Bhar / Hamori Empirical Techniques in Finance

2005
ISBN: 978-3-540-27642-5
Verlag: Springer
Format: PDF
Kopierschutz: Wasserzeichen (»Systemvoraussetzungen)

E-Book, Englisch, 243 Seiten, eBook

Reihe: Springer Finance

This book offers the opportunity to study and experience advanced empi- cal techniques in finance and in general financial economics. It is not only suitable for students with an interest in the field, it is also highly rec- mended for academic researchers as well as the researchers in the industry. The book focuses on the contemporary empirical techniques used in the analysis of financial markets and how these are implemented using actual market data. With an emphasis on Implementation, this book helps foc- ing on strategies for rigorously combing finance theory and modeling technology to extend extant considerations in the literature. The main aim of this book is to equip the readers with an array of tools and techniques that will allow them to explore financial market problems with a fresh perspective. In this sense it is not another volume in eco- metrics. Of course, the traditional econometric methods are still valid and important; the contents of this book will bring in other related modeling topics that help more in-depth exploration of finance theory and putting it into practice. As seen in the derivatives analysis, modern finance theory requires a sophisticated understanding of stochastic processes. The actual data analyses also require new Statistical tools that can address the unique aspects of financial data. To meet these new demands, this book explains diverse modeling approaches with an emphasis on the application in the field of finance.

Zielgruppe

Research

Autoren/Hrsg.

Bhar, Ramaprasad

Hamori, Shigeyuki

Inhaltsverzeichnis

Basic Probability Theory and Markov Chains.- Estimation Techniques.- Non-Parametric Method of Estimation.- Unit Root, Cointegration and Related Issues.- VAR Modeling.- Time Varying Volatility Models.- State-Space Models (I).- State-Space Models (II).- Discrete Time Real Asset Valuation Model.- Discrete Time Model of Interest Rate.- Global Bubbles in Stock Markets and Linkages.- Forward FX Market and the Risk Premium.- Equity Risk Premia from Derivative Prices.

Leseproben

4 Non-Parametric Method of Estimation (p.31)

4.1 Background

In some financial applications we may face a functional relationship between two variables Y and X without the benefit of a structural model to restrict the parametric form of the relation. In these situations, we can apply nonparametric estimation techniques to capture a wide variety of non- Hnearities without recourse to any one particular specification of the nonUnear relation. In contrast to a highly structured or parametric approach to estimating non-linearities, nonparametric estimation requires few assumptions about the nature of the non-linearities.

This is not to say that the approach is free of drawbacks. To begin with, the highly data-intensive nature of the process can make it somewhat costly. Further, nonparametric estimation is poorly suited to small samples and has been found to over fit the data. A regression curve describes the general relationship between an explanatory variable X and a response variable Y. Having observed X, the average value of Y is given by the regression function. The form of the regression function may teil us where higher Y-values are to be expected for certain values of X or where a special sort of dependence is indicated. A pre-selected parametric model might be too restricted to fit unexpected features of the data. The term "non-parametric" refers to the flexible functional form of the regression curve.

The non-parametric approach to a regression curve serves four main functions. First, it provides a versatile method for exploring a general relationship between two variables. Second, it gives predictions of observations yet to be made without reference to a fixed parametric model. Third, it provides a tool for finding spurious observations by studying the influence of isolated points. Fourth, it constitutes a flexible method for substituting missing values or interpolating between adjacent X values. The flexibility of the method is extremely helpful in a preliminary and exploratory Statistical analysis of a data set. When no a priori model Information about the regression curve is available, non-parametric analysis can help in providing simple parametric formulations of the regression relationship.