Potential yield is another important caveat. The U.S. all cotton yield of 798 pounds per acre in 2006 strikes me as remarkable given the widespread hot and dry growing conditions.
It is, however, in keeping with a strong up trend in yields in recent years, presumably from improved varieties, technology, and boll weevil eradication. So given "normal" growing conditions (i.e., when it rains a little in June and July) the U.S. could have regular over supplies of cotton unless planted acreage shrinks accordingly.
I am approaching this by assuming the ten year low for planted acres (13.4 million) and only a two year average yield (i.e., equally weighting the exceptional growing conditions in 2005 with the harsh conditions of 2006).
What would it take to have a reduction in U.S. stocks-to-use and higher prices? From the supply standpoint, it will require historically low planted acres and bad growing conditions as in 2006. Both of which are possible, but I have an easier time believing that 1) capital constraints and farm bill influences will keep U.S. planted cotton acres well above 13.5 million, and 2) 2007 yields will be near or above 800 lbs/acre, as in the last three years.