Advanced

Pricing and Hedging of Swing Options in the European Electricity and Gas Markets

Hedestig, John (2014) FMS820 20141
Mathematical Statistics
Abstract (Swedish)
This report outlines a method to price and hedge a generalized swing option based on the European natural
gas and electricity markets. The method is model free in that is does not assume a certain spot price dynamics.
It only requires a forward curve and European call options prices that cover the delivery period of the
swing option. The results is a lower bound derived from the price of forwards and call options on the market.
A method to approximate the Greeks of the swing option by evaluation on the lower bound is suggested.
This approximation is based on a Finite dierence method. Two hedges are constructed from the information.
The first hedge is based on the weights of forwards and call options from the calculation of the lower
... (More)
This report outlines a method to price and hedge a generalized swing option based on the European natural
gas and electricity markets. The method is model free in that is does not assume a certain spot price dynamics.
It only requires a forward curve and European call options prices that cover the delivery period of the
swing option. The results is a lower bound derived from the price of forwards and call options on the market.
A method to approximate the Greeks of the swing option by evaluation on the lower bound is suggested.
This approximation is based on a Finite dierence method. Two hedges are constructed from the information.
The first hedge is based on the weights of forwards and call options from the calculation of the lower
bound. The second hedge is a Greek neutralizing hedge.
We conduct an empirical study on the lower bound. A Delta and Delta-Gamma Hedge is evaluated and
we find the Greeks of the lower bound to be a blunt approximation of the swing option uncertainties. The
evaluation reveals that the lower bound approximation eects the dynamics and hence the Greeks of the
swing option. In addition the Finite dierence method is unstable in its approximations, especially for finer
granularity.
The lower bound is also compared to an existing Least Square Monte Carlo (LSMC) method. It is much
faster than the LSMC and a price comparison give inconclusive results. Additional studies reveal a pricing
defect lie within the much more complex LSMC. Finally a granularity study is implemented. Finer granularity
increases the price of the lower bound slightly but the eects on the Greeks are more significant.
Whether this eect on the Greeks is due to changes in the swing option or due to better approximations with
the lower bound cannot be concluded. (Less)
Please use this url to cite or link to this publication:
author
Hedestig, John
supervisor
organization
course
FMS820 20141
year
type
H2 - Master's Degree (Two Years)
subject
language
English
id
4584529
date added to LUP
2014-08-06 12:08:32
date last changed
2014-08-06 12:08:32
@misc{4584529,
  abstract     = {This report outlines a method to price and hedge a generalized swing option based on the European natural
gas and electricity markets. The method is model free in that is does not assume a certain spot price dynamics.
It only requires a forward curve and European call options prices that cover the delivery period of the
swing option. The results is a lower bound derived from the price of forwards and call options on the market.
A method to approximate the Greeks of the swing option by evaluation on the lower bound is suggested.
This approximation is based on a Finite dierence method. Two hedges are constructed from the information.
The first hedge is based on the weights of forwards and call options from the calculation of the lower
bound. The second hedge is a Greek neutralizing hedge.
We conduct an empirical study on the lower bound. A Delta and Delta-Gamma Hedge is evaluated and
we find the Greeks of the lower bound to be a blunt approximation of the swing option uncertainties. The
evaluation reveals that the lower bound approximation eects the dynamics and hence the Greeks of the
swing option. In addition the Finite dierence method is unstable in its approximations, especially for finer
granularity.
The lower bound is also compared to an existing Least Square Monte Carlo (LSMC) method. It is much
faster than the LSMC and a price comparison give inconclusive results. Additional studies reveal a pricing
defect lie within the much more complex LSMC. Finally a granularity study is implemented. Finer granularity
increases the price of the lower bound slightly but the eects on the Greeks are more significant.
Whether this eect on the Greeks is due to changes in the swing option or due to better approximations with
the lower bound cannot be concluded.},
  author       = {Hedestig, John},
  language     = {eng},
  note         = {Student Paper},
  title        = {Pricing and Hedging of Swing Options in the European Electricity and Gas Markets},
  year         = {2014},
}