Three essays on financial markets

[ACCESS RESTRICTED TO THE UNIVERSITY OF MISSOURI-COLUMBIA AT AUTHOR'S REQUEST.] This dissertation consists of three essays. The first essay extends Kyle (1985)'s informed trading model to allow for multiple informed traders, noise in informed traders' private signals and the probability of information events. The first essay's main theoretical findings are: (a) Other than the intuitive effect of incurring loss, noise in private signals may preserve informed traders' profit by dampening competition. When the noise level is relatively low, the competition dampening effect dominates. What's more important, there is an optimal noise level at which informed traders' profit is maximized. (b) If the optimal noise level can be implemented, then strong form informational efficiency is not achievable in days with information events, and the size of potential bias depends on both the probability and magnitude of information events. The theoretical section also includes simulation results which support theoretical finding (b). The empirical part of this essay extends Odders-White and Ready (2008) and finds that: (1) Estimated noise in private signal is higher for small firm stocks, which is intuitive since information leaks are less likely for small firms and small firms are more secretive. (2) Estimated noise level is larger for stock/year combinations with larger estimated number of informed traders and larger estimated magnitude of information events. Result (2) confirms theoretical finding (a), which states that optimal noise level is increasing in the number of informed traders and magnitude of information events. The second essay explains how an airline can hedge more effectively employing dynamic copula models. Since there is no corresponding futures contract available, airlines may have to cross hedge jet fuel with either crude oil or other refined products. Because the dependence between jet fuel and its hedging instrument is time-varying, airlines' cross hedging problem needs a dynamic solution. This essay estimates timevarying optimal cross hedge ratios using dynamic copula based GARCH models. Estimation results show that copula models outperform static hedging models by achieving a smaller hedged portfolio variance in both in-sample (by about 5%) and out-of-sample tests (by about 9%). This essay also finds that heating oil futures is a more effective cross hedging instrument for jet fuel than WTI crude oil futures. The third essay seeks to investigate the impact of corporate hedging. First of all, this essay examines whether oil/jet fuel price increases impact Southwest stock/Dow Jones Transportation index returns and return volatilities significantly. Assuming Southwest airlines' hedging effort is effective, this essay tries to find supportive evidence showing that Southwest stock return and return volatility are less sensitive to positive oil/jet fuel price shocks than the DJT counterparts. While intuition suggests oil/jet fuel price increases, especially the large ones, may reduce transportation firms' profitability substantially, the link between positive oil shocks and stock returns as well as the link between positive oil shocks and stock return volatilities are found to be weak. This essay shows that positive oil shocks' impacts on Southwest/DJT returns and volatilities are very limited, though there is a tail dependence exception: extreme positive oil price increases do make large negative stock returns more likely. Although intuition suggests Southwest stock return and return volatility may be less sensitive to oil/jet fuel price shocks due its hedging effort, the analysis of this essay does not find any profound evidence supporting this conjecture.

Name thesis:degree_name

Ph. D.

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Economics (MU)

Grantor dc:publisher

University of Missouri--Columbia

Year dc:date.issued

2014

Author dc:creator

• Cheng, Tengda

Advisor dc:contributor.advisor

• Miller, James

Statement dc:rights

• Access to files is limited to the University of Missouri--Columbia with SSO login.

Language dc:language.iso

eng, English

DOI

https://doi.org/10.32469/10355/46415

OAI identifier oai:identifier

oai:mospace.umsystem.edu:10355/46415