When I’ve written about the pernicious influence of high-tech finance, I’ve tended to emphasise the brain-drain and the flim-flam (the destructive influence on capital that gets drained into the pockets of the quants, and the embarrassing perversion of research priorities in mathematical sciences). But capital flows have no effect until they are realised in real-world allocation of resources. This is why I’ve been so dismayed by recent reports on investments being made in new fiber-optic lines connecting New York with Chicago and London. According to Businessweek private firm is spending $300 million to lay a transatlantic cable that is slightly shorter than the existing line, allowing round-trip transmission times to be reduced from 65 milliseconds to 59.6 milliseconds. And Forbes reported a few years back that another company was spending $300 million to reduce the New York-Chicago latency from 16 to 13 milliseconds.
What’s the point? Well, it’s pretty obvious that getting information quicker enables you to profit. Mark Twain was writing about this more than a century ago: In “Cecil Rhodes and the Shark“, where a young Cecil Rhodes makes his first fortune speculating in wool, based on foreknowledge of the Franco-Prussian war, gleaned from a newspaper found in the belly of a shark in Sydney, 50 days ahead of any ship-bound news reports from Europe. Making money off 3 milliseconds rather than 50 days is technically more demanding, but in principle no different. Equities are traded in New York, and futures contracts on the same equities in Chicago. If the price in New York rises, the price in Chicago must rise as well, but only 16 milliseconds later, after the information has reached the traders. So if you have the information about the rise after 13 milliseconds, you have 3 ms to buy the futures and then cash out the tiny profit before the next wave. “This is all about picking gold coins up off the floor–only the fastest person is going to get the coins.”
So, here you see the cost of high-tech finance, the cost of the economic distortions produced by the vast quantities of money leaking out of the markets: no longer just the minds wasted that might otherwise have been engaged in productive activity, but real productive capacity being sunk into construction as ridiculous as any Stalinist prestige project. At least when Stalin had marble transported from Central Asia to build the Moscow Metro, average Muscovites actually got efficient mass transit they could use. The cost of this fibrous folly is simply being extracted from all parties to the financial markets, using the speed of light as a weapon of extortion.
Although this raises another question in my mind: Suppose I know that company X has a special fiber-optic line, meaning that it has superior information to what I have. I’ll have to assume, then, that any transaction they propose is going to be a bad deal for me. So I wouldn’t want to trade with them. Presumably you don’t get to choose your counterparties on an exchange, but I don’t see that it would be impossible for exchanges or government regulations to ban particular communication technologies, treating them as a kind of insider trading. And given that such a superfast line would effectively serve as a kind of tax on high-speed traders, and as a way for high-speed traders to bilk slower traders, it’s heard to see why anyone on Wall Street would object. Anyone who buys access to the new cable will have to know that all their competitors will be compelled to buy the same access, and then everyone is worse off, except the people who own the cable.
Communication speed is a positional good, just like social status. It doesn’t really matter (up to a point) how fast you get the information, only whether you get it faster than someone else. Building the new cable is like standing up to see better at a concert: Pretty soon everyone is standing, and no one can see better.
Except that standing up doesn’t cost $300 million.