Essays on Incomplete Information in Financial Markets
Author
Summary, in English
The first essay, "The Effect of Information Quality on Optimal Portfolio Choice," analyzes the portfolio choices made by three types of agents, having access to different information sets. These three types of agents correspond to executives, stockbrokers and small investors, and we call them "fully informed agents," "insiders," and "outsiders," respectively. The fully informed agents are assumed to know the true dynamics of the economy, while the insiders and outsiders have to learn it from the realizations of different signals. While the outsiders only use the realizations of the stock returns in their learning, the insiders also have access to a private signal. The contributions of the first essay are threefold. First of all, we derive an explicit closed-form solution to the insider's problem. Second, we provide a comparative static analysis of the agents' solutions. The third contribution is a calibration to US data where we make a comparison of the best estimates as well as the optimal portfolios across the three types of agents.
The second essay, "Expected Life-Time Utility and Hedging Demands in a Partially Observable Economy," analyzes the expected life-time utility and the individual agents' hedging demands in a Lucas (1978) economy, in which the dividend drift term is unknown and mean-reverting. An expression for the partially informed agent's expected life-time utility is derived, and his hedging demand is analyzed. It is shown that the hedging demand consists of two components. In the case of a negative correlation between the true drift term and the dividend growth rate, the two hedging components work in opposite directions, and, assuming a positive equity premium, a conservative investor may end up having a positive hedging demand. Interestingly, this is in contrast to the findings in Brennan (1998), who analyzes optimal portfolio choice when agents learn about a constant drift term in the stock return.
In the third essay, "Can An 'Estimation Factor' Help Explain Cross-Sectional Returns?," it is first shown in a theoretical model that the expected excess return on any asset depends on its covariance not only with the market portfolio, but also with changes in the representative agent's estimate. Then, this model is tested in a GMM framework, and compared to the three-factor Fama-French model. We find some evidence suggesting that the estimation factor (the representative agent's best estimate of the drift in aggregate dividend growth) is priced. Further, a conditional version of our model is found to perform on a par with a conditional version of the Fama-French model, and the static version of our model is found to perform on a par with the static version of the Fama-French model. We demonstrate that the impact of the estimation factor is statistically and economically significant.
Department/s
Publishing year
2005
Language
English
Publication/Series
Lund Economic Studies
Volume
no. 124
Document type
Dissertation
Publisher
Department of Economics, Lund University
Topic
- Economics
Keywords
- ekonomiska system
- ekonomisk politik
- Financial science
- Finansiering
- ekonomisk teori
- economic policy
- Nationalekonomi
- ekonometri
- Economics
- econometrics
- economic theory
- factor pricing models
- hedging demands
- estimation risk
- partial information
- learning
- economic systems
Status
Published
Supervisor
ISBN/ISSN/Other
- ISSN: 0460-0029
Defence date
14 January 2005
Defence time
10:15
Defence place
Sal EC3:210, Holger Crafoords Ekonomicentrum i Lund
Opponent
- David Feldman (Professor)