Behavioural Finance relaxes the neoclassical assumption that investors consistently apply Bayes Rule when updating their expectations, and identifies the behavioural attributes that affect asset prices. This thesis extends this literature by examining deviations from the Bayesian model that arise due to i) ambiguity aversion, ii) investor sentiment and iii) decision heuristics. Bayesian Updating assumes that investors are able to always estimate a single generating process for expected returns. However, in reality investors analyze noisy information signals that relate to this unknown distribution in a latent way, and it is likely that they are not always able to determine a single probability distribution. Behavioural economists have shown that in such conditions of uncertainty about probabilities people become pessimistic. The first chapter examines whether the pricing of analyst earnings is affected by ambiguity aversion, offering confirmatory evidence. A behavioural literature shows that people in good sentiment make optimistic choices, relative to objective probabilities. The second chapter examines whether investor sentiment affects the performance of the momentum trading strategy, an anomaly related to the pricing of good and bad information. The results indicate that sentiment strongly affects the momentum phenomenon, suggesting that it is triggered from investors’ behavioural biases. It has been suggested that deviations from Bayesian Updating arise due to heuristics triggered by the characteristics of the information used. The last chapter examines the validity of one such important hypothesis proposed by Griffin and Tversky (1992) using rigorous experimental economics techniques. The results confirm this hypothesis, indicating that investors are likely to overreact to salient information signals with low predictive validity.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:524035 |
| Date | January 2010 |
| Creators | Antoniou, Constantinos |
| Publisher | Durham University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://etheses.dur.ac.uk/428/ |
Page generated in 0.0158 seconds