Bootstrapping a time series model
The bootstrap is a methodology for estimating standard errors. The idea is to use a Monte Carlo simulation experiment based on a nonparametric estimate of the error distribution. The main objective of this dissertation was to demonstrate the use of the bootstrap to attach standard errors to coefficient estimates and multi-period forecasts in a second-order autoregressive model fitted by least squares and maximum likelihood estimation. A secondary objective of this article was to present the bootstrap in the context of two econometric equations describing the unemployment rate and individual income tax in the state of Oklahoma. As it turns out, the conventional asymptotic formulae (both the least squares and maximum likelihood estimates) for estimating standard errors appear to overestimate the true standard errors. But there are two problems: 1) the first two observations y/sub 1/ and y/sub 2/ have been fixed, and 2) the residuals have not been inflated. After these two factors are considered in the trial and bootstrap experiment, both the conventional maximum likelihood and bootstrap estimates of the standard errors appear to be performing quite well. At present, there does not seem to be a good rule of thumb for deciding when the conventional asymptotic formulae will give acceptable results.
- Research Organization:
- Oklahoma State Univ., Stillwater (USA)
- OSTI ID:
- 5596975
- Resource Relation:
- Other Information: Thesis (Ph. D.)
- Country of Publication:
- United States
- Language:
- English
Similar Records
Bootstrapping an econometric model: some empirical results
Analyzing the Impact of Residential Building Attributes, Demographic and Behavioral Factors on Natural Gas Usage