Applying Time Series Analysis and Autoregressive Moving Average Model in Security Fraud
| Autor: | Lai, Cheng-Chieh, 賴正捷 |
|---|---|
| Rok vydání: | 2018 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 107 Event study plays an important role in securities fraud cases especially since Halliburton II. One of the core dispute in securities fraud is whether the security’s market price in question had been distorted by the alleged misstatement or omission, while an event study serves as an useful tool in analyzing certain impact on the aforementioned situation. Through the use of market model, event studies predict the security’s would-be price return based on the past market returns, which in turn provides evidence for materiality, loss causation and the calculation of damages. All are essential steps in deciding securities fraud cases. The use of regression analysis in market model is to calculate the price relationships between the security’s price return and market returns. Through these relationships, an event study predicts what the price would have been based on overall market. This method of price prediction eliminates the effect of firm specific misstatement or omission by the use of outside information as its variables. Time series is a sequence of observation taken through time. Data such as a security’s price usually exhibits some form of correlation across time. In this thesis, autoregressive moving average model is discussed to produce a better price prediction based on historical price value, by combining both linear regression and ARMA model, a model can contain movements of both the movement of overall market and historical price value. As a popular common method in model-building, this thesis expect the use of ARMA will be more important for price prediction in future securities fraud cases as a key tool in damage calculation. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|