Is the smoke clearing for expert networks?
From a modest hotel room in the Omni Shoreham in Washington DC, the SEC gathered a smattering of reporters, lobbyists, and others for a best practices seminar. But the somber title belied a dramatic observation to be made by an SEC official on expert networks. They’re not the problem.
Carlo di Florio, director of the U.S. SEC Office of Compliance Inspections and Examinations, spent just over six thousand words on reforms made under Chairman Mary Schapiro, the implementation of the Dodd-Frank Act, the focus on examination and training, and various enforcement actions in the advisory community. And then, almost 6000 words in and nearing his final remarks, he decided to “briefly mention the ‘Expert Network’ insider trading cases that the Commission and the Department of Justice have recently brought, and that have received much recent press coverage.”
Contrary to some reports that I have seen, I believe these cases do not represent some inherent hostility by the Commission toward expert networks
Expert networks are not the problem behind the insider trading investigation. The problem is bad behavior, and he sought to clear up “some urban legends and misunderstandings” that had developed. The secret behind those implicated in the recent insider trading cases: bad actors, doing bad things. Rather than condemn expert networks, di Florio sensibly directs the discussion toward how to use them. Investment advisers need to focus on training and policy: “it underscores the need for advisers to have reasonable policies to prevent insider trading.” The insider trading scandals weren’t brought on by expert networks. They were brought on by the incipient lack of controls and policies of those implicated.
di Florio proposes two sets of controls and divides them across the front-end and the back-end. The front-end focuses on the engagement of the network. If an adviser is going to use an expert network, they need to understand their own insider trading policies, for example, so they mistakenly seek out material nonpublic information. No one wants an analyst who does nothing but ask everyone for the numbers, but it’s even worse if they don’t know that they’re crossing the line. Given the allegations against Primary Global Research, an expert network and proprietary research firm implicated in what appears to be a systematic conspiracy to sell inside information, one would expect all investment advisers to interest themselves in the policies, procedures, and general probity of the firm with which they are dealing. di Florio also recommended additional controls for engaging experts who are employees from public companies – from outright avoidance to additional controls, such as a chaperone, perhaps.
The back-end focuses on the application of an expert network by the adviser. It entails monitoring. It’s the adviser’s responsibility to monitor the research activities and trading decisions of their employees. di Florio recommends a handful of potential tactics: random screening of trades following conversations with outside experts; comparing trades with coincident events, such as 8-k filings, earnings announcements, and press releases; reviewing trades in personal accounts.
The CFA Institute commented on the speech and shared a reserved sense of relief on di Florio’s recommendations. They echoed earlier sentiments expressed by Preet Bharara and confirmed that the SEC’s concern rests with the breach of confidentiality and delivery of material nonpublic information. Both expert networks and the mosaic theory remain an essential element of the investment research tool-kit.
di Florio’s comments brought much needed clarity to the discussion around insider trading, expert networks, and mosaic research in general. While the speech is footnoted to reflect that the SEC “disclaims responsibility for any private statements by its employees,” it does suggest the outlines of some safe-harbor for expert networks and mosaic research.
But investment advisers must go further. They must ask themselves if di Florio’s front and back-end controls should also apply to all engagements with experts, not just those organized through an expert network. Those calls and engagements that don’t go through a network, after all, would fall outside of the systems, procedures, agreements, and paper-trail of an expert network.
Is the corollary of di Florio’s remarks, therefore, the assertion that primary research would be better managed through the framework of an expert network? That any conversation with an outside expert would benefit from the systems, procedures, agreements, and paper trail of an expert network?
It’s perhaps beyond the scope of his comments, but it is nonetheless, worth asking.