BusinessWeek uncovers that the cable channel's own design flaw may be behind the investigation into its million-dollar stock-picking contest
In the past few months, Jim Kraber became more than a little obsessed with CNBC's "Million Dollar Portfolio Challenge." At the peak, the 42-year-old was spending 12 hours a day on the contest, using three computers in his Greenwich Village apartment to trade 1,600 different portfolios, all in an effort to win the $1 million grand prize. He even dropped his studies for the chartered financial analyst (CFA) exam, given once a year, so he could have more time for the financial news channel's game.
He made it into the group of 20 finalists, but in mid-May, as the last round of trading opened, he noticed an unusual pattern in the picks of other contestants. One trader had a stream of near-perfect picks, consistently placing huge bets on shares that soared in after-hours trading. Kraber suspected the trader and perhaps others were getting help from someone who was changing their picks after the stocks' increases—and he quickly notified CNBC. "I went back and looked at his trades and thought, 'This is pretty much statistically impossible,'" says Kraber, who holds master's degrees in business and statistics from New York University.
Kraber says CNBC rebuffed him at the time, but now it looks like he may have been right. Several contest participants have told BusinessWeek that there was a flaw in the design of the CNBC game that allowed certain players an unfair advantage. As many as four of the top contestants in the million-dollar contest may have exploited the flaw, according to the participants interviewed by BusinessWeek. On May 30, two weeks after Kraber says he notified the cable channel, CNBC posted a notice on its Web site that there have been allegations of trading strategies "in violation of the contest rules" and that it is investigating the issue. It has not disclosed the nature of the alleged problems.
"An Aggressive Investigation"
For what is essentially the American Idol of stock picking, the stakes are sky-high. A million dollars is on the line for the ultimate winner of the contest, along with fame and future opportunity. CNBC, which bills itself as the essential news channel for investors and businesspeople, has a reputation to protect. With constant promotion from channel anchors like Becky Quick and Joe Kernen, the contest attracted 375,000 participants and nearly tripled the traffic to CNBC's Web site. Now, the channel, which is part of General Electric (GE), may have to publicly acknowledge mismanagement of the contest and could face potential lawsuits from disgruntled participants.
"I figured out very quickly what was going on—and they were looking at all the trades," says Kraber. "It's not negligence that they might have made bad software. It's negligence if they knew this was going on and did nothing to stop it."
CNBC declined to comment specifically on Kraber's allegations. A spokesman for the cable channel says, "Once these issues were raised, we launched an aggressive investigation immediately. The integrity of the contest is very important to us." CNBC, run by President Mark Hoffman, has been a money machine for GE, even as NBC overall has struggled (see BusinessWeek.com, 2/5/07, "Jeff Zucker Takes Charge at NBC Universal").
Too Good to Be True?
The performance of some participants does look unbelievable, literally unbelievable. Over the first nine trading days of the final round, the top five stock pickers tallied average returns of 45%, according to a BusinessWeek review of their trading portfolios. If that kind of performance was stretched out over a year, it would work out to an annual return of more than 1,200%. "Obviously if you have knowledge of what's happening with the stock, that would really skew the results," says Lubos Pastor, a professor of finance at the University of Chicago, who is speaking generally and has not studied the trading in the CNBC contest.
The top traders are reluctant to discuss their performance. Four of the top five performers declined to discuss their trading in detail with BusinessWeek, and the fifth could not be located for comment. "I don't want to jeopardize anything by saying something stupid," says Joe Dondero, who was fourth in the standings on the last day for which results have been made public.
How could traders exploit CNBC's glitch? According to several participants, the technique was relatively simple, but not obvious to all participants. A trader could go to the CNBC Web site and select a number of stocks to buy, but hold off on executing those trades. If you made the selection before the close of regular trading at 4 p.m. EST and left your Web browser open, you could execute those trades after hours and still receive the 4 p.m. closing price. For example, if a company whose stock closed at $20 a share rose to $25 in after-hours trading, you could buy the stock at $20, even though it was already worth 25% more (see BusinessWeek.com, 6/8/07, Slide Show: "How to Game CNBC's Stocks Contest").
The allegation is that certain traders may have used the technique with companies that were reporting earnings and other important news after the market's close. They could select as many as 50 stocks and then execute trades for only the one or two best performers.
Serge Amelyan, a real estate investor from Mequon, Wis., was one of the 20 finalists with prescient trades. On the first day of the final round, he placed a big bet on Mindray Medical International (MR), which was reporting earnings that afternoon. Amelyan cashed in as the stock rose 7% in after-hours trading, more than any other stock with a late earnings announcement. The very next day, he invested heavily in Compuware (CPWR), which again reported strong earnings and surged 7%. In all, seven of the nine stocks Amelyan invested in during the finals announced earnings after the markets closed. Six of those saw sharp increases, while one had a slight decline. Amelyan says he didn't use any unusual trading practices. "I didn't play that way," he says.
