Measuring contagion with a bayesian time-varying coefficient model
- Rebucci, Alessandro
- Ciccarelli, Matteo
Année de publication: 2003
Número: 20
Type: Working Paper
Résumé
We propose to use a time-varying coefficient model to measure contagion. The proposed measure works in the joint presence of heteroskedasticity and omitted variables. It requires knowledge of the source of the crisis but not its timing. The estimation procedure is Bayesian and is based on Markov Chain Monte Carlo methods. We asses the performance of the proposed measure both with simulated and actual data.