New possibilities for securities market research.

摘要

This dissertation describes a novel approach to data acquisition in securities market research that facilitates interdisciplinary research. This new approach involves writing data acquisition programs that automatically extract securities market data from financial websites. Data acquisition is presented first as an independent chapter, and then four subsequent chapters illustrate these data acquisition techniques. Each of these remaining chapters make original contributions to the field of securities market research, adding value to this thesis beyond the overriding theme of data acquisition.; A chapter on the profitability of closed-end fund trading strategies introduces a causality error present in previously published strategies that would greatly reduce their profitability if they were to be applied during 1998–2003. This study develops techniques to modify previous strategies to remove the causality error and demonstrates that these modified strategies are still significantly profitable.; A chapter from the field of options theory develops a new calculation technique for the extraction of the risk-neutral density from options prices. Strengths of this new technique are demonstrated relative to existing calculation methods. Particularly strong results are found for estimates of the density's tails, which are precisely where related calculations found in the literature become unstable.; A chapter from the field of volatility prediction develops a new model based on a relationship between volatility and momentum that is inferred from a principal components analysis of equity markets. The model is tested on both ultra-short and daily times scales, and shows promise as a volatility prediction tool.; A chapter on index arbitrage introduces a dynamical systems model that connects index arbitrage to volatility. This model demonstrates that index arbitrage opportunities can explain a large portion of short-term volatility (regardless of past volatility). Additionally, the model describes a procedure for predicting the level of short-term volatility that will result from a given arbitrage opportunity.