Insider Trading Versus Savvy Investing

Lately, legal action has been heating up against investors for alleged insider trading. As a result, now is a good time to consider what sorts of research practices should be considered improper, as it might be an important question in some of these cases. That's the subject of Andrew Ross Sorkin's column today in the New York Times. He ponders whether all not-explicitly-public tidbits an investor might learn about a company should be considered inside information. They shouldn't, as a better criterion can be devised.

In the column, Sorkin refers to something that investors sometimes rely on called mosaic theory. The idea is that if you get lots of pieces of information and put them together, you might have a clear picture that tells you whether or not to invest in a company. For example, if you talk to a firm's suppliers, creditors, customers, managers and regulators, then you will probably be able to determine how it's doing better than someone who just read its financial statements.

The problem that Sorkin points out is that not all that information is public. He turns to on an example of talking to a manager at the Gap to determine how the retailer is doing. If that manager tells you something that the company hasn't publicly stated, isn't that nonpublic information that shouldn't be used for trading? It depends.

If this piece of information is something so utterly impactful that could move markets, then acting on it would obviously be insider trading. For example, if that Gap manager said that he's worried about his job because management is considering selling the company to Banana Republic, then that's clearly material. But if the manager says that sweaters are selling better this year than last winter at his store, then you probably can't immediately determine whether or not to buy or sell the stock based on that. Indeed, this conforms to what Sorkin writes:

Materiality, according to (law firm) Fried Frank, is "information that a reasonable investor would consider significant in deciding whether to purchase or sell a company's securities."

Yet that seems like a pretty awful criterion to me. I think it would be hard to find much information that isn't in some way significant in deciding whether to purchase or sell a company's security. Indeed, the whole point of mosaic theory is that information adds up. So how do you clearly differentiate a kernel of information that an investor gathers as legitimate mosaic detective work? I would rely on two tests.

First, is that information explicitly nonpublic? The easy example here is an acquisition. If the company you work for is mulling acquiring another, then that process is meant to be kept under wraps until a deal is publicly announced. But there are many pieces of information that happen to be nonpublic only because they aren't important enough to management to include in their quarterly reports. This information may be nonpublic, but isn't kept that way by design.

Second, was that information obtained through unique circumstances? For example, if an investor's aunt happens to work for a company he is researching and she tells him something that no other investor can easily learn -- even if it's not explicitly nonpublic -- then that stretches the bounds of fairness. But if any investor could call up an employee of her company and easily learn the same piece of information, then it's just a matter of doing the work, not having special connections to management.

Just because an investor happens to learn something that not every other investor can glean from a company's filings doesn't mean material, nonpublic insider information must have been provided. Going back to the Gap example, let's say the manager tells investor X that sweaters are selling quite well this season in all of the stores in the northeast region. Now let's say that investor Y asks the same question a day later and the manager provides the same answer. This is information is not meant to be kept secret by corporate management and is transparently provided upon request; thus, relying on it should not violate securities law.