Volatility Dynamics Under Duration-Dependent Mixing
Scholarly Editions
Overview
Overview
abstract
This paper proposes a new approach to modeling volatility changes
and clustering. In particular, we use a parsimonious high-order
Markov chain which allows for duration dependence. As in the
standard 1st-order Markov-switching model, this structure can
capture turning points and shifts in volatility due, for example,
to policy changes or news events. However, unlike the 1st-order
model, the duration-dependent Markov switching model is suited to
exploiting the persistence associated with volatility clustering.
To highlight the features of our model, we compare it to a popular
benchmark, the GARCH model. Unlike the latter, the proposed
parameterization allows time-varying persistence, includes a
stochastic component for volatility, and incorporates anticipated
discrete changes in the level of volatility. The empirical
distribution generated by our proposed structure works well for
the samples of data used in this paper. Implications for forecasts
relevant for risk management are emphasized.
