Date of Award

8-2004

Degree Name

Doctor of Philosophy

Department

Economics

First Advisor

Dr. Matthew L. Higgins

Second Advisor

Dr. Onur Arugaslan

Third Advisor

Dr. Debasri Mukherjee

Abstract

My dissertation consists of three essays on the econometric analysis of financial volatility.

My first essay is titled "The Runs Test for Volatility Forecastibility: Extensions and Comparisons with Tests for GARCH." Recently, Diebold and Christoffersen (2000) introduced a test for forecastable volatility. In this paper, I compare the size and the power of the runs test and the optimal LM test for GARCH by Monte Carlo simulation. For high frequency returns the LM test has superior power to the runs test. For low frequency returns however, the tests have very similar power. I also propose a switching variance model. For this process, I find that the runs test has greater power than the LM test.

The second essay of the dissertation I drive the population moments of criteria commonly used to evaluate accuracy of volatility forecasts from GARCH models. I state the existence conditions for the population moments. The criteria include the mean squared error (MSE), the mean absolute error (MAE) and a heteroscedasticity adjusted mean square error (HMSE). Using Monte Carlo simulation, I analyze the sampling properties of these criteria and the sampling properties of the R2 's and t-statistics from the Mincer and Zarnowitz (1969) regression. When volatility is highly persistent, I find that the majority of thesampling distribution of the R 2 lies below the population R2 . Also, the t-statistics for testing forecast efficiency are unreliable. For a logarithmic version of theMincer and Zarnowitz regression, I find that R 2 's tend to be smaller, but inference concerning forecast efficiency are valid. Among the accuracy criteria I find that the HMSE is preferable.

My third essay considers situations when the loss function is asymmetric. Most of the forecast criteria used in the literature consider symmetric loss functions like MSE because of mathematical convenience. However, forecast evaluation results are very sensitive to the proper specification of the loss function. In this paper, I use both parametric and nonparametric estimation techniques for the optimal predictor of volatility when the loss function is asymmetric. The results suggest that a constant/time-varying bias term is important.

Access Setting

Dissertation-Open Access

Included in

Economics Commons

Share

COinS