which typically are assumed to be the same for each random variable, but not in the copula approach;
(3) the analytical approach used in this paper could find the dependency and hence the correlation without assuming linear
correlation, and as a result we could obtain the transformation invariant correlations, especially where variances do not
exist (for some heavy-tailed distributions); (4) this paper was able to find the marginal effects of the random variable in
correlation with other random variables without imposing the conventional strong assumption of some distributions of the
random variables through the conditional density function; (5) because agricultural price and production indices are by
their natures correlated, the revenue distribution could not be found by the product of price and production indices. This
paper can then pave theway for obtaining the revenue index distribution; (6) this paper used a newapproach for conditional
forecasting, which focuses on what will happen to the price under the given output; (7) we put forward the model of time
varying (rotate) Joe copula for analyzing dynamic Kendall’s tau and policy implication; (8) finally, an important motivation
for this paper is the lack of any previous study in agricultural economics, and especially in Thailand, that would use the
copula approach to predict the agricultural price as a function of agricultural production.
3. Econometric