$300M To Save 6 Milliseconds
whoever57 writes "A new transatlantic cable (the first in 10 years) is going to be laid at the cost of $300M. The reason? To shave 6ms off the time to transmit packets from London to New York. The Hibernian Express will reduce the current transmission time — roughly 65 milliseconds — by less than ten percent. However, investors believe the financial community will be lining up to pay premium rates to use the new cable. The article suggests that a one-millisecond advantage could be worth $100M per year to a large hedge fund."
If you're in London and you know 6ms before anyone else that the price of oil in New York just shot up, you can buy oil right now and then sell it in 6ms for a tidy profit.
I heard some european head of state (Sarkozy perhaps) suggest that stock transactions be taxed based on speed, i.e. speculators who buy and sell very fast to make a quick buck get taxed a lot, but real investors who're in for the long run and keep their stock for a long time don't. That sounds like a great idea to me.
This concept actually was first proposed in 1972 by Nobel-winning economist James Tobin, with the idea that it would apply to currency transactions to prevent speculators from rapid trading like the kind you're describing. Basically, the concept is that with such a tax in place, traders would have to hold onto the asset long enough that they could pay for the tax, plus whatever gains they were anticipating, so that meant that they'd have to expect to own something for longer than a few minutes. There have since been discussions of applying the same idea to stocks, bonds, mortgage-backed securities, and other assets.
The purpose of that tax isn't so much to generate revenue (although this definitely would happen), it's to slow down the markets enough so that the assets could be properly valued rather than people making money on millisecond-level differences.
I am officially gone from