It’s the Data, Stupid
So Thomson Financial has finally done something that truly reflects the age - using computers and software to churn out commodity-type stories in the wake of U.S. companies’ earnings releases. They have the historic data, they have the “training set,” and it was only a matter of time before they would connect the dots and give customers what they really want - basic accuracy with lightning speed. In short, to give customers the data they require, in real time (well, 0.3 of a second is pretty good), in order for them to use their expertise to make a decision. I knew this day would come and applaud Thomson for using available technology to manage workflow and optimize scare resources - the time of their premium reporters and writers.
I had an “ah ha” moment concerning this issue when visiting a leading Boston-based VC about nine months ago. This VC, knowing the kinds of things I was interested in, dropped a 50-page research report in front of me and said “this was produced by a machine.” Wow. I scanned it, and it looked pretty good. Clearly the software had been trained to understand the writers’ style and diction, not to mention having an appreciation of historic context from previous analyst reports. When given new data, the machine was able to craft a report in the style of the analyst that provided the facts and added commentary in a very straight-forward manner. Interesting. But the fact is, reporting the news and providing commentary is a commodity, and precious few analysts add enough value to warrant the buy side shelling out significant dollars for their services. Thomson’s recognition of this fact is one of the reasons I find their move worthy of praise. What the VC showed me was an interesting techno-trick, but to my way of thinking way off-point. Leading analysts and traders want to make their own decisions - “give me the data and I’ll decide what to do with it” is the mantra I’ve heard time and time again.
At the end of the day it’s clear - smart investors want data. They want interesting, unique data, and they want it fast. Not the same junk everybody else gets, unless it is delivered markedly faster than the speed at which others receive it. This angle is clearly a race to the bottom, as ever-more sophisticated streaming data providers, faster chips and better hardware are splitting milliseconds into ever smaller units to try and create an edge. This is a limit that will soon be reached. This is not a game I would choose to play.
Thomson is giving traders and analysts what they want. But, in fact, they want more. Earnings reports and their interpretation are a necessary but insufficient data point for managing a portfolio. What they are providing is called a commodity.
So where is the edge? Different data, in real time, is what will create the edge. It has to be fast. It has to be different. It has to be relevant. It has to be discoverable. Without these things it will fall into the vortex of the thousands of unread emails or the stack of papers sitting on the left hand side of your desk. So Thomson got it right that it’s the data, stupid. But are mechanized earnings reports going to change the world? I think not.
