Volatility and dependence structure are two main sources of uncertainty in many economic
issues, such as exchange rates, future prices and agricultural product prices etc. who fully
embody uncertainty among relationship and variation. This paper aims at estimating the
dependency between the percentage changes of the agricultural price and agricultural production
indices of Thailand and also their conditional volatilities using copula-based GARCH
models. Themotivation of this paper is twofold. First, the strategic department of agriculture
of Thailand would like to have reliable empirical models for the dependency and volatilities
for use in policy strategy. Second, this paper provides less restrictivemodels for dependency
and the conditional volatility GARCH. The copula-based multivariate analysis used in this
paper nested the traditional multivariate as a special case (Tae-Hwy and Xiangdong, 2009)
[13]. Appropriate marginal distributions for both, the percentage changes of the agricultural
price and agricultural production indices were selected for their estimation. Static as well
as time varying copulas were estimated. The empirical results were found that the suitable
margins were skew t distribution and the time varying copula i.e., the time varying rotate
Joe copula (270
◦
) was the choice for the policy makers to follow. The one-period ahead
forecasted-growth rate of agricultural price index conditional on growth rate of agricultural
production index was also provided as an example of forecasting it using the resulted
margins and time-varying copula based GARCH model.