Summer quarter feed and residual use of wheat - quarterly pattern of wheat feed and residual use over period of several years indicates that 1991 summer quarter disappearance of wheat may be record high - U.S. Dept. of Agriculture, Economic Research Service report

Abstract: The quarterly pattern of wheat feed and residual use in recent y ears

shows this use tends to be large and positive in the summer quarter. Equations

were developed to estimate summer-quarter wheat feed and residual use.

Various specifications were examined. Using an assumed range of wheat and

corn prices, the equations imply a 1991 summer-quarter feed and residual

disappearance of between 450-475 million bushels -- which if realized, would be

record high.

Keywords: feed and residual use, summer quarter, prices In regions where wheat and livestock production is concentrated, as in much of the Southwest and Southern Plains, it is quite common to use wheat, especially lower quality wheat, in feed rations because of the high cost of transporting feed grains into the region. In 1990/91, low wheat prices relative to corn/sorghum prices encouraged large-scale wheat feeding, not only in regions that traditionally feed wheat, but across much of the country. The quarterly pattern of feed and residual use in recent years shows this use tends to be large and positive in the first quarter of the marketing year, negative in the second quarter, and either positive or negative in the second half of the year. In 1990/91, low wheat prices favored large wheat feeding and the implied feed and residual disappearance was a record 407 million bushels in the summer quarter. This equaled the previous record annual total in 1984/85. With wheat prices remaining low in the following months, feed and residual use was large and positive in the third quarter of 1990/91. The feed and residual use category for wheat includes all nonfood and nonseed domestic uses, including the statistical residuals (errors) from all categories of wheat supply and demand. It is possible to have large negative numbers in the quarterly feed and residual use estimates.

Why Focus on the Summer Quarter?

Most feed use of wheat occurs during the summer quarter (June-August). Winter wheat is largely harvested during June and July and, as farmers market their crop, prices are generally at their seasonal low for the marketing year. In addition, wheat with quality problems is likely to be marketed or fed quickly instead of being put into storage for later sale. Since neither domestic mills nor exporters are interested in this type of wheat, it usually moves into feeding channels or is used on the farm for feed. Most U.S. feed grains are not harvested until fall, so during the summer, their prices have not yet fallen to harvesttime lows. The price advantage of wheat relative to feed grains is even more pronounced during the summer in many major wheat producing areas because corn has to be shipped in from the corn belt, while wheat is local. This combination of seasonally low wheat prices and seasonally higher corn and sorghum prices often makes it attractive to use wheat in feed rations. The other quarters were not included in this analysis. While no one knows what actual feed use of wheat is, it appears to be smaller during the last 3 quarters of the marketing year. Thus residual considerations, possibly measurement quirks in the supply and demand estimates, often dominate during these quarters.

Price Ratios Are the Key Variables

Demand for an input to a production process is related to the price of the input, the price of competing inputs, and the price of the product. This basic framework was used to estimate the demand for wheat used in livestock production. Wheat prices represent the input price, while corn prices are used to represent the prices of competing feed grains. Sorghum and other feed grain prices tend to follow corn prices. For both wheat and corn, average prices received by farmers were used. An index of prices received by farmers for livestock and products (as published in Agricultural Prices) was used as a measure of prices in that sector, and represents the value of output that the feed is used to produce. Prices by themselves do not fully explain wheat feeding. Non-price factors also can affect feeding, especially the number of animals to be fed and the size of the wheat supply. Accordingly, several other variables, with varying success, were used with the price variables in an attempt to estimate summer wheat feeding. The most interesting included: Southern Plains cattle on feed, Arkansas broilers, and the wheat supply size. The size of the wheat supply is an important variable, not only because of its influence on wheat feeding, but also because it may affect residual disappearance. For example, losses in handling and storage are likely to be larger if the supply is large. A time trend was included in some specifications to represent growth in the scale of the sector.

Model Specification and Estimation

The ratio of the livestock price index divided by the wheat price was used as a variable to capture the attractiveness of feeding wheat to animals, relative to other uses. This variable is expected to be positive as wheat feeding is most likely to occur when livestock prices are high and wheat prices are low. The ratio of the livestock price divided by the corn price was used to represent the attractiveness of feeding alternative grains instead of wheat. This variable is expected to have a negative sign, because low corn prices in the denominator would encourage use of feed grains in the feed ration.

Other Variables Also Crucial

The regression using the price variables and a simple time trend is shown in equation 1. All three variables were statistically significant and had the expected signs. Although the time trend is indicating the general growth in the size of the livestock and the wheat sectors, it does not identify specific developments. A Southern Plains cattle-on-feed variable was developed by adding the July 1 inventory of Texas, Oklahoma, Kansas, and Colorado. This region is thought to feed more wheat than other regions because it is a major wheat-producing area with large cattle-feeding operations. Recognizing that some wheat, especially soft red winter, is fed to poultry, a poultry variable was used to capture that effect. Annual broiler production in Arkansas was chosen because that State is a large broiler and wheat producer. The size of wheat supplies was measured in two ways - total beginning stocks plus winter wheat production, and beginning stocks plus total production. Both worked well with the price, cattle on feed, and poultry variables. Equation 2, using stocks plus winter wheat production, produced a slightly higher R-squared (0.973), but resulted in a lower t-statistic for the cattle on feed variable. Equation 3, using stocks plus total wheat production, had nearly the same explanatory power, with an R-squared of 0.969, and the t-statistic of the cattle on feed variable was higher. In both equations all variables had the expected signs. Time trend was not used in equations 2 or 3. Poultry numbers have increased much like a time trend, so putting them both in an equation would result in multicollinearity.

Forecasts Using Equations

The wheat/corn breakeven price relationship for feed use, with both expressed in dollars per bushel, is generally thought to range up to a 12-percent price premium for wheat, depending on what type of animal is being fed. This is because the standard wheat bushel is 60 pounds, while a corn bushel is only 56 pounds. Moreover, wheat has higher protein content per pound, making it attractive in some rations. Data on June and mid-month July prices for corn, wheat, and the livestock and products index, as well as cattle on feed numbers are available. For this analysis, the summer quarter prices received by farmers are assumed to be $2.55-2.60 for wheat, $2.30-2.35 for corn, and a 163 index for livestock prices. July 1, 1991, cattle on feed in the Southern Plains was reported at 4.9 million head, up 9 percent. Arkansas broilers are assumed to be up about 5 percent from 1990 levels. Wheat supply forecasts are available, with increasing stocks more than offset by lower production. Using the assumed range of wheat and corn prices, the equations imply a summer-quarter feed and residual disappearance of between 450-475 million bushels -- record high.

Table A-2, Actual and Estimated Feed and Residual Use

     Historical   Equation 1   Equation 2   Equation 3
year                    mil. bu.
1977   117.1      110.1        96.2      90.9
1978    80.8       79.5        79.9      86.4
1979    38.3       51.1        74.4      76.7
1980    48.1       78.7        54.5      48.3
1981   144.9      135.4       138.1     143.0
1982   131.3      158.7       143.4     149.4
1983   196.1      185.1       189.5     182.4
1984   279.6      226.4       256.6     259.1
1985   235.5      242.0       253.4     258.6
1986   352.3      354.2       348.0     343.7
1987   363.8      327.9       358.3     355.7
1988   282.2      274.6       295.5     282.3
1989   263.8      265.7       224.1     229.6
1990   407.4      451.5       429.0     435.1

[Figures 1a Omitted] [Tabular Data A-1 Omitted]

COPYRIGHT 1991 U.S. Department of Agriculture
COPYRIGHT 2004 Gale Group