Essays in Financial Economics and Computational Finance

This thesis contains three chapters. In particular, it examines the opaqueness of financial securities, risk contagion among asset markets and learning behaviours and trading skills in financial markets. Chapter 1, Opacity in Security Design: The Role of Derivative Markets, builds a theoretical model to explain the contribution of financial derivatives, especially Credit Default Swap, to the opaqueness of financial securities. We find a trade-off in the opacity design of financial securities: opacity increases adverse selection in the primary market while gives the originator information advantage in the derivative market, the former decreasing the originator's profits while the latter increasing his profits. Hence, the optimal opacity involves a balance between the two effects, implying derivative markets do play a role in deciding the opacity of financial securities. In addition, a liquid derivative market tend to induce more opacity as the originator makes more profits on the derivative market, which makes him more willing to sacrifice his profits in the primary market and increase opacity. Chapter 2, \Statistical Arbitrage and Risk Contagion", builds a computational model to investigate the risk contagion mechanism provided by statistical arbitrageurs among asset markets. Statistical arbitrageurs arbitrage mispricings but also follow a market-neutral rule. We find that statistical arbitrageurs help stabilise markets in normal periods. However, they may also act as the mechanism for the spread of shocks, making the whole system expose to extreme events. While statistical arbitrageurs may play a role in risk contagion, we find the effects are limited. The reason is that statistical arbitrageurs in markets not shocked trade against shocks, which helps the system to recover fast. In addition, statistical arbitrageurs persistently make positive profits from trading, consistent with the fact that financial institutions heavily rely on this strategy. Chapter 3, The Role of Heterogeneous Beliefs in Trading Skill Acquisition", also creates a computational model to examine how chartists, who are effectively noise traders, affect fundamentalists' learning and trading skill acquisition, where fundamentalists do not have the skill to fairly price financial assets and learn to do so from trading. We find that the presence of chartists can facilitate the learning of fundamentalists by stabilising the market price. However, chartists tend to reduce the accuracy of the learning outcome by misleading the market price. Hence, an increase in the amount of chartists increases the proportion of fundamentalists holding trading skills, which makes markets more resilient. However, the less accurate learning outcome of fundamentalists tend to increase market volatility.