Use this URL to cite or link to this record in EThOS:
Title: Essays in decision making
Author: Zhang, Jie
Awarding Body: Lancaster University
Current Institution: Lancaster University
Date of Award: 2011
Availability of Full Text:
Access from EThOS:
In this thesis, we illustrate with a parametric specification how the Markowitz model, supplemented with a small degree of probability distortion, is able to explain majority choices in experimental evidence (e.g. Allais paradox), as well as gambling at actuarially unfair odds. We employ a graphical method to reveal that the different parameters employed in Allais experiments have vastly different implications for the ranges of probability distortion and parameters of the value function for which agents will act as if expected utility maximisers, even when they exhibit probability distortion. Regarding children's risk attitude, there was only one study, Harbaugh, Krause and Vesterlund (2002), prior to our work. Our experimental results in China for children and high school students are similar in some aspects, and broadly consistent with those of Harbaugh, Krause and Vesterlund (2002) in that over a range of low probabilities they appear to underweight small probabilities. Our results for Chinese adults are not consistent with the results reported for western respondents, but are consistent with those reported for some non-western respondents. We demonstrate that the assumption of an expo-power value function is able, depending on the degree of probability distortion assumed, to explain all or nearly all the majority choices in ten Allais experiments involving both small and large payoffs. We report two experiments to demonstrate that coefficient of variance (cv) is not a proper candidate for the basis of choice in general. Through estimates based on simulated and real data sets, we demonstrate how estimated parameter values are implicitly constrained to fit either expected utility or the cumulative prospect theory when power utility is assumed. Harrison et al. (2010)'5 data set is also used to show that the majority responses of agents reported in the experiment of Harrison et al. (2010) are consistent with the Markowitz model and that the estilTjated Markowilz model is the most parsimonious specification. We also conduct experiments and our results indicate that subjects' risk attitude is consistent with Markowitz (1952)'s assumption, especially that the degree of risk aversion rises with rising payoffs. We use a model employing an expo-power value function and Prelec's probability weighting function to explain our experimental results. The model also explains subjects betting on longshot unfair gambles, odds-on unfair gambles and 50-50 symmetriC unfair gambles, which Cumulative Prospect Theory (CPT) could not do.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available