by Christopher Leonard / February 15, 2017
Inside Agri Stats, the poultry business’s secretive info-sharing service.
On Dec. 8, 2009, Joe Sanderson, the chief executive officer of Sanderson Farms, held a routine earnings call with bank analysts. His aim was to reassure them that the company’s recent profits weren’t about to disappear, as the chicken industry’s usual business cycle dictated they would. For decades, poultry had been volatile in a frustratingly predictable way: When times started getting good, companies flooded the market with chicken, causing prices to crash. Sanderson Farms was having a pretty strong year, so naturally the analysts were expecting a downturn.
Sanderson told them that the industry had learned from its mistakes. There wouldn’t be a bust this time. Then he said something rather extraordinary: His competitors weren’t planning to ramp up production. He knew this because it had been communicated to him by a virtually unknown company. “I see a lot of information from Agri Stats that tells me nobody’s going to ramp up,” he said. Sanderson was right. The following year, Sanderson Farms reported that its profits had surged 64 percent. For the next six years, production cuts and skyrocketing profit margins were the norm in the $90 billion chicken business.
At first the transformation puzzled industry watchers. Some speculated that a merger spree during the 1980s and 1990s was responsible—with fewer decision-makers in charge and fewer competitors, the remaining companies could more easily survey and predict the landscape. But Sanderson’s conference call suggested another source for the shift: Agri Stats, a private service that gathers data from poultry processors, produces confidential weekly reports, and disseminates them back to companies that pay for subscriptions.
Many industries, such as health care and retail, make use of information-sharing services, but Agri Stats provides chicken producers with a rare level of detail, in uncommonly timely fashion. The company’s reports, portions of which Bloomberg Businessweek reviewed, contain exhaustive data about the internal operations of the nation’s biggest poultry corporations, including bird sizes, product mixes, and financial returns at participating plants. According to a 2011 presentation prepared by Agri Stats, the company gathers information from more than 95 percent of U.S. poultry processors.
Agri Stats has for years maintained that its reports don’t violate antitrust laws, in part because the information provided is historical. A typical report doesn’t say how much a company plans to charge for a cut of meat, only what it charged last month or last week. But historical data can be used to gauge future production levels, as Sanderson, who declined to comment for this story, demonstrated when he said he saw no evidence of a forthcoming ramp-up. He was referring to Agri Stats data showing the number of egg-laying hens, or pullets, that his competitors were placing on farms. This figure largely determines the number of eggs that will be laid and therefore how many chickens will be hatched and grown—a key marker of future production.
Last September, Lockridge Grindal Nauen, a law firm in Minneapolis, filed a class-action lawsuit against more than a dozen of the nation’s largest chicken companies, alleging that they colluded to inflate chicken prices from 2008 to 2016. The suit was filed on behalf of Maplevale Farms and other wholesalers, who claim that the price they were paying for chicken was inflated by an illegal, if tacit, agreement among the big companies. Agri Stats, the suit says, “acted as an agent and/or co-conspirator” of the defendants, serving as a kind of digital evolution of the proverbial smoke-filled rooms where collusive schemes are said to be hatched and orchestrated. If successful, the suit could cost chicken companies billions of dollars in damages; they’ve filed a joint motion to dismiss. A separate lawsuit filed last year builds on the Maplevale case, making the same allegations of collusion but expanding the pool of plaintiffs to retail purchasers of poultry—meaning the 95 percent of Americans who eat chicken.
The U.S. Department of Justice, which is chiefly responsible for antitrust oversight, doesn’t appear to have taken any significant price-fixing actions against Agri Stats. But on Feb. 6, Tyson Foods announced that it was cooperating with a subpoena by the U.S. Securities and Exchange Commission seeking information about the company’s pricing practices. Some media coverage focused on the role played by the Georgia Dock Index, an obscure price-setting mechanism, but Tyson itself said the investigation had likely been sparked by the Maplevale lawsuit, which alleges a central role for Agri Stats.
In keeping with its insider business model, Agri Stats seems to relish its low profile. The company’s bland and thinly detailed website says, beneath an image of a small red barn in a cornfield, that it was founded in 1985, “on a kitchen table in the USA’s heartland,” by a man named Jim Cox. This down-home picture wouldn’t lead most people to imagine a sophisticated agribusiness-to-agribusiness data company whose development coincided with, and helped spur, a revolution in the poultry industry.
Cox was born during the Great Depression, in the town of Essex, Mo. When he was 12 his father died, leaving a widow to raise five children. “We were very poor, OK?” Cox recalls. He developed a strong work ethic, paying his way through Purdue University by laboring for as many as 80 hours a week on a dairy farm. After college he sold feed and feed additives for an Indiana company called Central Soya, which entailed spending a lot of time driving around to service large poultry companies across the South.
