Each member of a packer team was assigned to a feedlot and instructed to act as a
regional buyer, just like in real cattle procurement markets. This was intended to allow
enough time for packers to inspect and submit bids for cattle among spatially dispersed
feedlots. Each trading period lasted about ten minutes and was called a “week.” The
winner of each auction was the packer who submitted the highest bid.
During the trading period, paper cards representing completed trades are returned
to the instructors who scanned them into a computer. The information on each card from
a completed trade includes the price and quality of cattle sold, and identity of the seller
and the buyer. This information is summarized for market participants before the next
trading period. Thus, feedlots and meatpacking managers are informed about the volume
of cattle trade, cattle placed on feed, and the wholesale price of processed beef in the
previous trading period.
A total of 1,788 transaction data were collected during fourteen trading weeks,
after allowing for a training period of two weeks. After the first seven weeks of cattle
trades, two mergers were simulated. Packer one merged with Packer two, and Packer
three merged with Packer four. These mergers represented the smallest packers (1 and 2)
and the largest packers (3 and 4). Overall, the structure of the game remained essentially
the same after the mergers except that there were two bigger packers instead of four
smaller ones. Descriptive statistics of the variables used in the analysis are reported in
table 1.
As reported in table 1, the average cattle price after adjusting for dressing
percentage (121$/cwt) is greater than the price of beef (119$/cwt). Further, the spread
between boxed beef price and the dressed cattle price is negative in 536 out of 1066
18