From Pharmaceuticals to Finance
Kraber, the player who helped uncover the CNBC contest's possible flaw, grew up in Dixon, Ill., about 100 miles west of Chicago. He majored in chemical engineering at the University of Illinois, where he says he began to develop an interest in the markets. But when he graduated in 1989, he found he didn't have the stomach for the risk he saw in a career trading. Instead, he took work in the pharmaceutical industry and moved to New York.
Kraber's interest in finance remained strong, however. So after five years working for Merck (MRK) and Schering-Plough (SGP), he decided to pursue an MBA in finance at New York University and then followed that with a master's in statistics. He says his three degrees have quite a bit in common: The knowledge needed to analyze heat transfer and, say, trade complex derivatives are closely connected, maintains Kraber. "I see it all as very much related," he says. He now trades for his own account and does market risk analysis for others on an ad hoc basis.
When Kraber heard about CNBC's million-dollar challenge earlier this year, he knew he wanted to enter. But it wasn't until he read the rules of the game that he figured he had a pretty good shot at making the finals. The key was that CNBC put no limit on the number of portfolios a player could manage, and only the best-performing one would count. So Kraber, with his expertise in statistics, computer-programming, and stock selection, could set up hundreds of different portfolios, all pursuing high-risk, high-return strategies. By sheer chance, at least one of his portfolios would do well, and he figured that with smart strategic picks he'd rank near the top of all the participants. "I realized I had an almost 100% chance of making the finals," he says.
The first round of the CNBC challenge, which began on Mar. 5, consisted of 10 one-week contests. The winner of each weekly game won $10,000 and automatic entrée into the final round. In addition, the top 10 finishers overall made it into the final round. In mid-April, Kraber grabbed the weekly prize for the sixth week of the contest, with a portfolio that included WD-40 (WDFC), Cascade (CAE), and Apogee Enterprises (APOG). He became one of 20 contestants for the big-money prize.
"Very Suspicious" Trades
But it didn't take him long to notice curious behavior. On the morning of May 16, after the results of the final round's first day of trading were posted on CNBC.com, at least two traders posted big gains from Mindray. As Kraber saw what he thought were uncanny picks by other finalists he became incensed. He called CNBC and requested an investigation into any trades processed after 4 p.m., the official deadline laid out in the contest rules. "It's obviously against the rules, which say there's no after-hours trading," Kraber says. When he pointed this out to a CNBC marketing rep he managed to get on the phone, "she went off on me," he says.
Unbeknownst to Kraber, other contestants began to notice peculiar trading patterns. Two other finalists, who spoke only on the condition of anonymity, say they also saw that several of the top contenders were consistently picking the stocks of companies with strong earnings reports after the market close. "It was very suspicious that people were making such good calls so consistently," says one of the two finalists.
The two traders say that it wasn't until they started talking with each other and two other finalists that the four of them together developed the theory of how the CNBC glitch may have worked. They hypothesized that if a contestant left open the browser window with the challenge's trading interface past the 4 p.m. deadline, it might still be possible to reorder, but not add, cancel, or change the quantity of trades. The two sources say that members of their group have demonstrated how this could be done.
Winner to Be Named
The two sources flagged CNBC near the end of the competition, when they became convinced they had figured out the puzzle. On May 24, the day before the contest closed, the pair scanned all the companies that were reporting earnings after the market's close, and noticed that Verigy (VRGY), which makes semiconductor testing equipment, had soared 20%. They thought that if their theory was right, many of the leaders would end up getting the stock at the 4 p.m. closing price. The group alerted CNBC to look for Verigy picks. The next day, they saw that three of the four players they suspected had picked Verigy. Amelyan didn't buy Verigy, opting instead for Red Robin Gourmet Burgers (RRGB), which gained 8.4%.
Now, it's left to CNBC to sort through what to do about the contest's troubles and who will get the million-dollar prize. The cable channel has said that it hopes to name the winner by July 8, but it may push back the deadline if more time is needed to get to the bottom of the trading issues.
Kraber realizes that he is unlikely to receive the prize money, even if the people suspected of exploiting the software glitch are disqualified. He finished in 12th place, behind several other people who don't appear to have used the loophole. He says he just wants CNBC to do what's right. "I think they knew about this and let it go," Kraber says. "I just want some accountability."
Check out the slide show to see how some participants believe the loophole may have worked.