Essays in financial econometrics

Abstract

Chapter 1 is the product of joint work with Ferhat Akbas and it provides a behavioral explanation for monthly negative serial correlation in stock returns. For the first time in the literature, this work reports that only low momentum stocks experience monthly negative serial correlation. Using a recently collected dataset, this finding provides the basis for a behavioral explanation for monthly negative serial correlation. Chapter 2 uses mean squared error (MSE) criterion to choose the number of instruments for generalized empirical likelihood (GEL) framework. This is a relevant problem especially in financial economics and macroeconomics where the number of instruments can be very large. For the first time in the literature, heteroskedasticity is explicitly modelled in deriving the terms in higher order MSE. Using the selection criteria makes GEL estimator more efficient under heteroskedasticity. Chapter 3 is the product of joint work with Victor Chernozhukov and Konrad Menzel.This chapter proposes new ways of inference on mean-variance sets in finance such as Hansen-Jagannathan bounds and Markowitz frontier. In particular standard set estimation methods with Hausdorff distance give very large confidence regions which are not very meaningful for testing purposes. On the other hand confidence regions based on LR-type statistic and wald type statistic provide much tighter confidence bounds. The methodology is also extended to frontiers that use conditional information efficiently.