Park, Seungmoon; (2021) Essays on uncertainty and asymmetric information in financial markets. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text Essays on uncertainty and asymmetric information in financial markets.pdf Download (1MB) | Preview

Abstract

This thesis analyses the behaviour of traders in financial markets with various structure of traders, and presents how their actions are affected by uncertainty and information asymmetry. In Chapter 2, a sequential trading model with ambiguity aversion is studied. Traders with and without ambiguity trade an asset in sequence with a market maker updating the expected value of the asset according to a history of actions and their private signal. Ambiguity is only assumed on the precision of the private signal of traders about the value of the asset. The ambiguity averse traders update the set of precision with the recursive multiple prior preference, while the traders without ambiguity and the market maker update the distribution of the precision from uniform distribution as a prior. We find that the behaviour of the ambiguity averse traders differs depending on their updating rule on the set of precision. With the Full Bayesian Updating, herding or contrarian behaviour never occurs while the traders choose no trade in equilibrium even with the informative private signal asymptotically. With the Maximum Likelihood Updating, herding can happen, while no trade is less likely to be chosen by the traders in equilibrium, and they tend to act following their private signals like traders without ambiguity as trade goes on in the limit. In Chapter 3, a sequential trading model is introduced to estimate ambiguity aversion in the financial market. As with the model in Chapter 2, it has traders with and without ambiguity and they trade an asset with a market maker in sequence. Also, the ambiguity is only on the precision of the private signal received by the traders, and the ambiguity averse traders have recursive multiple prior preference. In addition to that, there is an event uncertainty and the private signal of the traders is continuous. In the model, traders choose not to trade with any signal, or show herding or contrarian behaviour depending on their updating rule on the set of priors about the precision. We estimate the model with trading data on NYSE stocks. The estimation result shows that strict reevaluation generates more herding or contrarian, while loose reevaluation brings no buying or selling more likely. In Chapter 4, we consider a sequential trading market with potential manipulator. As in the previous chapters, traders exchange an asset with a market maker. In the market, there are three types of traders: noise traders, informed traders and a potential manipulator. The manipulator receives private information on the asset value like other informed traders, but she has the opportunity to trade twice in a trading day, while the other traders trade only once. We show that, under some conditions, the manipulator does not follow her signal in the first period of action in equilibrium. Instead, she trades against her signal, suffering a loss to distort the expectation of other rational traders and the market maker. The price path is manipulated with the distortion. The manipulator then re-enters the market and makes profits by trading with the price. In this process, the uncertainty is essential as it generates a gap in expectation between the informed traders and the market maker and enables the distortion.

Download activity - last month

Export as XMLCSVJSON

Download activity - last 12 months

Export as JSONXMLCSV

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item