The investment landscape is changing rapidly and asset managers are finding themselves in an arms race for new forms of alternative data. Everything from mobile phone location data to internet of things (IoT) data about crop yields, are being scraped, sanitized, packaged and sold on to investment managers looking for an edge when making decisions.
But the brave new world of big data-driven investing is something of a regulatory grey area. At what point does some exclusive, real-time look through into a company's performance constitute insider trading? In legal terms this means trading on material, non-public information received in violation of a duty to keep it confidential.
Highly respected Wall Street lawyer Jonathan Streeter has said a legal case involving the purchase of data for trading purposes could be brought soon. Streeter, a New York-based partner at the law firm Dechert, was the federal prosecutor who led the insider trading case against Raj Rajaratnam, co-founder of Galleon Group.
Streeter says the ways and means of using alternative data is becoming a hot topic among the hedge fund clients he advises. "My clients are often calling me up and saying, I have this vendor who wants to sell me this data set; do you see any problems with that?
"Oftentimes the way it works is, there's the original kind of creator of the data or owner of the data, the cell phone company or the company that takes the satellite images or whatever, and then there's a vendor in between who buys the data from them, massages it in various ways and turns around and sells it to investors."
There are two lines of legal analysis here: one is the securities law angle which concerns the risk of insider trading on material, non-public information. Secondly, there's the question of internet laws and data privacy rules.
The hedge funds and asset managers who scope out alternative data are generally interested in aggregates; not only don't they require personal details within the data, they don't want it.
Regarding the risk of insider trading, it's perhaps useful to differentiate along the lines of public and non-public data. I can stand in Walmart's car park and count cars all day, so using satellite imagery to do this may give me an unfair advantage, but doesn't change the fact that it's public data.
However, imagine a very big retailer, for instance, selling transaction data pinpointing how many Panasonic televisions it sold this quarter versus last quarter.
Streeter said an example like that may catch the eye of regulators. "There is stuff that's happening, or is about to start happening that I think the SEC, the FCA and other regulators are going to be interested in," he said. "There are some of these data sets that start to look like, wow, someone is really going to have a huge advantage if they have this data set.
"If you're a hedge fund you may be able to buy this data set and ordinary investors don't have access to that information. And that is material, non-public information about product sales that a public company is going to announce at its next quarterly earnings."
Tammer Kamel, CEO of Quandl, a platform for financial, economic and alternative data, agreed that there are situations where a certain data set if used or sold, would constitute insider information.
He said: "My understanding is that there has to be some breach of a fiduciary duty or some breach of a confidentiality promise before you are really in danger of breaking those kinds of regulations and laws.
"The big retailer example you gave could flirt with that situation because that retailer is a large enough channel for Panasonic, it would be in possession of information that is of material importance to Panasonic.
"So if that retailer went and traded on that information, or some employee of the retailer went and traded on that information, for example, I think they are breaking the law.
"But if the retailer sells that data set to anyone who wants to buy it, then yes, it is material information but it's being made available to anyone who wants to pay for it—in the same way that Bloomberg has lots of material information on their platform and it's not available to everybody; it's available only to those people who pay for it."
Kamel also pointed out that if the retailer did that to Panasonic, "the CEO of Panasonic would probably call up and say, 'if you do that again I'm not selling another TV on your platform ever again!'
"Notice, it's a business recourse; it's not a legal recourse," he added.
This argument that a big retailer or a credit card company would never risk their core business by selling data about major customers—even if it is worth a lot of money to investors—does not convince Streeter.
He said the price of data is increasing and it's becoming more of enticing value proposition, and the rules are murky. He agrees that whether there is insider trading often turns on whether there is a violation of a duty of confidentiality, but explains, "The contract between these companies was probably written at a time when no one anticipated this problem. The next time they negotiate their contract, Panasonic might say, you can't sell our data, or Panasonic might say, you can sell our data but you have to give us a cut of it or something.
"But at the same time public companies have to think about SEC rules that prohibit them from selectively disclosing their financial results to investors. There are all kinds of ways in which you can imagine this playing out," noted Streeter.
"Airlines, retailers, cell phone companies, credit card companies—they all have data sets that could be very valuable, and there's a vendor out there who'd like to buy it off them, slice and dice it and sell it on to an investor.
"There are all kinds of legal questions embedded in there about whether they can do that. And I would expect investors like hedge funds with significant assets under management and sophisticated compliance departments to be held to a pretty high standard."
Both Tammer and Jonathan will be talking about alternative data at Newsweek's AI and Data Science in Capital Markets conference on December 6-7 in New York, the most important gathering of experts in Artificial Intelligence and Machine Learning in trading. Join us for two days of talks, workshops and networking sessions with key industry players.