Llicenciatura en Administració i Direcció d'Empreses
Llicenciatura en Economia

Tècniques de Previsió (10145)

The aim of the course is building, fitting, checking, and predicting with sophisticated linear and non-linear time series models, with emphasis on prediction. Additionally, the study of the empirical characteristics of certain financial time series. The coursework intends to train (at least at an intermediate level) two more fundamental skills: statistical programming using R, in order to carry out statistical computations and data analysis,  and scientific writing using LaTeX, for preparing presentations and writing projects/articles.

Course structure

There are three main components:

i) Lectures, where the core of the theory is developed, accompanied by
practical sessions using computer demonstrations

ii) 30-min student presentations during the course, where more
technical and specialized  aspects related to time
series are presented. The students are supervised to a certain
extent by the lecturer for the preparation of the material

iii) Student projects. The students start working on their projects as
early as the 3rd week of the course (and typically their presentations
in ii above relate to the methodological component of the chosen topics)


Contents

Tema 1. Background: probabilistic modelling and prediction, and the Black-Scholes model

Tema 2. Empirical analysis of financial time series

Tema 3. Elements of Markov chains

Tema 4. ARIMA models

Tema 5. Kalman Filter and state-space models

Main References

The are lectures notes which have been prepared for the course. Additionally,


Tsay (2005) Analysis of Financial Time Series
Harvey (1993) Time Series models
Davis and Brockwell (1987) Time series: theory and methods