Wheat: Forecasting season-average wheat prices using futures prices – formula and procedure for forecasting wheat prices

Forecasting season-average wheat prices using futures prices – formula and procedure for forecasting wheat prices – U.S. Dept. of Agriculture, Economic Research Service Report

Linwood A. Hoffman

Abstract: A method is developed which uses futures prices to forecast the

season-average price of U.S. wheat. A historical monthly average basis is computed

and deducted from the nearby futures price resulting in a monthly farm-price

forecast for each month in a crop year. Next, a weighted-season-average price is

computed. Results provide timely and reasonably accurate forecasts of season-average

producer prices for wheat.

Keywords: Basis, wheat, forecasts, futures prices, futures-method forecast,

season-average prices.


Commodity price forecasting is an important and ongoing activity conducted by both the private and public sectors. Forecasting methods range from sophisticated econometric models to expert qualitative judgement. Policymakers constantly seek to understand the effects of domestic or international events upon producers’ season-average prices. Producers’ price expectations influence planting decisions, which, in turn, affect harvested supplies and market prices. Thus, commodity price forecasts are important to taxpayers, producers, and consumers. A short-run change in farm prices depends upon numerous factors that affect commodity supply and demand functions. Estimates of commodity prices should be based on expected supply and demand conditions. While some have questioned the impact of technical traders on the futures market, futures prices are still considered as a composite indicator of expected supply and use and, thus, can be used to forecast short-run farm prices (1, 2, 3, and 4). Futures prices are determined by the interaction of current and expected demand for, and supply of, a commodity. Hedgers and speculators evaluate a number of factors, including, but not limited to planting intentions, weather factors, government policies, and the potential for domestic or export consumption. Hedgers deal with the actual commodity, as well as with futures contracts. Frequently, speculators have no-direct connection to the cash commodity, but expect to profit from changes in futures prices. Current futures prices provide important information about expected cash prices on future dates. However, most participants in the futures market need to be able to forecast a price at the location and time when they plan to buy or sell. Thus, they need to predict the “basis,” the difference between the futures price and their local price. Similarly, in making decisions about farm programs, policymakers benefit from accurate forecasts of a national-average farm price. This article describes the methodology used in forecasting monthly and season-average prices. Then, weekly updates of season-average price forecasts are presented for the 1991/92 crop year. Forecast accuracy results are presented for previous crop years. To assess forecast accuracy, forecasts are compared with actual season-average prices, and an alternative published forecast. The alternative forecast used in this article is the U.S. Department of Agriculture’s (USDA) season-average price forecast, released in the World Agricultural Supply and Demand Estimates (WASDE).

Forecasting Method

Forecasts are made of the monthly average cash wheat price received by farmers for each of the 12 months of the crop year, starting with June. Each month’s forecast is based on the current futures price for the nearest contract maturing after the month being forecast (referred to as the “nearby futures contract”). The forecast for each month is obtained by adding a historical average-price-difference “basis” (cash price minus futures price) to the nearby futures price. Monthly price forecasts are then weighted by a historical percentage of sales by month to calculate the weighted season-average price forecast. Relationships within the forecast method are expressed as: (1) [P.sub.m] = [F.sub.mt] + [b.sub.m] where: [P.sub.m] = Forecast U.S. farm price of wheat in month m for 12 months, June through May. Thus, this method provides a short-term forecast based on the availability of futures contract prices. [F.sub.mt] = Futures settlement price of wheat observed on day t of the first contract to mature after month m. Each crop year contains five futures contracts: July, September, December, March, and May. [b.sub.m] = Expected basis, in month m, equals the U.S. farm price less the price of the nearby futures contract for wheat averaged for month m over the previous 5 years. The forecast of the weighted season-average price (SAP) is computed as:

12 (2) SAP = [Sigma] [w.sub.m] [P.sub.m]

m = 1 where: [w.sub.m] = monthly weight for month m. [P.sub.i] = the average actual farm price for past months and/or ([F.sub.mt] + [b.sub.m]) for future months.


