The soft commodity is a high-frequency trading market. We use the right-tailed unit root tests of Phillips et al. to investigate asset bubbles within and to detect explosive episodes on each soft commodity sample date set. Our empirical test uses weekly data from New York soft commodity market and finds bubbles existed in each of the soft commodities. The test model allows us to exam each bubble’s origination and termination dates using date-stamping in the recursive procedure. We found that the soft commodities bubble is more relevant to the man-made factor.
