Asymptotic inference about predictive ability
Article Abstract:
An evaluation is conducted on the predictive ability of various nonnested and nonlinear models, based on an asymptotic inference on the moments of smooth functions of predictions and out-of-sample prediction errors. The use asymptotic tests are considered better than many of the simulation methods to test the effectivity of various predictive models since such tests need not require the formulation of a null model from which data for evaluation are generated. The test uses kernel techniques to test for unknown forms of serial correlation and heteroskedasticity in predictive errors.
Publication Name: Econometrica
Subject: Mathematics
ISSN: 0012-9682
Year: 1996
User Contributions:
Comment about this article or add new information about this topic:
Risk vulnerability and the tempering effect of background risk
Article Abstract:
Improvements are made on the risk aversion model, wherein the sensitivity of the model towards the interaction of independent risk variables and its subsequent effect on the risk taking behavior is accounted for. This was achieved by improving the risk vulnerability portion of the model, where an absolute condition was formulated which makes absolute risk aversion decrease with any increase in the interaction of independent risk variables. This enables an optimization of the risk-taking behavior, by making the risk behavior responsive to risk factors.
Publication Name: Econometrica
Subject: Mathematics
ISSN: 0012-9682
Year: 1996
User Contributions:
Comment about this article or add new information about this topic:
- Abstracts: Existence and regularity of equilibria in a general equilibrium model with private provision of a public good
- Abstracts: Asymptotic normal and bootstrap inference in structural VAR analysis. A re-evaluation of the quasi-Bayes approach to the linear combination of forecasts
- Abstracts: Optimal tests for parameter instability in the generalized method of moments framework. Monitoring structural change