09 December 2014

Widget Wonderland

Three semi-related posts went past my eyeballs today. The thread which hangs them together is the value of value. How do we figure it?

First, Blankenhorn takes on Amazon, among my favorite fubar companies. As is often the case, the comment stream is as significant as the post. The poster child for the "we lose money on each sale, but make up for it in volume" approach. Either Amazon attains monopoly, i.e. pricing, power over retailing or it goes belly up.

Second, one Peter Greulich (not familiar) takes on IBM. Here the issue is both the danger of goodwill and the structure of companies. How to value IP, services from IP, and goodwill (e.g. the sale of WhatsApp)? IBM has been shedding real capital and real production for many years. It wants to see the ROI of software driven services. The WhatsApp home run as paradigm for Big Companies, if you will. So do many corporations. Good luck.

Third, Rob Hyndman takes on Data Science. What is it? Who does it? Is it just stats with a new name? Is it just OR, with a new name? Is it more, in some sense, than just stats (Janert says so)? Likely.

Taken together, some musings on the value of value. Micros and quants invariably practice reductio ad absurdum (in the second sense, and on themselves) by looking only at the financials and ignoring the real world aspects. How else would they have blessed liar loans? If quants and micros took the "science" part of data science seriously, may haps we wouldn't go through Great Recessions? One of the comments on the Hyndman post is on that point.

If science viewed thought as property, we'd still be cavemen.

No comments: