Capital One Fraud Researchers May Also Have Done Some Fraud

The best small fundamental equity long/short fund was two guys misusing a database.

Good thing they didn't have my credit card data.

Photographer: Patrick T. Fallon.

Part of why people don't like insider trading is that it seems too easy. Some people spend their days slaving over a hot spreadsheet, trying to figure out if a company will make money or not, and then you just waltz in with a tip from your buddy at the golf club and buy some call options on the company just before it announces a merger. It's just unfair. 

Say what you will about Bonan Huang and Nan Huang, but they (allegedly) worked hard for their hot tips. You don't see a lot of this on the golf course: 1

query CMG

That's a heavily redacted list of search queries that they allegedly ran in Capital One's database of credit card sales, looking to see how many people were using their Capital One cards at Chipotle. 2   Bonan Huang and Nan Huang worked at Capital One "as data analysts tasked with investigating fraudulent credit card activity," but, I mean, the database was just sitting there, how could they resist taking a peek? They could not.

Their queries seem to have revealed that a lot of people were putting burritos on their Capital One cards, because on July 21, 2014, the day after querying the database, Bonan Huang and Nan Huang between them apparently bought call options on 5,500 Chipotle shares for a total of just less than $100,000. Chipotle released earnings after the market closed that afternoon. Earnings were good; in particular, revenue was up 28.6 percent quarter-on-quarter. 

The next day Bonan Huang and Nan Huang allegedly started selling their options, making a profit of about $278,000. For three days' work. But at least they wrote the queries. Usually when I say "for three days' work," the work was golf. These guys did work.

If you believe the Securities and Exchange Commission, actually, they did a ton of work:

Defendants worked for a large credit card issuer as data analysts tasked with investigating fraudulent credit card activity. While employed there, Defendants searched their employer's nonpublic database that recorded the credit card activity for millions of customers at numerous, predominantly consumer retail corporations. The Defendants conducted hundreds, if not thousands, of keyword searches of this database. These searches, which were not done in furtherance of their employment duties, allowed the Defendants to view and analyze aggregated sales data for the companies they searched.

Isn't that sort of sweet? I mean, these guys appear to have done fundamental research on a bunch of companies, and then bought stock 3  in the companies whose fundamental performance was better than market expectations, while selling stock 4 in the companies whose performance was worse than expected. The SEC singles out Chipotle, as well as Cabela's and Coach, where they bought put options because sales were decreasing, though there seem to have been quite a few other trades as well. 

Apparently -- unsurprisingly -- there was a pretty strong relationship between people buying burritos or guns or purses with Capital One cards, and people buying burritos or guns or purses with cash and other credit cards, so their research proved profitable. Ridiculously profitable. From the SEC complaint:

From January 2012 to January 2015, defendants Bonan Huang and Nan Huang deposited a total of $147,300 into their six OptionsHouse accounts. During this time period they transferred approximately $1,763,500 out of these six accounts. As of January 15, 2015, the total balance in the six acounts was approximately $1,063,000. Accordingly, Bonan Huang and Nan Huang made approximately $2,826,500 trading options during this period in their OptionsHouse account. This represents a three-year return of approximately 1,819%. 

That's amazing! These two like customer-support guys at Capital One were seemingly running an incredibly successful fundamental research-driven long/short equity hedge fund. A small fund, but still. The average equity hedge fund returned 25 percent -- total, not annual -- during that period. 5  You sometimes see insider-trading cases where someone makes like a thousand-percent return in a few days by buying call options just before a merger. Every so often a network of tippers will yield multiple big scores like that. But to do hundreds of searches and trade multiple stocks over three years based entirely on raw consumer spending signals, and to make 1,819 percent doing it, is just phenomenal. Even if the consumer spending signals were, you know, stolen.

People have asked me if this is insider trading and, you know, sure it is? (If the allegations are true, I mean.) This is not "classical" insider trading -- trading or tipping by an insider at Chipotle or whatever -- but rather "misappropriation" insider trading:

The "misappropriation theory" holds that a person commits fraud "in connection with" a securities transaction, and thereby violates § 10(b) and Rule 10b-5, when he misappropriates confidential information for securities trading purposes, in breach of a duty owed to the source of the information. ... Under this theory, a fiduciary's undisclosed, self-serving use of a principal's information to purchase or sell securities, in breach of a duty of loyalty and confidentiality, defrauds the principal of the exclusive use of that information. In lieu of premising liability on a fiduciary relationship between company insider and purchaser or seller of the company's stock, the misappropriation theory premises liability on a fiduciary-turned-trader's deception of those who entrusted him with access to confidential information.

