A recent Facebook conversation made me want to post this idea I've had in one form or another for the past few months.
First, at an abstract level. When you get above the level of menial labor, a lot of the work of big organizations consists of conversations. Conversations in board meetings. Conversations in merger negotiations. Conversations in which someone is trying to sell something through persuasion. Conversations about strategy. Conversations about divisions of responsibility. Conversations that end in signing written contracts, or that comprise oral contracts. Conversations in classrooms, with clients, with consultants, in which a well-informed party is informing a poorly-informed one. Conversations in print, in academic journals or corporate financial reports. Some of these conversations have a very high dollar value. Why? What is worth paying so much for? For now let's just say quality in some sense. Conversations differ in quality, from stupid to lucid, from false to true, from irrelevant to crucially urgent, from vague to clear, from rambling to disciplined, from rude to polite, from trivial to very important.
Now, notice that the internet has created enormous new modalities of conversation. E-mail. Chatrooms. The blogosphere. Podcasts. Skype. Social networking sites. And sometimes the conversations that occur in these new contexts are very high quality. I noticed in college that chat, in particular, was sometimes better for deep, personal conversations than talking face-to-face. The reason was that you could write something and see how it sounded first, and then delete it, so you were more willing to take risks and go out on a limb. Since then, I've noticed that the blogosphere is often excellent for high-level intellectual discussions, sometimes better than academic journals. Of course this is very hard to measure, but my sense is that the quality of conversations that take place for free in the blogosphere, on Facebook, and in general through new media, are generally very often superior to the quality of conversations that take place in more traditionally structured environments where capital allocation decisions are made, and that there is an opportunity, so to speak, for "arbitrage"-- for buying, channeling, re-incentivizing various kinds of online conversations so as to inform decision-making and improve it, while cutting the costs of decision support.
If this is such a good idea, why hasn't it already been done? Well, it has, sort of, in "expert networks." So I want to study expert networks more, figure out what's gone well or badly and how they're likely to evolve. Interestingly, expert networks seem to be at the heart of a recent major scandal on Wall Street. The Wall Street Journal reports:
Hedge funds didn't wait for a guilty verdict in Raj Rajaratnam's insider-trading trial to tighten controls on the use of so-called expert networks.
The charges against Mr. Rajaratnam, founder of Galleon Group, and the testimony at his trial shed light on both a shady web of contacts at some prominent companies and on expert-network firms, which connect hedge-fund employees with industry specialists. He was convicted Wednesday on all 14 counts of securities fraud and conspiracy.
Mr. Rajaratnam's lawyer has described his client's strategy of researching stocks through a private network of individuals as building "a mosaic of public information about the companies he follows."
Hedge-fund managers, however, are re-examining how they develop and use expert networks, says Guy Talarico, chief executive of Alaric Compliance Services, a New York-based consultancy that advises private-equity and hedge funds on compliance matters.
"I'm going to assume that there will be a real decrease in the use of expert networks, because people will not want to run the risk of that exposure," he says.
Many hedge funds have already implemented more rigorous procedures, Mr. Talarico says. They are more carefully reviewing the experts' credentials and the training they have received on insider trading, he says. They are requiring compliance officers to listen in on calls between experts and fund employees and obliging the expert to agree to a fund's insider-trading restrictions, he says.
Still, Ron Geffner, a hedge-fund lawyer for Sadis Goldberg LLP in New York, says that Mr. Rajaratnam's conviction "shouldn't be viewed as a conviction of expert networks, but rather practices by certain individuals that went beyond their intended use."
"This is not the trial of expert networks, although it may seem this way," says Mr. Geffner, who is also vice president of the Hedge Fund Association, a trade group based in Aventura, Fla.
The Economist wrote about expert networks last June:
THE inside information she gave him was “absolutely perfect”, a former hedge-fund executive testified on June 6th. Winifred Jiau, a consultant for Primary Global Research, an “expert network” firm, is on trial for peddling confidential information about companies to traders.
Ms Jiau’s industry is also undergoing its own trial, of sorts. Expert networks are matchmakers that link clients with experts. A hedge fund that trades pharmaceutical stocks, for example, might use an expert network to find a doctor who can explain how a new cancer drug works. The network would set up a phone call and pay the doctor handsomely.
Such networks have recently caught the eye of American regulators, who fret that investors may be using them to ferret out illegal inside information. Since November more than a dozen people with links to expert networks have been arrested for insider trading.
The hullabaloo has spooked the industry. Some firms, such as Primary Global, may not survive. Others, such as Gerson Lehrman Group (GLG), which has not been accused of any wrongdoing, are looking for a way to distinguish themselves from tarnished rivals and win back business from jittery clients.
GLG is by far the largest network. In 2008 it generated two-thirds of the industry’s global sales of $433m, according to Integrity Research Associates, a consultancy. As clients switch from smaller networks to ones with stricter safeguards, GLG could grab an even bigger share of the pie.
But the business will be less lucrative than before. GLG’s clients spend on average $170,000 a year to subscribe to its database of 300,000 experts. This yields rich profits. But some hedge funds are suspending their use of such networks, for fear of falling foul of the law. Others are making their traders jump through legal hoops before allowing them to speak to an expert. A few are abandoning networks altogether, and finding their own experts. “It would be odd to think that doing research on your own or finding someone online and sending them a cheque is better than the framework we provide,” grumbles Alexander Saint-Amand, GLG’s boss.
GLG has invested in technology to help clients search its database of experts. Compliance chiefs can set filters and block some experts if they worry there may be a conflict of interest. GLG keeps a record of who talks to whom, about what, down to the minute, which they could easily hand over to prosecutors if asked. “If you want to sell inside information, this is a terrible place to do it,” says Mr Saint-Amand.
But there is still scope for things to go awry. Professionals can sign up for GLG without a background check. They have to sign an agreement not to share any material, non-public information, but GLG does not listen in to the calls to enforce this. Clients who are determined to do wrong probably can. For example, they can meet experts through a network and then set up a separate “consulting” relationship, which enables them to buy tips without GLG ever knowing.
Compared to expert networks, the model I have in mind is considerably more public. I'm thinking of the Wikipedia model applied to business. Firms jump in with a question, participants offer answers, firms give spot awards for good ones. Firms that are known to be good tippers get better advice, so that have an incentive to pay. Participants can start out "on spec" as it were, giving it their best shot and then hoping for handouts, but those with a history of saying smart stuff might bid on a question: "I'll comment for $1,000," or something. There might also be a rating system, where only experts that had beat certain metrics-- total volume of spot awards, total volume of thumbs-up from commenters, or perhaps resume details like years of experience, education, etc.-- would be authorized to comment. And the whole thing would be out in the open. The disadvantage of this is that if a company wants to get an edge on its competitors, wikiconsulting might not work. Their competitors could see it. But I think a lot of firms face questions that are very customized to their particular lines of business and they might not be hurt a lot by doing their consulting in public.
An interesting question is how this would affect insider trading laws. If wikiconsulting became important, a well-reputed expert's thumbs-down would likely affect the stock price. They could exploit this by, say, shorting a stock, then pooh-poohing the company's latest idea, then covering the position at a discount after the share price takes a hit. The SEC might be able to mitigate this somewhat. The honor system would probably have to do some of the work too.