Date of Award

1995

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Finance

First Advisor

G. Geoffrey Booth

Abstract

This dissertation examines the price discovery and extreme value processes found in Germany's stock index and stock index futures markets. These two concepts are framed within the fundamental relationship between risk and return found in the financial economics literature. Results from the price discovery analysis indicate that the stock index futures market processes information more quickly than the underlying spot market. However, this processing can be characterized by a feedback loop because sometimes the spot market processes information more quickly than the futures market. An indepth analysis of the information processing relationship implies that the futures market processes information faster than most of the individual stock index component stocks. However, two securities sometimes lead the futures market. The reasons these two securities lead the futures market are of particular interest. Additionally, the processing speed of the futures market tends to be increased when there is market-wide, as opposed to security-specific, information affecting the securities market. Also, analyses of up and down markets and different trading activity proxies are performed. They reveal that the lead-lag relation is conditional on the information set available to market participants. The results of the extreme value section indicate that extreme price declines for most FDAX contracts are larger in absolute terms than extreme price increases. The results of the extreme value analysis also indicate that the data generation process of the extreme price changes originate from a Type II extreme value process. An examination of prudent margin setting procedures is made as a practical application of extreme value theory. The results of this part of the study indicate that the extreme value distribution approximates the empirical extreme value observations better than the normal distribution process.

Pages

202

DOI

10.31390/gradschool_disstheses.5999

Share

COinS