When is a scoop non-public information?

April 25, 2012
provocation yesterday, where I suggested that the NYT could sell advance access to its stories.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google,mail" data-share-count="false">

Many thanks to everybody who responded to my provocation yesterday, where I suggested that the NYT could sell advance access to its stories. John Gapper summed it up well, in a tweet: “If scoops don’t matter to most readers, as the digerati claim,” he said, it’s logical to sell them to those who do value them. Which, in this case, would be hedge-funds capable of front-running the news and making a profit when the news moves markets.

In a sense, there’s something very economically inefficient about scoops like this one. The NYT story came out in the middle of the weekend, when markets were closed; when they opened on Monday morning, both Walmart and Walmex had billions of dollars shaved off their market capitalizations, but no one was given the opportunity to short those stocks at their prior level. After many months of diligent and valuable work on this story, one would think that a genuinely capitalist economy wouldn’t just leave money on the table like that. After all, buy-side institutions pay millions of dollars to analysts who research companies like Walmart in depth; isn’t that exactly what the NYT was doing?

For years, short-sellers have briefed journalists when they find out something damning about a company: think of Jim Chanos, for instance, putting Bethany McLean on to Enron and other companies. More recently, a group of people ranging from Mark Cuban to John Hempton to Muddy Waters to Anonymous Analytics has merged the shorting and the reporting functions, putting on short positions before releasing their own research on a company in the hope of seeing that company’s shares fall as a result.

But while the world doesn’t seem to have blinked very much at shorts helping reporters, there’s a much more visceral opposition to the idea that reporters might ever help shorts. If the NYT were to give any hedge fund an advance peek at its reporting, goes the argument, well, that would be bad.

The journalism-ethics angle to this hasn’t really been fleshed out, though. Mathew Ingram, for instance, says that if news is being put out in the public service, then it shouldn’t be “just another commodity”; if the NYT were to go down this road, then “that would make it a very different type of entity than it is now”. It’s all very vague and hand-wavey.

The Epicurean Dealmaker, in the comments to my post, is a bit more on point:

Why in God’s name would you want to give the reporters and editors of the New York Times even more of an incentive to break market-moving news. Surely you know there are many sides to any story; emphasis is critical. Why would news consumers trust editors and journalists who could directly profit by making a complicated story just a little more controversial, by shading facts and presentation to put a company in a worse light, by selectively releasing (or suppressing) information? Newspapers like the Times have always relied on a not quite accurate but nevertheless crucial image of impartiality for their authority. This would disappear if they were seen to be tools of Steve Cohen or Ken Griffin.

The fact is that the reporters and editors of the NYT already have an incentive to break market-moving news. Talk to pretty much any business reporter, and they’ll tell you the same thing: the story everybody wants to get is the scoop which moves markets. If you develop a reputation as someone who can get those scoops with any regularity, you’ll rise far and fast. It’s not uncommon for business news services to even put out charts of a stock price, showing when a story came out and what happened to the price after it did. Those charts can mean real money in terms of new subscriptions, and also in terms of pay rises for the journalists in question.

So journalists already indirectly profit from moving markets, and my suggestion was that the relationship should remain indirect: the NYT would sell advance access to its feature stories as a package, ex ante, just like other high-end news services. If a hedge fund wanted to pay a very large amount of money for that package, then it could then do with the information as it wished. But in any case if the hedge fund wanted information to flow the other way, and wanted to influence the NYT’s journalists at all, then it would have to do that the old-fashioned way, just like Chanos did with McLean.

But the real problem with my idea, it turns out, has nothing to do with journalistic ethics at all. Instead, it’s the insider-trading rules of markets around the world.

I wrote about insider-trading rules back in 2008, and came to the conclusion that while I wouldn’t necessarily implement such laws if they didn’t exist, I’m not a huge fan of abolishing them, either. Certainly there are well-formed arguments why insider-trading laws should be abolished, but let’s ignore the philosophical arguments for the time being: they haven’t been abolished, they’ve been in force since the 1960s, and everybody has to abide by them.

And the effect of substantially all insider-trading laws is to, in effect, ban precisely the kind of thing I’m suggesting, where a small group of people can take advantage of information asymmetries to make money. The way that most (but not all) stock markets are set up, the ideal is a level playing field, where all players get exactly the same information at exactly the same time, and then act accordingly; attempts to act on information before it’s public are criminalized.

One thing that both the ethical and the legal approaches have in common, however, is the concept of “public information”: both of them object to my idea because the NYT is in the business of putting out public information, and giving hedge funds advance access to that information — before the rest of the public gets a look — would in some way be fundamentally unfair.

The concept of “public information” is not a well defined one, and probably can’t be well defined. Certainly it’s not a function of price: once a piece of information hits the Bloomberg wire, it’s public, even if you need to pay Bloomberg $20,000 a year to see it. Josh Benton raises the example of Footnoted Pro, which costs $10,000 a year — but that service is explicitly based on the analysis of public information which is released to and by the SEC. Michele Leder’s product simply provides a smarter way of finding the needles in the haystack of EDGAR filings.

Is a tweet public information? Yes. Is a Facebook status update? I don’t know, but I suspect it probably isn’t. But here’s something which definitely isn’t public information: hours of interviews with a former Walmex executive detailing exactly when and where the company paid bribes.

In the comments to my post, Daniel Davies makes an impassioned argument that what I’m suggesting would almost certainly be illegal in many markets, even if it might be allowed, in some circumstances, in the US. He also says that hedge funds would never pay $1 million a year for this service. These two things are related. The value of advance notice of NYT stories, to a hedge fund, is inversely proportional to the number of other hedge funds who are also getting that advance access. And here I think is one of the key ways to distinguish between public and non-public information.

Let’s say the NYT prices the service at $100,000 per year, and you’re a hedge fund wondering whether the service is worth paying for. It’s way more than you normally pay for public information, so you’re inclined to say no. On the other hand, if everybody else makes the same calculation and you end up being the only fund to subscribe, then at that point the $100,000 might well be worth it: you could make many times that on one trade. The problem is that if you’re the only fund to subscribe, then the information can’t be considered public any more. And if it’s non-public information, then you risk putting yourself in legal jeopardy by acting on it.

In other words, if the information is public then it’s worth very little, and if it’s non-public then it might be illegal to trade on.

This also explains why it’s so common for executives to complain that what short-sellers are doing is illegal. Jim Chanos had information about Enron which was damning, and he acted on it before it was made public by Bethany McLean. That looks like material non-public information to me. What he did was legal, if he didn’t have any insider sources within the company. But McLean certainly talked to a lot of people within Enron, and she was also talking to Chanos all the while. Which raises some legal grey-area issues.

My feeling is that it would be astonishing, in practice, to see an insider-trading prosecution based on information which the New York Times Company had sold on a subscription basis. It’s the NYT’s business to sell information; doing so can’t sensibly be considered illegal. And similarly, once someone has legitimately bought that information from the NYT, it’s a bit crazy to say that they can’t act on it.

Similarly, I’d be equally astonished to see Sharesleuth, Mark Cuban’s operation, ever prosecuted for insider trading, even if they quite explicitly had sources inside the company they were reporting on. Sharesleuth’s model is not intrinsically unethical; the problem with it is rather that the model just doesn’t seem to work. Still, I’m sure that if and when it does work, the company being targeted would try extremely hard to get Cuban investigated for insider trading. And that’s almost certainly a risk that potential subscribers to any advance-news NYT product have no interest in taking.


Comments are closed.