When all is said and done, predicting the damage done to U.S. corn and soy crops from the worst drought in half a century may have been the easy part.
With the harvest imminent and plants mature, most traders are fairly confident they have a handle on this year's supply. Whether corn yields are 120 bushels an acre or 130, it's clear that demand will outstrip supply, possibly by a wide margin. Demand, in trade parlance, will have to be "rationed".
What's far less clear -- and harder to discern now than ever before -- is just how much less food, feed or fuel will be made from corn as buyers cut back. From food companies to livestock ranchers to ethanol plants, the calculations are complex: Can end consumers withstand higher prices? Can they sustain production with cheaper grain alternatives?
For traders, that complexity is multiplied. The unpredictability of dry weather is nothing compared to the vagaries of consumption by livestock producers, exporters, ethanol makers and other industrial users that turn corn into scores of products including plastics, adhesives, explosives and pharmaceuticals.
So after two months of relatively steady price gains as every passing hot, dry day withered the crop a little more, some are bracing for a bumpy spell in which traders attempt to second-guess the price point at which demand is rationed.
"Demand occurs in so many different categories that it's a little hard to get your hands around," said Darrel Good, a respected agricultural economist with the University of Illinois and a foremost authority on the topic.
"The thought process is pretty clear, but quantifying things is pretty subjective."
The U.S. Department of Agriculture cut its 2012/13 U.S. corn crop forecast by 4.011 billion bushels, or 27 percent, over the past two months and slashed its estimate of corn use across all demand segments by 2.55 billion bushels, or 19 percent.
U.S. inventories at the end of next summer are now expected to fall to 650 million bushels, a 17-year low and considered near the bare minimum required to prevent an unprecedented scramble for the last kernels. As recently as June, the USDA had forecast 2 billion bushels.
Few traders expect those numbers to be final, with the agency fine-tuning for months to come.
In theory, the mystery of demand should offer opportunities for smart traders to capitalize on volatile markets. In practice, the evidence needed to quantify demand can be so disparate and piecemeal that it defies order.
U.S. export sales are the easy part, strictly reported on a weekly basis. For everything else it's a case of scouring data on livestock slaughter rates, chicken egg sets or sow liquidation; anticipating government policy on ethanol; or seeking private market intelligence on how much wheat is replacing corn among pig farmers.
Cargill CARG.UL said this week that it was delaying the announcement of its 2013 sweetener pricing, suggesting the agribusiness giant may be anticipating difficulty in sourcing corn.
The change in sentiment from a supply-induced panic to a period of demand uncertainty is already weighing on prices. From Friday through Tuesday, corn for December delivery on the Chicago Board of Trade (CBOT) fell 4.2 percent to post the biggest three-day drop since June 13, just days before traders first fixated on the drought.
From mid-June to mid-July, corn embarked on a remarkably orderly rally, surging for several days and then pausing for another push. Rarely did prices close at their daily maximum. Never did they fall more than two days in a row. For the past two weeks, however, they've just bounced around.
"The first phase is unidirectional, where the market goes into a little bit of a panic mode, trying to price the day-to-day weather," said Malinda Goldsmith, a partner at agriculture-focused and Dallas-based Four Seasons Commodities, where she oversees a $50 million portfolio that gained 10 percent in July.
"Then you get to a point where the demand gets to be destroyed. And the market gets more choppy. Then you get into a very volatile period, which is the one we're in now, where we have sharply higher, sharply lower prices."
She concluded: "Now it's going to get dicey."
Modeling demand for food is a tricky task in the best of times. It is generally regarded as largely "inelastic", meaning consumption tends to resist the effects of higher prices. Historical assumptions can provide a useful guide.
But the growing role of the ethanol industry -- which consumes 40 percent of the U.S. crop -- has introduced a new set of variables. Even setting aside the possibility that President Barack Obama could offer a waiver of ethanol blending quotas this year, corn traders must now take stock of New York gasoline prices, Brazilian sugar exports and ethanol credits.
"We really didn't have ethanol when we had this kind of a shortfall in production in the past, so how that sector will respond is something new," Good said.
"There's a lot of difference of opinion out there in how much corn will eventually get consumed for ethanol."
The USDA cut its forecast of corn use for ethanol in the 2012/13 marketing year by 400 million bushels this month to 4.5 billion bushels, the lowest in four years. At that rate, output would be short of the government's 2012 mandate of 13.2 billion gallons (50 billion liters).
