E-Locus - Institutional Repository of the University of Crete

Home Search

Results - Details

Search command : Author="Λουλάκης" And Author="Μιχαήλ"

Current Record: 3 of 4

[Add to Basket]

Identifier

000347020

Title

Υπολογιστικές τεχνικές στη χρηματοοικονομία

Alternative Title

Computational methods in finance

Author

Μωραΐτη, Μαρίνα Μιχαήλ

Thesis advisor

Λουλάκης, Μιχάλης

Abstract

Monte Carlo methods are widely used in the simulation of various random phenomena. In the area of Finance they constitute an essential tool both in derivative pricing as well as in risk management. They are based on the analogy between volume and probability. This thesis examines and implements a variety of Monte Carlo methods in Finance and consists of four parts. The first part discusses the various methods for simulating random variables and random processes pahts, with an emphasis to normally distributed random variables and to Brownian Motion and Geometric Brownian Motion paths. The second part reviews central concepts from the theory of derivative pricing and introduces Monte Carlo estimates to price several derivatives. Part three focuses on techniques for reducing the variance of the regular Monte Carlo estimator, by introducing new estimates that share the desired properties of Monte Carlo estimators and at the same time manage to reduce variance. Finally in part four we talk about some basic models for the dynamics of the interest short rate and brie y about the VaR measure,i.e. the value at risk measure. We also estimate the 1-month VaR with 99% certainty of a specific portfolio, whose constituents' dynamics follow the CIR (Cox-Ingersoll-Ross) model for the evolution of the interest short rate.

Language

Greek

Subject

Brownian motion simulation

Derivative pricing

Estimator efficiency

Monte Carlo

Random variable generation

Value at risk

Variance reduction

Αξία σε κίνδυνο

Αποτελεσματικότητα εκτιμητριών