MODELING AND RISK MANAGEMENT WITH APPLICATIONS IN FINANCIAL AND WEATHER DERIVATIVES

The purpose of this dissertation is to conduct an in-depth analysis to modeling, pricing and risk management for financial instruments traded in financial and weather market. This work is composed of one scientific report and four working papers that focus specifically on the following topics: The scientific report presents and discusses a review of recent literature on financial products used to protect against climate change. Through comparisons of different methodologies proposed by different researchers this study shows the necessity and importance of using weather derivatives and insurance contracts to better protect and develop world wide financial markets. In the first paper, we propose temperature-based risk management using hybrid financial instruments built on weather derivatives. %We model temperature time series and price one-month forward option contracts for hedging adverse outcomes and then show how a “negative” weather performance can be counterbalanced by the “positive” performance of a hedging over-the-counter financial instrument, which can be tailored to meet specific needs. The Value-at-Risk technique is exploited, in order to define the level of the critical temperature, to estimate the prices of such derivatives. The results presented in the paper are promising, and show that such contracts can positively cover temperature risk. In the second paper, we present a way to hedge temperature risk exploiting weather derivatives contracts by considering ‘tail events’ and the standard financial approach to tackle them such as Value-at-Risk and Expected Shortfall. We perform the application of risk measures through historical and parametric methods and analyze the effectiveness of Expected Shortfall based approach for hedging meteorological risk. Even, according to risk measure theory, Expected Shortfall captures diversification while Value-at-Risk does not, numerical results show that it is more convenient to enter a single contract that covers more months rather than monthly contracts spanned on the same period. In the third paper, we propose forecasting based on a stochastic model of the probability distribution. We suggest to incorporate model uncertainty by considering forecasting using dynamical stochastic evolutions of the probability distribution of the model in question. We have considered and compared the results of two different classes of autoregressive models where the so-called stationary distribution is and is not normal. At the end, we have compared the uncertainty measure based on worst case approach and the proposed approach using a Value-at-Risk based measure giving tighter estimates. In the fourth paper, we propose some functional linear models predicting wind speed from temperature data in which both the response and the covariate variables are functions. We have proposed the functional linear model observed in different patterns which are powerful enough to describe the dynamics of wind speed in time.