Chicken had long been a pricey specialty meat, reserved for Sunday dinners, but in the postwar era, technology and a new generation of entrepreneurs was making it cheaper and more available. The traditional barnyard coops were replaced by automated warehouses, where chickens were hatched by the million. Scientists at the University of Arkansas and elsewhere bred hyperproductive birds whose weight doubled on half the feed.
As Cox made sales, he collected information that helped him and his customers figure out which mixes of feed and additives helped birds add the most weight at the lowest cost. By the 1980s he’d persuaded his customers to fill out a standardized form that gathered key performance metrics. He didn’t realize at first that this information might be more valuable than the feed itself. As he recalls, one of his customers approached him and said, “Well, you know, Jim, while you’re getting that, why don’t you just make a little comparison? Don’t tell us who else is on the report or anything. Just make a comparison of the companies so that we can see how we’re doing.” Cox’s first report was about 10 pages long. His clients loved it, and soon he was handing out reports with feed orders.
When Central Soya was purchased in 1985, Cox sat down at his home in Fort Wayne, Ind., and decided to start a business focused entirely on data. He hired his secretary away from Central Soya, and the two set up shop with a borrowed copy machine and data-sharing agreements from 16 companies. That April he incorporated Agri Tech Systems (later changing the name to Agri Stats). Then he hit the road and persuaded more poultry producers to join his network, developing methods to standardize their data sets until his monthly report ran to about 500 pages. With each new page, Agri Stats’ value grew.
Cox’s long history in the chicken business gave him an in with poultry executives, who were a quintessential old-boy network. Many were Southern men who’d grown up in small towns during the Depression. When one of them trusted Jim Cox, others tended to follow. “They’re very, very competitive,” he says. “But at the same time, it’s like a family when it comes to relying on one another.”
As Agri Stats expanded, chicken was moving from the era of drumsticks to the era of McNuggets, overtaking beef and pork to become America’s staple meat. Chicken carcasses were being divided and subdivided into a complex map of economic opportunity: The breast meat was worth so many cents per pound, the thigh meat far fewer; the value of livers fluctuated wildly.
During the 1990s, Tyson became the world’s largest poultry producer in part by embracing data. Sensing an opportunity, Cox pitched Tyson senior executive David Purtle on his service. Purtle leveraged his company’s market position, striking a deal for an Agri Stats subscription, he recalls, “where I didn’t have to pay as much as everybody else.” Cox confirms this but adds, “I came back to him about a year later, and he anted up some more.” Whatever the price, it was worth it: Tyson added rich data on a significant number of plants to Agri Stats’ reports, making them a de facto industry standard.
Cox kept at it for another decade and a half before retiring in the mid-2000s. In 2013, according to SEC filings, Eli Lilly purchased Agri Stats for an undisclosed sum and folded it into its farm animal drug division. Today, at Agri Stats’ headquarters in Fort Wayne, dozens of software engineers, data specialists, and marketing managers oversee one of the largest private storehouses of information ever compiled about a single sector of the agricultural economy. According to Cox, clients submit their sales invoices in real time—when someone sells a truck of chicken to the Kroger grocery chain, for example, the invoice goes to Agri Stats soon after. He says the company tracks and stores data on the health and profitability of 22 million chickens every day.
This data trove is remarkable not only for its size but also for the secrecy with which it’s kept. The U.S. Department of Agriculture publishes reams of statistics on the hog and cattle businesses, which is of great value to farmers and commodities traders. But this isn’t the case for the chicken business because it’s so tightly integrated, with companies such as Tyson owning the birds and the feed, thus leaving fewer commercial exchanges during which the government can gather data. Instead, what is almost certainly the world’s largest repository of chicken business data sits on servers belonging to Agri Stats—for the benefit of its customers and, in rare cases, academics and industry groups who gain limited access for research purposes.
C. Robert Taylor, a professor of agricultural economics at Auburn University, has been an expert witness in many cases against big meatpackers, and over the years he’d heard rumors of a secretive poultry industry information-sharing service. But when attorneys provided him with a copy of an Agri Stats report for a case, he says, “I was shocked at the incredible detail.” His next thought, he adds, was that if it were public, it would be a gold mine for researchers.
Taylor couldn’t discuss the contents of Agri Stats’ reports, citing legal confidentiality rules, but a review by Bloomberg Businessweek of a report produced in 2014 confirms that the documents offer extraordinarily detailed information. One page shows an extensive revenue breakdown for 33 poultry plants, covering granular data in a number of areas, including product mixes—something most companies would characterize as proprietary. At the first plant listed—identified as “52” (only the company that has paid for the report sees its plants identified in an immediately recognizable way)—24.3 percent of its chicken was sold as deboned breast, 5.8 percent as wings, and 16.4 percent as whole leg.
