Quarterly U.S. feeder cattle and fed cattle prices were estimated within a rational distributed lag framework, utilizing nonstochastic difference equations to minimize problems with specification errors in the disturbance structure. The behavior of certain input and output variables in the beef market explained the price variation in each reduced form equation. The results revealed dynamic stability, with a geometric rational lag structure describing fed cattle prices and a polynomial rational lag structure characterizing feeder cattle prices. These dynamics were found to be better than both static‐serial correlation and purely autoregressive specifications.
Loading...
John M. Marsh (1983) studied this question.