Weather Derivatives: Modelling, Pricing and Applications
Date
2014-07-21
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
This thesis investigates the modelling of weather dynamics, pricing and applications of weather derivatives. Since the dominant weather effect in the current weather related financial market is temperature, our focus relies on temperature derivatives modelling, pricing and applications in Canadian temperature futures market. A brief overview on weather markets, weather indices and the literatures of modelling and pricing weather derivatives are introduced in Chapter 1. Then we first apply a standard Ornstein-Uhlenbeck process driven by general L\'{e}vy process for modelling the daily average temperature to Canadian cities data in Chapter 2. The model is characterized by mean-reverting property with time-dependent seasonal mean level and volatility and the driving noise in the model is extended to the general L\'{e}vy process. As an extension to the first model, continuous-time autoregressive (CAR) model driven by general L\'{e}vy process is considered and calibrated to Canadian data in Chapter 3. The discretization of the stochastic differential equation makes the calibration of these two models more convenient and applicable. These two models are also applied to derivation of explicit futures price of CAT index, and numerical prices of CDD and HDD futures. In Chapter 4, by analyzing some key properties of daily average temperature dynamics, we propose two two-state regime-switching models to capture the movement of temperature. Both these two regime-switching models have a ``normal" regime along with a ``jump" regime. The ``normal" regime is characterized by a standard Ornstein-Uhlenbeck process. For the ``jump" regimes, we use a Brownian motion with more extreme drift and volatility and a L\'{e}vy jump as the process which driving the abnormal positive or negative jumps in the temperature dynamics. Unlike the step by step calibration procedure for the first two models, we employ a modified Expectation-Maximization method to obtain the hidden Markov states and estimate parameters in these two regime-switching models. The CAT, CDD and HDD futures pricing approaches are also studied for both regime-switching models. For applications of the temperature futures, firstly, we give out a static, simple strategy example to hedge the volume risks in practical. Then, by building up a system of models for energy and temperature, we propose a dynamic hedging strategy for the hedging of the energy futures using temperature futures. Finally, in Chapter 6, we conclude and state the future works in this research topic.
Description
Keywords
Economics--Finance, Mathematics, Energy
Citation
Cui, K. (2014). Weather Derivatives: Modelling, Pricing and Applications (Doctoral thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca. doi:10.11575/PRISM/28679