Perhaps more important, the table shows which product mixes yielded the highest prices per pound. Plant 52’s deboned breasts fetched an effective sales value of 84.51¢ per pound, whereas Plant 53 sold 31.2 percent of its mix as deboned breast and received only 77.27¢ per pound. Plant 53 also processed a much larger bird, which affects the overall value of the products sold. Executives with access to the report could learn, too, that a rival made more money the previous month selling breast meat than wings, which could help them anticipate what the competitor might do next month as well.
Another page reveals an invaluable industry metric, one that poultry companies experiment with obsessively in search of a competitive edge: the caloric breakdown of the feed used in each plant. Plant 52 was using a feed mix with a calorie count of 2,401, while Plant 89 was using feed with 2,889 calories. Using these figures, a company could push its count closer to a competitor’s, in search of higher profits per bird.
Cox, Purtle, and others in the industry argue that such benchmarks make the industry more efficient on the whole. “If Agri Stats did anything,” Purtle says, “it got people focused on their numbers and trying to do better.”
Armed with Agri Stats data, the biggest chicken producers have been enjoying an unprecedented era of stability and profitability. At Tyson, operating margins in the chicken division have risen sharply since 2009, when they were 1.6 percent, according to SEC filings. The next year they were up to 5.2 percent. After a brief dip, they climbed to 7.9 percent in 2014, an astounding 12 percent in 2015, and 11.9 percent in 2016. A similar trend has been under way at Pilgrim’s Pride, where operating margins went from 3.08 percent in 2012 to 14.02 percent in 2014 and 12.77 percent in 2015, according to data compiled by Bloomberg. The recovery from recession accounts for some of the gains, but the poultry industry’s profit margins still have been abnormally fat and long-lasting by historical standards.
Accusations of collusion have dogged chicken companies for decades, but they’ve taken on new momentum in recent years, as three producers have come to control almost half of the U.S. market. According to Tyson’s most recent industry fact book for investors, which covers the 2013 fiscal year, Tyson and Pilgrim’s Pride account for roughly 40 percent of the market, while Sanderson accounts for 7 percent. These companies achieved their market positions partly as a result of the 1980s and 1990s merger wave, but also by integrating their operations. Tyson owns and controls its own hatcheries, feed mills, slaughterhouses, and trucking lines. Nominally independent farmers raise the birds under strict contracts, whose conditions the company sets.
Although this integration has made chicken cheaper and more plentiful, it also makes it easier for companies to implement production cuts. Because the life span of an eating chicken is so short—about six weeks in most cases—the big producers can quickly throttle back production by reducing the number of eggs their birds hatch when markets look weak. In the face of lower demand following the financial crash, executives openly discussed cutting back supplies. There’s nothing untoward about that, as long as they don’t coordinate cutbacks. Illegal collusion occurs when companies plan with one another to cut production ahead of time with the specific intent of raising prices.
The lawsuit filed by Lockridge Grindal Nauen on behalf of Maplevale Farms and other wholesalers names more than a dozen companies, including Tyson Foods, Pilgrim’s Pride, and Sanderson Farms. The financial stakes are enormous: Plaintiffs who prove their antitrust claims can recover damages equal to three times the harm caused by the collusion. The Maplevale lawsuit alleges that a price-fixing scheme took place across eight years, starting in 2008, and estimates that the wholesale chicken market during that period was worth at least $21 billion annually. So if the plaintiffs can prove that poultry companies colluded to raise prices by, say, 5 percent annually over eight years, they could win $24 billion in damages. (The suit doesn’t specify how much prices were allegedly inflated; technical facts such as these are usually established by economic experts at trial.)
Tyson, Pilgrim’s Pride, and Sanderson Farms declined to comment for this story, but the lengthy motion the defendants filed in late January seeking to dismiss the suit calls the collusion accusations a “conspiracy theory.” The filing doesn’t discuss Agri Stats in detail, but it points out that the company has been active for decades, while the collusion was alleged to have started around 2008. Eli Lilly declined to answer a detailed list of questions about Agri Stats and the antitrust litigation.
Cox disputes the idea that Agri Stats’ business model enables collusion and says he consulted with an antitrust attorney in New York when he was developing the concept as an employee of Central Soya. “It didn’t take him but about three seconds to say, ‘You don’t have a problem. As long as you talk about history, you’re OK,’ ” Cox recalls. “So I just took it on that premise, from Soya, that we were OK, too. And we have been challenged on that. Always won.” He declines to discuss further the legal challenges to the company, save to emphasize that the reports maintain the anonymity of participating plants.
