The Rising Influence of Rising Affluence

[DewDiligence]

Faulty USDA Forecasts Roil Corn Market

[As a long-term investor rather than a trader, the machinations described in this article don’t interest me much; nevertheless, it’s helpful to understand them to be ready to take advantage of temporary price weakness in such stocks as MON and DE caused by USDA forecasts.]

http://online.wsj.com/article/SB10001424052970203752604576641561657796544.html

›DECEMBER 5, 2011
By LIAM PLEVEN And TOM MCGINTY

Government reports about the U.S. corn crop have become increasingly unreliable of late, contributing to wild swings in corn prices, a Wall Street Journal analysis shows.

Over the past two years, the Department of Agriculture's monthly forecasts of how much farmers will harvest have been off the mark to a greater degree than any other two consecutive years in the last 15, according to a Journal analysis of government data. This year's early-season forecasts also appear to have been way off. The next monthly report is due on Friday.

At the same time, periodic stockpile reports—government estimates of how much corn is stored in farm silos and other storage facilities—have generated big surprises. The average monthly swings in stockpile estimates between May and October, the heart of the growing season, have been greater this year than in any year since 1996, according to the Journal analysis.

The stockpile reports have had a big effect on markets. On Sept. 30, the USDA said a quarterly survey showed corn stockpiles were 23% higher than it had estimated earlier that month. Corn prices fell 6.3% in the futures market that day, shaving $5 billion off the value of corn in the fields.

"There is very much a lack of confidence right now among farmers" in the government data, says Bill Christ, who harvests roughly 100,000 bushels of corn a year in Metamora, Ill. "Can't they get it right?"

USDA officials blame unpredictable weather for recent errant production forecasts. They say the figures are snapshots that change based on fresh information, such as damage caused by heat waves or changes in consumption patterns.

"If somebody's going to be in these commodity markets, they better understand these things are subject to change," says Gerald Bange, chairman of the USDA's World Agricultural Outlook Board, which is involved in producing the data.

The U.S. grew 38% of the world's corn in 2010, when the domestic crop was worth $67 billion. Strong demand from foreign buyers, and from the growing ethanol industry, has added to market volatility. Jerry Norton, who tracks corn for the USDA, says because corn supplies are tight, "the market is much more sensitive" to changes in the department's reports.

The Chicago Board of Trade has long had limits on single-day price moves on corn and other commodities in the futures market. Corn prices have hit the basic limit 20 times since the start of 2009, with eight of those instances, or 40% of them, coming on the day of a USDA report, according to a Journal analysis of price data. Between 1996 and 2008, only 20% of such limit moves came on reports days.

Like many farmers, Mike Yost, a former USDA official who is now a partner in a Murdock, Minn., dairy farm, pays close attention to the reports. In January 2010, the USDA forecast a record crop, causing corn prices to plummet. Mr. Yost figured a record crop would keep corn prices low, so he didn't lock in his feed prices.

Six months later, the department said corn stockpiles were smaller than he and many others had expected, despite the bountiful crop. Prices shot up, and Mr. Yost's dairy operation had to pay an extra $200,000 to buy feed. "We just got too comfortable going on the government numbers," he says. "There's something not quite right in their formula."

Critics of the reports contend that weather alone doesn't explain the erratic numbers, especially in the stockpile reports.

Darrel Good, a professor at the University of Illinois who has written extensively about USDA data, says that a number of recent quarterly stockpile reports have been simply incorrect—the numbers have been higher or lower than the actual amounts in storage. "Things went haywire in June and September," he says. "The cumulative numbers just don't make sense."

A USDA spokesman says the agency stands by the numbers.

The USDA data drive all sorts of decisions in the agricultural economy and beyond. Farmers use the information to help decide how much their corn is worth and when they should try to sell it. Ranchers, ethanol producers and food companies all pay attention when making their own corn purchases.

Many other nations rely on U.S. corn for feed, and some foreign buyers key on the reports. China, for example, swooped in to buy U.S. corn when prices fell after this year's June 30 stockpile report.

For all their shortcomings, USDA production figures are considered the best domestic agricultural estimates available.

Corn buyers and sellers say the department's task has gotten harder in recent years. With global demand for corn on the rise and prices soaring, U.S. farmers have been planting more. Production has grown by 30% over the past decade. The growth of the ethanol fuel industry, which has become a massive consumer of corn, has further complicated matters.

Accurate information about agriculture has been a national priority since the Civil War. Today, two measures of the corn harvest that the USDA focuses on are how big the next one will be and how much is in storage.

In eight of the 11 years from 1998 to 2008, the difference between monthly production estimates made between May to October—when the size of the crop is less certain than late in the year—and final U.S. production totals averaged less than 3%, according to the Journal analysis.

Last year, the USDA overestimated the size of the harvest in its monthly forecasts by an average of 6%, the widest average overestimation since 1995. That means the final harvest was about 750 million bushels smaller than what was forecast—equal to about 20 days of consumption. In 2009, the department underestimated the crop by an average of 4.5%, the biggest underestimation since 2004. Current USDA figures for this year suggest that the department overestimated the total crop by nearly 10% early in the season.

The USDA is trying to improve its production estimates, which it provides each month starting in May. In May, June and July, the estimates are produced by Mr. Bange's Outlook Board, and typically are based on recent yield trends, weather conditions and surveys of farmers by the department's National Agricultural Statistics Service, which asks them how many acres they plan to fill with corn.

The USDA's statistics service takes over each August. As the crop begins to take shape, it asks tens of thousands of farmers how many bushels per acre they expect to produce.

To fill out the picture, the statistics service sends corn counters into hundreds of fields in corn-growing states. They stake out 15-foot sections, then count stalks and measure the length and diameter of cobs. [Not exactly my idea of an exciting job, LOL.] Their goal is to extrapolate total production [duh].

Joseph Prusacki, a top official at the USDA's statistics service, says the department is trying to figure out a more reliable guide to what developing corn will weigh when it is mature.

The stockpile reports pose different challenges. The statistics service produces the quarterly reports, using surveys of farmers and commercial grain facilities to determine how much corn the nation has in storage. The Outlook Board issues monthly estimates of the stockpile size at the beginning and end of the marketing year, which starts with the fall harvest. Quarterly surveys can lead to changes in monthly estimates.

Frustration over the data has mounted in recent years. After the June 2010 stockpile report that caught Mr. Yost off guard, the statistics service that August increased the Outlook Board's July production forecast by 1%, even as analysts were raising concerns about the threat from a heat wave. The service then slashed its production estimates in September and October.

By the time the harvest was in, the August forecast turned out to be more than 900 million bushels too high.

This year, the stockpile figures released June 30 were higher than analysts expected. Corn prices, which had been nearly $7 a bushel, plunged to $6.29 in a single day.

"The USDA overstated the crop, and we believe that's why we've seen a correction since then," Bill Lovette, chief executive officer of Pilgrim's Pride, one of the nation's largest poultry firms and a major feed buyer, told investors in late July, after prices bounced back.

Mr. Yost, the Minnesota farmer, who headed the USDA's Foreign Agricultural Service from 2006 to 2009, says if the USDA numbers were more reliable, "it would serve the whole food chain better."‹