Here, Bonan Huang and Nan Huang allegedly got the information from their employer, Capital One, which was supposed to have exclusive use 6 of the -- hey, wait a minute, does that mean that Capital One was allowed to trade on this data for its own profit? Wouldn't that be amazing? Surely the answer is no: I assume that Capital One signed agreements with retailers (or rather, with Visa and MasterCard, which signed agreements with retailers) in which it promised not to disclose transaction data, or use it for nefarious purposes. 7 Really anyone who used this data would be misappropriating it from, ultimately, Chipotle. Which gets to keep its sales data to itself. Except once a quarter when it releases that data and the stock jumps.

Henry Manne, the pioneering scholar of law and economics who died last week, famously argued that insider trading should be legal, in part because it makes markets more efficient, and this case is a good example. Chipotle's and Coach's and Cabela's stocks were mispriced, the day before their earnings announcements, because those companies had earnings information that the market didn't have, and didn't tell anyone. (Until the next day.) People bought and sold those stocks at the wrong price all day long. Bonan Huang and Nan Huang seem to have done their research to figure out the right price. Illegal research, sure, but they were right -- spectacularly, and over and over again. 

That's how markets work: People do research to try to figure out the right price, and then if the price is wrong they trade, and eventually prices get to be right. And so there are tons of legal, yet somehow unfair-seeming, ways in which smart traders try to figure out the right price. There are helicopters with heat-sensitive cameras flying over oil tanks to help hedge funds get non-public oil supply information. There's a former Google engineer "selling analysis of obscure data sets" -- like "satellite images of construction sites in 30 Chinese cities" -- "to traders in search of even the smallest edges." Or there are like a billion people trying to use Twitter to predict stock prices. That is the business. You take the data that is out there, or find new ways to get new data, and then you analyze the heck out of it to find out if it tells you anything about companies that you didn't already know.

And usually the answer is that it tells you a teeny little bit, and you add a few basis points to your returns. The returns to discoverers of new data sets tend to dissipate quickly -- in part because others discover them too, but in part because, you know, the market is smart, lots of people have incentives to figure this out, how much more information could one more piece of information really give you, etc. Markets are basically efficient. (Right?) If you had asked me two days ago if raw Capital One credit-card usage data would be helpful in making excess returns in the stock market, I'd have said, sure, of course. If you'd asked me if you could use it to make consistent excess returns of 1,800 percent over three years, though, I would have been skeptical. Surely lots of Wall Street firms -- Chipotle is followed by 31 analysts -- and asset managers are doing tons of research to try to estimate Chipotle's sales. They're visiting branches and calling investor relations and talking to pork suppliers and surveying consumers and generally getting paid a lot of money to build a robust estimate of how many burritos Chipotle is selling. One more piece of data -- one credit card company's charges at Chipotle -- would be helpful, but come on, not that helpful.

Nope: Super helpful! I don't know what to tell you. It seems a shame that Bonan Huang and Nan Huang's research was apparently illegal. Because it was really good.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.
  1. This is from Exhibit O to the motion for a temporary restraining order in the case. 

  2. The redactions are to prevent you from seeing that it's Chipotle. For some reason. The SEC complaint doesn't identify the credit card company at which Bonan Huang and Nan Huang worked, or any of the companies whose stocks they allegedly traded. This is just "Company C." But the SEC provides tons of identifying details -- not just dates of earnings releases, stock prices and earnings data for each company, but also full earnings releases with just names blacked out. I have no idea why it would do that. Anyway this Bloomberg News article matches everything up. Company A was Cabela's, Company B was Coach and Company C was Chipotle Mexican Grill. 

  3. Umm, call options.

  4. Buying puts. And put spreads actually. See paragraphs 30-31 about their trading in "Company A" (Cabela's).

  5. That's the HFRI Equity Hedge (Total) Index, Bloomberg HFRIEHI <Index>, from Dec. 31, 2011 through Dec. 31, 2015. (So, roughly that period.)

  6. The SEC lays this out in paragraphs 46 through 50 of the complaint; in particular, Capital One required employees to agree not to "use any material nonpublic information you may obtain about their businesses" in "trading stock of [Credit Card Company] business partners."

  7. If that assumption is wrong, then why isn't Capital One licensing this database to hedge funds for vast piles of money?

To contact the author on this story:
Matt Levine at

To contact the editor on this story:
Zara Kessler at

Before it's here, it's on the Bloomberg Terminal.