Date of Submission


Date of Award


Institute Name (Publisher)

Indian Statistical Institute

Document Type

Doctoral Thesis

Degree Name

Doctor of Philosophy

Subject Name

Quantitative Economics


Economic Research Unit (ERU-Kolkata)


Sarkar, Nityananda (ERU-Kolkata; ISI)

Abstract (Summary of the Work)

Since the 1970's it has been observed in many economies that financial and macroeconomic variables like equity prices, treasury bill rates and exchange rates have become more and more volatile in nature. This may be due tumor flexible monetary policies pursued in these countries as well as due to their increasing exposure towards various international developments. Accordingly, economic agents are facing increasingly more and more risky environment. Re- searchers as well as professional economists in the area of capital and business finance have, therefore, been increasingly attracted in recent years towards studying the effect of risk and uncertainty on asset returns, and framing rational decision rules for individuals and institutions for the purpose of selecting security portfolios. The increased importance played by risk and uncertainty considerations in modern economics and finance theory has necessitated the development of new econometric time series modelling techniques that allow the higher-order moments to be time dependent as opposed to the traditional nacroeconomic and financial time series modelling approach which mainly centres on the conditional first moment.While it has been recognized for quite some time that the uncertainty in speculative prices, as measured by variances and covariances, changes through time ( see, for instance, Mandelbrot, 1963; and Fama, 1965), it was not until the introduction of what is now known as Modern Financial Econometrics that applied researchers in financial and monetary economics have started explicitly modelling time variation in second or higher-order moments. In a seminal paper in 1982, Engle introduced the autoregressive conditional het- eroscedastic (ARCH) model. This model allows the conditional variance to change over time as a function of past errors keeping the unconditional variance constant. It has been observed that such models capture many empirically observed temporal behaviours like thick tail distribution and volatility clustering of many economic and financial variables (see, Bera and Higgins, 1993; Bollerslev, Chou and Kroner, 1992; Bollerslev, Engle and Nelson, 1994; Shephard, 1996; and Gourieroux, 1997 for excellent surveys on ARCH model and its various generalizations).The basic ARCH model has been generalized in different directions so as to broaden its applicability. One important generalization of ARCH model is what is known as ARCH in the mean (ARCH-M) model which was first introduced by Engle, Lilien and Robins (1987). This generalization has, in fact, provided a methodology to estimate and to test the conditional version of capital asset pricing model (CAPM). The origin of ARCH-M model may be traced back to the unconditional CAPM of Lintner-Sharpe-Mossin, which is considered to be the most widely used theoretical model for specifying the relation between risk and return. In the early applications of mean-variance capital asset pricing of Sharpe (1964), Lintner (1965) and Mossin (1966), the expected return and its covariance with the market return were assumed to be constant. Over the last two decades, this relationship has been modified and reoriented in a different manner and used in awide range of financial applications. The basic risk-return relationship, called the capital asset priciing model of sharpe (1965) and Mossin (1966), Procides that for an asset i the relationship is expressed as E(yi)=bCov(yi,rm)


ProQuest Collection ID:

Control Number


Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.


Included in

Mathematics Commons