The Econometric Modelling of Financial Time Series

by Terence C. Mills

• ISBN13: 9780521624923
• ISBN10: 0521624924
• Edition: 2nd
• Format: Paperback
• Copyright: 1999-09-28
• Publisher: Cambridge University Press

Summary

Substantially revised and updated second edition of Terry Mills' best-selling graduate textbook The Econometric Modelling of Financial Time Series. The book provides detailed coverage of the variety of models that are currently being used in the empirical analysis of financial markets. Covering bond, equity and foreign exchange markets, it is aimed at scholars and practitioners wishing to acquire an understanding of the latest research techniques and findings, and also graduate students wishing to research into financial markets. This second edition includes a great deal of new material, and also provides a more in-depth treatment of two crucial, and related, areas: the theory of integrated processes and cointegration. The new material discusses the distributional properties of asset returns and more recent and novel techniques of analysing and interpreting vector autoregressions that contain integrated and possibly cointegrated variables. Data appendix available online at www.lboro.ac.uk/departments/ec/cup.

Author Biography

Terence Mills is Professor of Economics at Loughborough University. He obtained his Ph.D. at the University of Warwick, lectured at the University of Leeds, and has held professorial appointments at City University Business School and the University of Hull.

Table of Contents

Preface to second edition	vii
Introduction	1	(7)
Univariate linear stochastic models: basic concepts	8	(53)
Stochastic processes, ergodicity and stationarity	8	(3)
Stochastic difference equations	11	(2)
ARMA processes	13	(15)
Linear stochastic processes	28	(1)
ARMA model building	28	(9)
Non-stationary processes and ARIMA models	37	(11)
ARIMA modelling	48	(5)
Forecasting using ARIMA models	53	(8)
Univariate linear stochastic models: further topics	61	(61)
Determining the order of integration of a time series	62	(37)
Decomposing time series: unobserved component models and signal extraction	99	(8)
Measures of persistence and trend reversion	107	(7)
Fractional integration and long memory processes	114	(8)
Univariate non-linear stochastic models	122	(55)
Martingales, random walks and non-linearity	122	(2)
Testing the random walk hypothesis	124	(2)
Stochastic volatility	126	(5)
ARCH processes	131	(22)
Other non-linear univariate models	153	(18)
Testing for non-linearity	171	(6)
Modelling return distributions	177	(28)
Descriptive analysis of three return series	177	(1)
Two models for return distributions	178	(6)
Determining the tail shape of a returns distribution	184	(4)
Empirical evidence on tail indices	188	(5)
Testing for covariance stationarity	193	(3)
Modelling the central part of returns distributions	196	(2)
Data analytic modelling of skewness and kurtosis	198	(2)
Distributional properties of absolute returns	200	(3)
Summary and further extensions	203	(2)
Regression techniques for non-integrated financial time series	205	(48)
Regression models	205	(13)
ARCH-in-mean regression models	218	(3)
Misspecification testing	221	(12)
Robust estimation	233	(2)
The multivariate linear regression model	235	(3)
Vector autoregressions	238	(7)
Variance decompositions, innovation accunting and structural VARs	245	(3)
Vector ARMA models	248	(5)
Regression techniques for integrated financial time series	253	(53)
Spurious regression	253	(7)
Cointegrated processes	260	(8)
Testing for cointegration in regression	268	(5)
Estimating cointegrating regressions	273	(6)
VARs with integrated variables	279	(18)
Causality testing in VECMs	297	(2)
Fully modified VAR estimation	299	(3)
Impulse response asymptotics in non-stationary VARs	302	(4)
Further topics in the analysis of integrated financial time series	306	(34)
Testing for a single long-run relationship	306	(4)
Common trends and cycles	310	(5)
Estimating permanent and transitory components of a VECM	315	(3)
Present value models, excess volatility and cointegration	318	(16)
Generalisations and extensions of cointegration and error correction models	334	(6)
Data appendix	340	(2)
References	342