As previously mentioned, the difference between a cash price at a specific location and the price of a particular futures contract is known as the basis. The basis tends to be more stable or predictable than either the cash price or futures price. Several factors explain the basis and their magnitude varies from one location to another. Some specific factors that determine the basis include: local supply and demand conditions for the commodity and its substitutes, transportation and handling charges, transportation bottlenecks, availability of storage space, storage costs, conditioning capacities, and market expectations. Because the basis calculated for this analysis represents an average of U.S. conditions, it reflects a composite of these influencing factors. The basis in this study is the arithmetic difference between the monthly U.S. average cash price received by producers and the nearby futures settlement price. For example, the June basis is the difference between the June-average cash price received by producers and June’s average settlement price of the July futures contract. A 5-year moving average of these bases is used to reduce distortions that may occur in any given month and is updated at the end of each crop year.

Monthly Weights

Monthly marketings are used to construct the weighted season-average price. Each month’s weight represents the proportion of the year’s crop marketed in that month. A 5-year moving average of these monthly weights is constructed (1986/87 through 1990/91) and is updated annually after the release of USDA’s December issue of Crop Production. The monthly prices, actual or forecast, are multiplied by each month’s corresponding weight.


Historical daily settlement prices are obtained from the Commodity Futures Trading Commission (crop years 1981-89) of each wheat futures contract traded on the Kansas City Board of Trade. Current futures settlement prices are from the Wall Street Journal (crop year 1990 and 1991). Cash prices are from Agricultural Prices, published by USDA’s National Agricultural Statistics Service. Weights for monthly marketings are from various issues of USDA’s December Crop Production.


Table A-1 illustrates the method used in forecasting the season-average wheat price for the crop year 1991/92. This method produces a weekly forecast of the season-average price. A weekly futures settlement price (as observed on each Thursday) is used for each of the nearby contracts. Alternatively, a daily or monthly forecast of the season-average price could be made. Six steps are involved in the forecast process.

1. The latest available futures

settlement prices (line 1) are gathered for

the contracts that are trading.

Settlement prices for Thursday,

February 13, 1992, are used for

illustration (line 1). Futures quotes are

used for March, May, and July

1992 contract settlement prices.

Actual monthly prices received are

available and used for June 1991

through January 1992. (The

January monthly cash price represents a

mid-month price and is updated the

following month.)

If this forecast were started in May

1991 (concurrent with the start off

USDA’s price forecasts for crop

year 1991/92), the July 1992

futures price would not be available.

Thus, a 5-year-average spread

between the May and July contracts

would be used to compute an

implied July 1992 futures price.

Alternatively, if the forecast was

started in June 1991 (the beginning

of the crop year), all futures prices

needed (July 1991 through July

1992) would be available and

entered on line 1.

2. Monthly futures prices are the

settlement prices of the nearby

contracts. For example, the futures

price for February 1992 (line 2)

represents the February 13

settlement price of the March 1992

contract. The nearby (May) contract

price is used for March because

during any contract close month the

nearby contract has greater stability

than the contract-close month

(March), as contract liquidity

decreases during the delivery month.

Also, the contracts usually close

about the third week of the month

which would lower the number of

observations that could be used to

calculate the average monthly

closing price.

3. A forecast of the monthly average

farm price (line 4) is computed by

adding the basis (cash price minus

futures price) (line 3) to the

monthly futures price (line 2).

4. The actual monthly average farm

price is entered on line 5 as it

becomes available. If this 1991-92

forecast was made during May or

June 1991, all 12 monthly prices

would be forecast and line 5 would

remain blank.

5. The actual and forecast farm prices

are spliced together in line 6. For

the present marketing year, 1991-92,

8 of the monthly prices shown

are actual farm prices of all wheat

(June through December), while

the last 4 monthly prices are


6. The monthly percentage of wheat

marketings by producers (5-year

moving average line 7) is used to

weight the monthly farm prices

(line 6). A weighted season-average

farm price of wheat is then

computed (line 8).

