SWF Radar

O’Reilly Media’s first Money:Tech conference kicked off at the Waldorf-Astoria in New York today. Several presentations threw some interesting light on where financial research is going in the world of Web 2.0 and RSS feeds.

I am attending the conference. Here is a report on the highlights of the first day. (My report on the second day is here.)

> Jim Cramer of TheStreet.com

After the opening salvo by Tim O’Reilly and Paul Kedrosky, the first presentation was made by Mad Money’s Jim Cramer, introduced by Kedrosky as the “first professional blogger” (on account of his founding TheStreet.com in 1996).

Before falling into his Mad Money persona toward the end of the presentation, Cramer had something interesting to say about sell-side research, namely that it’s dead, killed by Google. Ad hoc research calls, he gave as an example, produce no useful information.

He’s a big fan of blogs, though: “If Implodometer is here,” he said, “God love you, you made us a fortune.”

He only had good things to say about Bloomberg and its information services.

He added that, “Everyone is in the business of denial on Wall Street.”

> Larry Tabb of TABB Group

TABB Group’s Larry Tabb gave a presentation focused not on the Web but on the shift from high-touch trading channels (phones and brokers) to low-touch channels (computers). This shift, he showed, is happening quickly: 79 percent high-touch in 2004 compared with 37 percent now. At the same time, he said, commission rates and broker revenues are plummeting.

This is leading to the creation of “dark pools” and “crossing networks” in which brokers attempt to match order flow on its way to the exchange. He said that 37 of these are currently being developed in the US, and that stock markets are already starting to lose market share to them.

Tim O’Reilly asked him whether dark pools could one day end up analogous to the Colorado river, siphoning off all potential trades before they reach the exchanges just like water in the Colorado river apparently never reaches the sea.

Tabb answered that the SEC would certainly step in as early as at the 15 percent mark to make sure such a situation did not arise, since price discovery would end up being impacted and mom-and-pop order flow would end if crossing networks took over completely.

> David Leinweber of Berkeley

In the final session before lunch, David Leinweber, now of the University of California, Berkeley, gave an overview of the impact of technology on financial markets. He presented a graph showing how, “after the introduction of the Netscape browser” in the mid-Nineties, reaction time to bad corporate news (reflected in a completed fall in the share price) dropped from around two weeks to a few minutes.

He said that by looking in the right places on the Web, you can gather better information than that presented in the mainstream media.

> Steve G. Steinberg of Steinberg Consulting

Steve G. Steinberg, a computer programmer, described some of the clever ways he has been producing actionable intelligence for his Wall Street clients over the past decade. I write “actionable intelligence,” as Wall Street’s gain was clearly the NSA’s loss: The techniques he described include setting up hundreds of vendor accounts on eBay to monitor pricing trends, tracking serial numbers of iPods and their parts to work out how many have been manufactured (a technique the Allies apparently used on German weapon parts in World War Two for the same purpose), making leftfield second-order inferences such as attempting to estimate Chinese Internet penetration by creating filters to monitor the level of spam emanating from China, and black-hat methods, such as cookie manipulation and the like.

> Eric Christiansen of Barclays Global Investors

Eric Christiansen of Barclays Global Investors (BGI) said that analysts are already drowning in data, but that we’re still seeing only the core of all possible data. He suggested that, furthermore, only a small subset of visible data “gives us alpha,” with the rest merely enabling us to stay competitive.

He said that if you find something that really works in terms of delivering particularly valuable information, it won’t last. And he said that you’ve always got to look for new sources.

> Robert Pessarella of Bear Stearns

Robert Pessarella of Bear Stearns then gave a presentation describing how his chosen premier sources of financial information have changed over the recent past. A few years ago he would read the Wall Street Journal and maybe the Financial Times on the train to work, then check Bloomberg.com, Dow Jones and Reuters when in the office, with perhaps a glance at CNBC. Now, he said, everything revolves around RSS and Web 2.0 (a term, incidentally, that the conference organizers, O’Reilly Media, invented), with Google, Yahoo, Technorati, Wikipedia and even Amazon and eBay being the first ports of call in his search for information, with the service providers Connotate, Capital IQ, Gerson Lehrman and Majestic Research (all conference sponsors) helping clean and sort the information.

He seemed a little more sceptical about these companies, though, when he then chaired a panel discussion featuring some of them. He wondered how they hope to compete with free-service providers and issued what looked like a light warning that if their services did not open up more, he’d look elsewhere for his information. Two of the panelists suggested they could compete with free by employing lots of people.

Pessarella finished the session by stressing how important it is “to live in the news flow.”

> Bo Cowgill of Google

Bo Cowgill of Google gave an interesting talk about how the company has for the last two and a half years been running internal prediction markets, using play money, in such areas as demand forecasting, performance and industry news.

It turns out that newer employees tend to be more optimistic than old-hands, that people tend to be influenced by those who sit near them, that the percentage of optimistic forecasts tends to increase the day after the share price rises and vice versa, that things good for Google tend to be overpriced, with bad things underpriced, that the forecasts tend to be around 90 percent accurate, and that those on the periphery tend to be more accurate in their predictions than those nearer the center.

Optimists, he added, tended to lose money, while, in Google play-money terms at least, pessimists tended to get rich.

More on the second day of the conference tomorrow…