Use this URL to cite or link to this record in EThOS:
Title: Model-free moment indices : theory, construction and application
Author: Leontsinis, Stamatis
ISNI:       0000 0004 2716 9631
Awarding Body: University of Reading
Current Institution: University of Reading
Date of Award: 2010
Availability of Full Text:
Access from EThOS:
The growing interest in volatility trading, by many types of financial institutions, has led to a recent surge of interest amongst academics. Variance swaps are the most popular pure volatility trade derivatives, so understanding them is of imperative importance. So therefore, we provide an extensive empirical work in variance swaps spotting trading patterns via using all of the most widely known volatility indices. Additionally, we perform a comprehensive distribution fitting exercise to the variance swap returns, which provides in sights to their accurate modelling. We also show that the 'model-free' volatility index formula, which the exchanges base their volatility indices upon, is actually model-dependent. The main theoretical contribution of this thesis lies in the derivation of a completely model-free risk-neutral moment generating function which is not constrained by the usual underlying and volatility dynamics assumptions. Our model-free risk-neutral moment generating function can generate higher moment indices for assets which are traded or not. Extending the literature beyond the second moment, we derive formulae for model- free volatility, skewness and kurtosis based on our moment generating function. The estimation of these formulae is free from the problems and constraints that the calibration techniques entail. We also test the numerical procedures that are used by the exchanges, finding that they can be considerably improved. A time period of more than 16 years of FTSE 100 options provides a challenging environment in which to implement our theoretical results. Important observations of these data help the adjustment of the model-free formulae to minimise any data related errors, particularly those relating to 'cabinet' options. The absence of a volatility index in the British market inspired the construction of the VFTSE, the first volatility index on the FTSE 100 index. We show that the standard methodology consistently overestimates volatility and that the standard distributions used for model-free volatility modelling are incorrect. Viable alternatives are proposed. Following this, we present for the first time in the literature, term structure of skewness and excess kurtosis time series indices, the SFTSE and KFTSE indices. Using those moment indices, we build and analyse FTSE 100 forward densities using four different methodologies. Our results have important implications for VaR analysis, volatility trading and risk management. Finally, we discuss an options trading application of our moment indices, using FTSE 100 forward densities to spot mispriced FTSE 100 options.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available