Evolutionary algorithms attempt to find solutions to problems where the solution space is too large to be examined in its entirety. These algorithms have been used in applications ranging from Biology to Economics. Genetic Algorithms and Genetic Programming are variants of evolutionary computation that can be applied to problems in Accounting and Finance. This dissertation evaluates the applicability of Genetic Programming to option pricing and time-series modeling and the applicability of Genetic Algorithms to financial statement analysis The Black-Scholes model is a landmark in option pricing theory and has found wide acceptance in financial markets. The search for a better option pricing model continues, however, as the Black-Scholes model was derived under strict assumptions that do not hold in the real world and model prices exhibit systematic biases from observed option prices. I successfully apply Koza's (1992) Genetic Programming methodology to develop option-pricing models. This method is well suited to the task and offers some advantages over alternative methods There is often a need in accounting research for a proxy of the markets' expectation of earnings. Quarterly earnings forecasts from both analysts and mechanical models have been traditionally used as such proxies. Mechanical models, although easy to implement, have unfortunately never been able to consistently beat analysts' forecasts, of which Value Line (VL) is an example. I use Genetic Programming to develop forecasting and forecast combining models for the time-series of quarterly earnings. However, the models developed using this technique fail to perform better than traditional linear models One of the goals of financial statement analysis is to extract firm-value-relevant information from financial statements. The process by which this information is processed can be considered a black box. In making forecasts and reports, analysts examine financial statement variables and derived quantities and aggregate this information with outside information. The process is subjective and it is a stylized fact that some information is always purposely or by necessity left out. It is humanly impossible to coherently aggregate information from so many variables. Therefore, a methodology in which information is extracted automatically is potentially attractive not just to analysts but also to lay investors. Ou and Penman (1989) propose one such automated methodology. They develop what they term the 'Pr measure' to aggregate financial statement information and predict the signs of changes in annual company earnings adjusted for drift. However, Ou and Penman's methodology may have some weaknesses. I develop predictive models that ameliorate some of the potential weaknesses in Ou and Penman's method. My methodology combines genetic algorithms and LOGIT to predict the signs of earnings changes / acase@tulane.edu
| Identifer | oai:union.ndltd.org:TULANE/oai:http://digitallibrary.tulane.edu/:tulane_23715 |
| Date | January 2000 |
| Contributors | Trigueros, Joaquin Rafael (Author), Jain, Prem C (Thesis advisor) |
| Publisher | Tulane University |
| Source Sets | Tulane University |
| Language | English |
| Detected Language | English |
| Rights | Access requires a license to the Dissertations and Theses (ProQuest) database., Copyright is in accordance with U.S. Copyright law |
Page generated in 0.0071 seconds