Integrating modelling techniques for financial time series
The use of a related series of historical values as evidence or a guide line to predict and to draw conclusions about future events is a practice common to researchers of many disciplines. In order to achieve a ' better' outcome, researchers are motivated to search for rules that can explain relationships between the past and future. However, any forecast is affected by the unpredictable nature of future events as well as by the limitations of past data. Moreover, a forecasting model used by one field may not necessarily be used by another, because there are problems which are unique to each field. In standard statistical treatment of time series, time domain and frequency domain are the classical techniques used as a basis for characterising an observed system and forecasting its future behaviour. However, due to complicated data structures of time series, recent research developments--along with the availability of high-speed computers--have facilitated the development of modern nonparametric methods such as the 'moving blocks bootstrap ' and the 'neural networks'. The inadequacy of the efficient market hypothesis and the instability of data structures found in financial market create serious challenges to the implementation of modelling strategies for financial time series. However, a combination of both classical techniques and modern nonparametric methods may offer a more effective solution to financial forecasting problems. This thesis will implement a financial time series model which utilizes both classical techniques and nonparametric methods for estimation and inference. It will go on to propose a modified bivariate transfer function model. Both time domain and frequency domain techniques will be used in order to determine a relationship between bivariate time series. After generating the ' less model dependent' samples using the moving blocks bootstrap, a neural network will be applied to achieve better point estimation. Integrating these techniques provides the best alternative solution to financial forecasting problems. The bivariate time series in question is interest rate spreads and the spot Canadian dollar. Simulation results showed undercoverage overall, but the integrated modelling technique appears robust to the choice of noise distribution as well as the sample size.