Computational Methods in Asset Valuation Problems

Postgraduate Thesis uoadl:1320590 1152 Read counter

Unit:
Κατεύθυνση Στατιστική και Επιχειρησιακή Έρευνα
Library of the School of Science
Deposit date:
2015-06-30
Year:
2015
Author:
Ζωσιμάς Ζωσιμάς
Supervisors info:
Απόστολος Μπουρνέτας (Καθηγητής - επιβλέπων καθηγητής), Αθανάσιος Γιαννακόπουλος (Καθηγητής), Κωνσταντίνος Μηλολιδάκης (Αναπληρωτής Καθηγητής)
Original Title:
Υπολογιστικές Μέθοδοι σε Χρηματοοικονομικά Προβλήματα Αποτίμησης Τίτλων
Languages:
Greek
Translated title:
Computational Methods in Asset Valuation Problems
Summary:
The dramatic development of the financial sector in the recent years has
resulted in the creation of new financial assets and the further development of
already existing ones. A consequence is the use of options on a regular basis.
Because of the large values traded a close scrutinization of this asset is
absolutely necessary. Taking the above into consideration, our objective is the
study of options and more specifically of American options, by presenting and
implementing various pricing methods. We present the binomial lattice model and
the extrapolation method using Richardson extrapolation. We examine the random
tree method. This method produces two consistent and asymptotically unbiased
estimators. In order to improve the effectiveness we use enhancement techniques
e.g. antithetic variables and Latin hypercube sampling. Furthermore, we analyze
the parametric approximations method. With this method, we find the best value
within a parametric class. Moreover we discuss the state – space partitioning
method by using finite – state dynamic programming. The latter three methods
are tested for various options and the results are compared for different
values of the parameters.
Keywords:
Option pricing, American option, Monte Carlo simulation, Dynamic programming, Financial engineering
Index:
Yes
Number of index pages:
xi-xiv
Contains images:
Yes
Number of references:
35
Number of pages:
xiv,143
File:
File access is restricted.

document.pdf
2 MB
File access is restricted.