Thursday, April 12, 2012
FINRA and the SEC announced fines and a censure of Goldman Sachs for supervisory failures relating to research "huddles." Under the settlements Goldman will pay $11 million each to FINRA and the SEC. According to the FINRA release:
In 2006, Goldman established a business process known as "trading huddles" to allow research analysts to meet on a weekly basis to share trading ideas with the firm's traders, who interfaced with clients, and, on occasion, equity salespersons. Analysts would also discuss specific securities during trading huddles while they were considering changing the published research rating or the conviction list status of the security. Clients were not restricted from participating directly in the trading huddles and had access to the huddle information through research analysts' calls to certain of the firm's high priority clients. These calls included discussions of the analysts' "most interesting and actionable ideas."
Trading huddles created the significant risk that analysts would disclose material non-public information, including, among other things, previews of ratings changes or changes to conviction list status. Despite this risk, Goldman did not have adequate controls in place to monitor communications in trading huddles and by analysts after the huddles.
Goldman did not adequately review discussions in the trading huddles to determine whether an equity research analyst may have previewed an upcoming ratings change. For example, an analyst said of a particular company in a trading huddle in 2008 that "we expect companies with consumer and small business exposure to be under pressure in the current environment, including [the company]." The next day, the analyst sought and received approval to downgrade the company from "neutral" to "sell," and to add the stock to Goldman's conviction sell list. Goldman published an equity research report making these changes that same day.
Goldman also failed to establish an adequate system to monitor for possible trading in advance of research rating or conviction list changes in employee or proprietary trading, institutional customer, or market-making and client-facilitation accounts. Accordingly, Goldman failed to identify and adequately investigate increased trading in proprietary accounts in advance of the addition of securities to the firm's conviction list, certain transactions effected in an account in advance of changes in published research that warranted review based on their size or profitability and/or atypical trading for that account, and certain spikes in trading volume that immediately preceded the addition of stocks to the firm's conviction list.