Testing for parameter instability across different modeling frameworks
Autor: | Drew Creal, Siem Jan Koopman, Francesco Calvori, Andre Lucas |
---|---|
Přispěvatelé: | Finance, Econometrics and Operations Research, Tinbergen Institute, Econometrics and Data Science |
Jazyk: | angličtina |
Rok vydání: | 2017 |
Předmět: |
Observation-driven and parameterdriven models
Economics and Econometrics Autoregressive conditional heteroskedasticity 05 social sciences Variance (accounting) Instability Regime switching Nonlinear system Time-varying parameters Generalized autoregressive score model 0502 economics and business Econometrics Range (statistics) Economics Structural breaks Multiplier (economics) 050207 economics Constant (mathematics) Finance Credit risk 050205 econometrics |
Zdroj: | Calvori, F, Creal, D, Koopman, S J & Lucas, A 2017, ' Testing for parameter instability across different modeling frameworks ', Journal of Financial Econometrics, vol. 15, no. 2, pp. 223-246 . https://doi.org/10.1093/jjfinec/nbw008 Journal of Financial Econometrics, 15(2), 223-246. Oxford University Press Calvori, F, Creal, D, Koopman, S J M & Lucas, A 2017, ' Testing for Parameter Instability across Different Modeling Frameworks ', Journal of Financial Econometrics, vol. 15, no. 2, pp. 223-246 . https://doi.org/10.1093/jjfinec/nbw008 |
ISSN: | 1479-8409 |
Popis: | We develop a new parameter instability test that generalizes the seminal ARCHLagrange Multiplier test of Engle (1982) for a constant variance against the alternative of autoregressive conditional heteroskedasticity to settings with nonlinear timevarying parameters and non-Gaussian distributions. We investigate the performance of the new test relative to both classic and recently proposed parameter instability tests, including tests against structural breaks and parameter-driven dynamics. We find that the recent test of Müller and Petalas (2010) performs best across a wide range of alternatives, particularly if parameter instability is slow. For time-varying parameters that exhibitmoremean reversion, our new test has higher power. We provide an application to a heavily unbalanced panel of losses given default for US corporations from 1982 to 2010 and provide evidence of significant parameter instability in the parameters of a static beta distributedmodel. |
Databáze: | OpenAIRE |
Externí odkaz: |