Aristotle University Of Thessaloniki (AUTH)
Kρυφές στοχαστικές διαδικασίες με άλματα και μοντέλα χώρου καταστάσεων με ανισοτικούς περιορισμούς. Εφαρμογή στη Χρηματοοικονομική
Abstract
dc:descriptionThe first chapter of the dissertation introduces the reader in the research field which is under study. The importance of stochastic jump processes in finance is highligthed and the basic principals of stochastic state space modeling are demonstrated. In the second chapter, a brief description of the achievements and contribution of the present dissertation in the pertaining research field is presented. In the third chapter, a new, semiparametric method for the estimation of the jump beta coefficients is presented. These coefficients measure the influence of the market jumps on the stock returns, and are latent components. It is proved that the proposed methodology admits always a unique real solution in terms of the jump betas, while it also serves as a criterion for testing the fit of any sets of the coefficients to the empirical returns. In the sequel, the topics of the following chapters are related to the field of stochastic state space modeling, when the hidden state vector is subject to inequality constraints. New, nonlinear methods are proposed for the estimation of the state vector satisfying the inequality constraints for every time step, and these methods are applied in finance for the estimation of the hidden return asset jumps. More specifically, the remaining chapters are organized as stated below. In the fourth chapter, a new methodology is proposed concerning the estimation of a hidden state vector, when its components are subject to inequality constraints. The proposed nonlinear method consists of three stages, which are summarized as follows: (a) the use of Kalman filter, (b) truncation of probability density functions, and (c) scaling. Thereafter, an application of the proposed method for the estimation of the hidden positive and negative jumps of asset returns is illustrated by constructing an appropriate state space model. In the fifth chapter, the method proposed in the previous chapter is modified in order to take into consideration the characteristic of conditional heteroskedasticity for the estimation of a hidden state vector, which is subject to inequality constraints. Moreover, a comparison between the two methodologies is presented concerning the results of the estimation procedure for the hidden two-sided asset return jumps. Finally, in the sixth chapter, an exponential state space model is presented for the estimation of a hidden state vector, when its components take values in a left or right bounded space. The recursive relations for the estimators are derived and the proposed model is used for the estimation of the hidden asset return jumps.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2019
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Theodosiadou, Ourania
- Θεοδοσιάδου, Ουρανία
Subjects
dc:subject × 12Rights
dc:rights- Statement dc:rights
-
- BY_NC_ND
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/46188