Some analysts say the estimate is too low. Rising oil prices mean ethanol is still cheaper than gasoline in some places, encouraging refiners to maximize its use in blending.
Production cuts by an estimated two dozen plants -- mostly older, less-efficient facilities located further from the corn belt -- have helped boost margins enough to sustain output.
"I think we'll see a fairly aggressive pace of demand rationing in the export sector first, followed by feed, followed by ethanol," said Shawn McCambridge, analyst with Jefferies Bache.
But others questioned whether use may decline further, citing a recent slowdown in ethanol production blamed on poor plant margins amid this summer's historic rally in corn prices.
Calls for the suspension of a government mandate on ethanol blending could be a swing factor, with the United Nations chiming in on the debate last week.
"I think the ethanol could be lower than (the USDA estimate) ... If the price of ethanol increases, which I think it will given the higher cost of production, it may not be as competitive an octane source," said Bill Lapp, president of Advanced Economic Solutions. Demand may be nearer 4 billion bushels, he said.
Overseas demand is another part of the puzzle. While theoretically made easy thanks to U.S. government rules that require exporters to report all sales within a week of a deal, predicting the pace of annual shipments is tricky work.
The USDA has already cut its outlook for 2012/13 exports to the lowest in 28 years, at just 1.3 billion bushels. It slashed 300 million bushels in each of the past two months.
Because this year's supply distress is domestic, unlike during the price spikes of 2008 and 2010, many importers are simply buying from other producers such as Brazil, which yielded a record corn crop this year.
But many of them have also begun more aggressive campaigns to replace corn as a source of livestock feed, relying on feed wheat instead. This trend has helped to whittle global wheat stocks, which hit a record 200 million tonnes two years ago.
South Korea, traditionally a top-five U.S. corn customer, has been a particularly active feed wheat buyer of late. The country's wheat imports hit a record 5.2 million tonnes this season as nearly 2.9 million tonnes were imported to feed livestock, while its corn imports fell to the lowest in three years, according to the USDA.
Emerging corn importer China was forecast to produce a record-large 200-million-tonne crop this season, which would limit its import needs to just 2 million tonnes, the USDA said. Earlier this year, China had been expected to import a record 7 million tonnes in 2012/13.
"I don't see a robust year for exports whatsoever. I wouldn't be surprised to see something around a 1.2-billion-bushel program. The first rationing is going to be done on what moves beyond the borders, not within the borders," McCambridge said.
LIVESTOCK FEEDERS FIRST TO SQUEAL
Predicting the amount of corn being used by livestock and poultry producers has always been challenging because the USDA releases corn stocks data only on a quarterly basis.
With the advent of the Internet, forecasters have access to scores of local corn and livestock prices, which help them get a clearer picture of profitability among meat and dairy producers.
But as corn prices climb and supplies become harder to source, producers are prone to explore alternatives such as feed wheat or replace a portion of their feed rations with something cheaper like soybean hulls or wheat middlings.
Domestic livestock and poultry producers, traditionally the largest consumers of corn, have seen their share shrink from more than 60 percent two decades ago to just 37 percent due to ethanol.
They have been squeezed by high corn prices. However, other than an increased sow slaughter, a sign of herd downscaling by the pork industry, there has been little liquidation of hogs, cows and chickens.
Some livestock and poultry operations in the southeastern United States have imported corn from Brazil while cattle feeders in the southwest have brought in Canadian feed wheat, but the volumes have been relatively minimal.
More imports are likely as long as corn prices remain elevated, but shipping costs may be prohibitively high in many cases, likely limiting imported volumes.
In the meantime, feed users are likely to maintain a hand-to-mouth corn-buying strategy, extending coverage on price dips and trying to minimize losses.
But the months ahead will be challenging, particularly for cattle operations. The longer life cycle of animals makes it difficult to change course when market conditions turn sour.
"The real question is not how much demand will drop with each dollar rally in the corn market, but what is the margin difference when you rally corn by a dollar, live cattle rally and feeder cattle drop," said Tregg Cronin, a market analyst with Country Hedging.
"You run all these crushes and margin calculations on a rolling basis and as long as you can pencil a margin, you have to assume that the demand center is going to use corn. It's a lot more art than science."