|
|
 |
| |
|
|
|
Comparative Bayesian Analysis of GARCH and Stochastic Volatility Models using R and Stan |
|
|
|
PP: 1489-1504 |
|
|
|
Author(s) |
|
|
|
Muhammed Navas T,
Mosab I. Tabash,
Shazia Farhin,
Athar Ali Khan,
|
|
|
|
Abstract |
|
|
| This study uses modelling and model comparison to compare three widely used GARCH models with their stochastic volatility (SV) counterparts in modelling the dynamics of inflation rates using the Bayesian approach. BRICS country consumer price index (CPI) data are used to assess these models. We find that the stochastic volatility models perform better than the GARCH models most of the time. The stochastic volatility in the leverage (SV-L) model is also demonstrated to be the most effective for the BRICS nations that we took into consideration. The article also looks at which model attributes are crucial in simulating inflation rates. It turns out that when modelling inflation rates, inflation volatility feedback is an important component to take into account. For each of the five countries we took into consideration, SV-L outperforms all other models. The study was done in rstan, a programming language for statistical inference, and the simulation uses the Hamiltonian Monte Carlo (HMC) algorithm of the Markov chain Monte Carlo (MCMC) to sample from the posterior distribution.
|
|
|
|
|
|
|
|