Forecast for 1991/92 Crop Year

Season-average price forecasts are based on expectations reflected in the futures market and, if available, actual farm prices. As of February 13, 1992, the futures method projection for the 1991/92 price of wheat was $3.11 per bushel (table A-1). The initial forecast was $2.75 per bushel, as of May 2, 1991, $0.14 per bushel above the season-average price for 1990/91 (figure A-1). Prices were expected to rise in 1991/92 because of reduced U.S. plantings and a below-average-yield forecast. However, price projections dropped in early July because of large beginning stocks, quality problems with soft red winter wheat, declining corn and soybean prices, and potentially large wheat exports from the major foreign competitors. However, since the middle of July, the futures price forecast has generally risen. The July 18, 1991, projection was $2.55 per bushel and it moved upward to about $3.11 per bushel on February 13, 1992, responding to tighter U.S. stocks and strong U.S. exports, lower than expected U.S. winter wheat plantings for the 1992 crop, and tightening world supplies.

Forecast Accuracy

Forecast accuracy was examined for crop years 1986/87 through 1990/91. A mean absolute percentage difference was computed for the crop year and a monthly percentage difference was computed between the monthly forecast and actual season-average-farm price. Accuracy of the futures method was also compared with the WASDE projections, an alternative season-average price forecast. Because the WASDE numbers are released monthly, the historical futures forecast was computed on a monthly basis. The midpoint of the WASDE forecast range is used to represent the WASDE forecast. The monthly futures projection uses the settlement price available the day after the release of the WASDE forecast. This procedure attempts to equalize information available to either method. The mean absolute percentage difference of the futures forecasts ranged from 2 to 4 percent over the past 5 years, compared with WASDE’s 3 to 5 percent (table A-2, figures A-2 to A-6). Based on the mean absolute percentage difference, the futures method performed about as well as the WASDE forecasts. Differences between the two forecasts were minor in each of the past 5 year’s projections, ranging from a low of 0.1 percent for the 1986-87 crop year to a maximum of 1.8 percent for the 1989/90 crop year. Based on the monthly percentage differences, the futures method performed better than the WASDE method in 3 out of 5 years. For example, the futures method had a lower percentage difference in 9 out of 13 monthly forecasts for 1986/87, 8 out of 13 for 1987/88, and 11 out of 13 for 1989/90. The WASDE forecast had the lower percentage difference in 9 out of 13 months in 1988/89 and in 12 out of 13 months in 1990/91.


This analysis suggests that the futures method can provide a timely and reasonable forecast of producers’ season-average prices. This procedure can provide a useful service to producer organizations, policy analysts, and consumer organizations. The futures forecast method can also provide a useful cross-check against other seasonage price forecasts. [Tabular Data A-1 to A-3 Omitted]


[1]Danthine, J. “Information, Futures Prices, and Stabilizing Speculation.” Journal of Economic Theory, 17(1978):79-98. [2]Gardner, Bruce L. “Futures Prices in Supply Analysis.” American Journal of Agricultural Economics, 58(1976):81-84. [3]Peck, Anne E. “Futures Markets, Supply Response, and Price Stability.” Quarterly Journal of Economics, 90(1976):407-23. [4]Rauser, G.C., and R.E. Just. “Agricultural Commodity Price Forecasting Accuracy: Futures Markets versus Commercial Econometric Models.” Futures Trading Seminar, Vol. 6. Chicago: Board of Trade of the City of Chicago, 1979, pp. 117-153. [5]U.S. Department of Agriculture, National Agricultural Statistics Service. Agricultural Prices. Annual Summaries. Various issues. [6]__. Crop Production. December issues, 1985-91. [7]U.S. Department of Agriculture, World Agricultural Outlook Board. World Agricultural Supply and Demand Estimates. Monthly issues, 1986-91. [8]The Wall Street Journal. Commodity Futures Prices. Various issues, 1990-91.

Linwood A. Hoffman Agricultural Economist, Commodity Economics Division Research Service, USDA.

COPYRIGHT 1992 U.S. Department of Agriculture

COPYRIGHT 2004 Gale Group