”
• Attrition models try to value such embedded options we are short to customers.

15
Natural Options Positions
Natural Gas storage:

• Calendar Spread Options (CSOs).

• Single commodity
– High dimensional optionality, occurring at the daily level, each decision point is
coupled with optionality in the future
– The decision to inject or withdraw from storage affects the state of system.

Transportation and Transmission:

• Spread Options

• Firm capacity transaction involves effectively granting the buyer the right,
but not the obligation, to flow gas between two points in the system
• Electrical transmission contracts often specify a path, which is used for
financial reconciliation of all transmission charge. They can be analyzed in
term of spread options

16
Option Valuation Intro: One-Step Binomial

• What’s the value of a call option with strike $100 in the following scenario? Assume 0
interest rate and dividends, a deep super-liquid market in underlying, and absence of
arbitrage opportunities. (Hint: it is not $9.90.)

17
Option Valuation Intro: One-Step Binomial

• Everyone needs to go through this argument at least once.

• The idea is that you form a riskless portfolio of the form C – D S such
that the payoff at expiration is independent of the value of the
underlying.
• –D * ( 98 – 100 ) = ( 110 – 100 ) – D * ( 110 – 100 )
• 2 D = 10 – 10 D
• D=5/6
• Thus no arbitrage a the fair premium is C = 2 * 5 / 6 = 5 / 3, approx
$1.67.
• Note that the probability p = 99% does not show up anywhere.
• Exercise: Do the same but with $98 replaced by $104.

18
Summary of Option Valuation

• The value of an option depends upon:

– Underlying price
– Strike
– Interest rates
– Time to maturity
– Volatility:
• This is the new ingredient.
• Intuitively if you are long a put or call you like higher volatility
• Movements in the positive direction make more than movements in the negative direction lose, due to convexity of the payoff
function.
• Measured in terms of standard deviation of annualized returns

• The theory underpinning option valuation (i.e. Black-Scholes and successors) is

predicated on ideal frictionless markets and the assumption that asset returns are
random walks with normal distributions.
• The results:
– Formulas (or algorithms) for computing option values
– Prescriptions of trading strategies which “replicate” option payoffs
– Sensitivities of option values (Greeks) to the key variables.

19
Summary of Option Valuation

• A picture is worth a thousand equations:

20
Summary of Option Valuation
• Dependence in value on underlying price
– The slope of the value function is called “Delta”
– If you sell Delta of the underlying you have neutralized your option value to small underlying price changes.

21
Summary of Option Valuation

• Dependence on time to expiry

– Value converges to terminal value as time to expiry falls
– This decay in value of long options positions is referred to as theta.
– When you are long a put or a call you will (generally) experience negative theta because time passage lowers the value of
your option. Exceptions: Deep ITM call, underlying pays dividends or Deep ITM put, positive interest rate.

22
Summary of Option Valuation
• Dependence on volatility
– Practically the same as changes in expiry
– Higher vol means higher value
– The sensitivity of option value to volatility changes is called vega.
– When you are long a call or a put you have positive vega because a rise in vol makes your option worth more (circular?).
– Actually, I think the real explanation is that roughly speaking increasing vol broadens the terminal distribution and pulls the “center of mass” of the
conditional densities f( S | S > K ) and f( S | S < K ) away from K.

23
 Summary of Option Valuation
Extracting value from a long option position:
• Hope that the price moves the right way (pure speculation).
• Go out to the market and sell a similar option if such trades.
– If identical options, cancels your option position
– Gets you premium from the option sale.

• Delta hedge: If your current option position has a given D, hold –D in the underlying at all times.
– This requires continuously changing your holding of the underlying.
– If you are long the option it rotates your option payoff so that all movements of the underlying are good for you (both positive and negative).

24
Summary of Option Valuation

Black-Scholes is simply a trade-off between what you gain from underlying price movements when you
are delta hedging due to curvature of your payoff (gamma) and time decay (theta).

25
The Role of Models and Limitations

• Ideally more sophisticated option models can be used to extrapolate from known option prices
(traded vanillas) to values of exotic structures in a portfolio.
– The more “removed” the exotic structure is from the structure of vanilla options the more important the form of the
option model becomes.
• If our portfolio only contained liquid vanilla options then sophisticated option valuation models
would be of less value.
– If options market trade liquidly then we would know the value of our options by simply observing the market prices.
• If done correctly, prevents from having arbitrages in the system (e.g. monthlies > dailies).

Limitations
• Price returns are not normally distributed.
– Argument for still using normal distributions is that it is analytically much more tractable and it gets you most of the way
there regarding option portfolio management.
– Arguments against include “Great Intellectual Fraud”.
– Also spot power prices are very non-normal, part of the reason a hybrid model is used for structured products valuation.
• Markets are not frictionless:
– This is really a problem in power; large bid-offer spreads
– Block size constraints: 50MW or 25MW blocks trade. If your delta is changing by 10MW day to day, there isn’t much you
can do with delta hedging.
• Markets are illiquid or don’t exist:
– How do you calibrate an option model (volatilities) to a nonexistent market?
– Possibilities include use of a proxy market (e.g. PJM), statistical analysis of historical data, or combination of both.

26