October 6, 2003
SOYBEAN PRICES: WHAT NEXT?
November 2003 soybean futures were trading near $5.20 before the USDA's August production forecast of 2.862 billion bushels, moved to near $6.00 in front of the September forecast of 2.643 billion bushels, and to $7.00 last week. The higher soybean prices since mid-September were fueled mostly by reports of disappointing yields. In addition, a rapid pace of export sales and some early talk of dry weather in parts of South America contributed to the price rally.
Reports of very low yields from early maturing soybeans have been replaced by some reports of more normal yields in later maturing soybeans. Eastern and southern areas may have near normal yields and western areas are likely to have below normal yields, but that is in line with the USDA September forecast. Does the market now accurately reflect production potential? Or has there been a significant under or over reaction to early yield reports? The USDA's October production forecast has the potential to move prices significantly in either direction.
In recent years, there have been some large differences between the USDA's September and October soybean production forecasts. Large declines in October (3 to 5 percent) occurred in 1995, 1998, 1999, and 2000. The declines in 1999 and 2000 followed declines in September, while the smaller October forecasts in 1995 and 1998 followed larger forecasts in September. This year, the market obviously expects a smaller forecast in October. Seemingly, the only question to be answered is how much lower?
Private estimates released last week reflected declines of 25 to 80 million bushels. Anticipating the change in the USDA production forecast this year is complicated by the announcement that the USDA will use administrative data (primarily from the FSA) to adjust acreage in the October report. This data has traditionally not been available until December. In addition, USDA will ask producers in the Missouri survey to update harvested acreage estimates.
The worries about U.S. crop size have been compounded a bit by the rapid pace of export sales. As of September 25, the USDA reported that 373 million bushels of U.S. soybeans had been exported or sold for export during the 2003-04 marketing year. Commitments were 35 percent larger than on the same date last year, while the USDA has projected a 10 percent drop in exports for the year. Both the European Union and China have purchased more U.S. soybean than at this time last year. The current pace of export sales (seasonally adjusted) cannot be maintained, even with further declines in the size of the domestic crush. Current prices, along with prospects for a large South American harvest in 2004, may be high enough to trim the rate of export sales. However, the magnitude of the needed cut in exports is still not known.
After the U.S. harvest is completed and the forecast size of the crop is updated in November, attention will focus more intently on 2004 South American production prospects. The USDA has forecast a 7 percent increase in South American soybean acreage and a 5.8 percent increase in production. The recent soybean price rally may lead to a larger increase in area, but the focus will be on yield prospects. Average yields in Brazil and Argentina have been trending higher, but year over year declines have occurred. Over the past 14 years, the average yield in Brazil was below that of the previous year four times. The declines ranged from 0.4 to 8 percent. Year over year declines in Argentina occurred in 6 of the past 14 years. The declines ranged from 0.9 to 13 percent. By comparison over the past 14 years, the average U.S. yield was lower than the previous year 4 times, in a range of 4 to 15 percent. The September forecast for 2003 was for a 2.7 percent decline. In general, it appears that yield declines in poor years are less severe in South America, particularly in Brazil, than in the U.S.
Since mid-September, soybean prices have increased a little more rapidly than product prices, soybean oil prices have been stronger than meal prices, basis in many areas has strengthened, and the inverse in the futures market (beyond March 2004) has increased. This pattern of price behavior is very typical of "short crop" years. In the short crop years of 1980, 1983, and 1988, the highest cash price of the marketing year occurred in September twice and November once. However, in the yield declining years of 1993, 1995, and 1999, the highest cash price occurred well after harvest - May once and July twice. The magnitude of the production decline this year, as revealed in the October and November reports, should provide more insight on whether or not to expect an early price peak.
Issued by Darrel Good
University of Illinois