Peter Carstensen, a law professor at the University of Wisconsin and former Justice Department antitrust lawyer who has studied Agri Stats while researching the modern poultry industry, casts the level of plant-by-plant detail in the company’s reports as “unusual.” He explains that information-sharing services in other industries tend to deal in averaged-out aggregated data—for example, insurance rates in a given state. Such services run afoul of antitrust law, he says, when they offer projections or provide data so detailed that no competitor would reasonably share it with another. Getting detailed information is a particularly useful form of collusion, Carstensen says, because it allows co-conspirators to make sure they’re all following through on the agreement. “This is one of the ways you do it. You make sure that your co-conspirators have the kind of information that gives them confidence—so they can trust you, that you’re not cheating on them,” he says. “That is what creates stability for a cartel.”
Whereas Cox and others argue that the anonymity of Agri Stats’ information mitigates this concern, lawyers in the Maplevale case assert that poultry companies can reverse-engineer the data, thanks in part to a special monthly document Agri Stats produces called the Bottom Line Report, which breaks down key financial metrics, such as interest expenses, on a single row. The lawyers say it’s relatively easy to match this anonymous data against numbers found in publicly available records.
Agri Stats and its president, Brian Snyder, declined to grant an interview or provide basic facts such as its annual revenue—information Eli Lilly also keeps out of its public filings and earnings statements. “We view our operations and strategy as confidential and proprietary information and not appropriate to discuss in a public forum,” Snyder wrote in an e-mail. He referred questions to the company’s website, with its scant detail and bucolic photography.
In the weeks after the Maplevale suit was filed, nobody paid it much attention. Then, on Oct. 7, a veteran stock analyst named Timothy Ramey of Pivotal Research Group downgraded shares of Tyson from “buy” to “sell” and slashed his valuation of the company’s shares. The downgrade note was perhaps unique in the antiseptic genre of analyst reports, in that it read more like a petition for divorce than a simple recommendation change. After professing deep admiration for Tyson Foods CEO Donnie Smith, Ramey wrote that the evidence Tyson and others had colluded using Agri Stats was “chilling.” There was no simpler explanation, he wrote, for the post-2009 “perfect harmony” of industry production that had delivered record profit margins for Tyson.
Prior to the note, Ramey had often written admiringly of Tyson and Smith. He says he considers Smith a friend—he’s even had him over for dinner—and one of the five best CEOs he’s researched in his three decades covering the food business. But as the industry’s boom proceeded, Ramey grew perplexed. “When you start getting 12 percent and 13 percent EBIT margins—when the returns on capital are just incredible in the business—then you just have to start to scratch your head and say, ‘How come this isn’t behaving like every other cycle in history?’ ” he says. The Maplevale complaint, he adds, “sort of sent shivers down my spine. I said, ‘I don’t know if this is right or wrong, but it definitely explains a lot about the last eight years if it is right.’ ”
Ramey acknowledges that other factors could be in play, including the Georgia Dock Index, a wholesale price index that poultry companies use to set prices. That index, which is based on industry self-reports, gained notoriety late last year when public records were released showing that the Georgia state employee who compiled it had told his bosses he feared poultry companies were giving him inaccurate numbers. (Some dubbed the episode “chicken Libor,” in reference to a scandal involving manipulation of an exchange rate set in London for interbank loans.) The price was suspended briefly, but it has since been reworked and relaunched.
The Georgia Dock price is a concern, Ramey says, but it’s exceedingly unlikely it could account for recent trends on its own. He points out that only a small share of Tyson’s chicken prices are even based on the index and that, absent production cuts, oversupply of chicken would swamp any price-rigging based on a single index.
The same day Ramey’s note was released, Tyson issued a statement disputing “the speculative conclusions reached by the analyst” and saying it would defend itself in court against the Maplevale case. Nevertheless, its stock fell about 9 percent that day. No other major analysts appear to have followed suit by downgrading Tyson to “sell,” but the company’s stock is down 12 percent since Ramey’s note, in part also because Smith stepped down as CEO in late November, at age 57. (Tyson declined to comment on Smith’s departure but has said elsewhere that it was unrelated to pending litigation.)
The Maplevale lawsuit is still in its early stages and could take years to resolve. The plaintiffs have until March 15 to respond to the motion to dismiss. In the meantime, a hearing set for Feb. 24 could determine if they’re able to begin full discovery, which would give them access to sensitive documents, such as internal communications and Agri Stats reports, that could reveal whether the companies used Agri Stats to help coordinate production cuts.
Meanwhile, Agri Stats continues to expand its business. The company has opened operations in Brazil to take advantage of growth by U.S. poultry companies there. It has also been branching out into the hog business, which has, over the past 30 years, started to look more and more like the chicken industry, with hogs being raised under contract for vertically integrated companies such as Smithfield Foods. It appears that demand for the service is strong. At a hog industry trade show in 2011, an Agri Stats employee pitched the company’s services. His slideshow indicated that 27 companies had already signed up.