Obsessing over strategic geography has a rather… twentieth century feel to it. Few now worry about the control of the Suez Canal, or the rights of warships to traverse the Bosphorus; far-flung scraps of land once valued as coaling stations and choke points are now important chiefly as tax havens and political distractions, and the various growths of Railway Imperialism have largely decayed back into the soil on which they were imposed. But there are a couple of areas that still pursue this approach to life. One, of course, is the subject of pipeline politics, amply discussed by m’colleagues, for example, here. Or here. Or here.
The other doesn’t get quite so much attention:– it’s fibre. Laying a pipe to move bits is a lot easier than laying a pipe to move natural gas, though it’s still no walk in the park. (If you have the spare time, here is an extremely long article on the subject by Neal Stephenson, the well-known writer of extremely long things.) But it’s not so much the expense of laying the fibre, it’s the limits that geography places on the most demanding users of that fibre. And that’s the traders.
This article from 2006 gives you a bit of background. Basically, algorithmic traders – using software to execute their trades at high speed – are at the point where the lightspeed lag is a significant factor in their profitability. Light travels at 300 000 000 metres per second (as you know, Bob), or 300 km per millisecond. And it turns out that algorithmic trading programs are fast enough that a few milliseconds delay, or latency, can have dramatic effects on profit – the (no doubt slightly hyped) figure is $100 million per millisecond.
The European reference is this: we are sitting on a small continent, most of whose major centres of population are pretty close – certainly most of the centres of finance are close. As a rough measure, NY-Chicago is almost twice London-Frankfurt, for example; east Asia, of course, is even more spread out. We also (well, some of us) have the friendly North Sea to lay cables through – no problems with very deep water, no unstable geography, no problems with intervening mountain ranges or unfriendly nations. I think this could be one of the neglected advantages that geography gives Europe (along with all the others) – realtime, high-bandwidth data services are only going to expand in the future, and that means smaller delays will be at a premium. Shipping data off to cheap centres in India is all very well for some applications, but for fast interaction you need as close to co-location as you can get, and a European country with lots of cheap, clean electricity and lots of cold water to cool data centres with will do rather well. Possibly one for the Nordics to look at, when the oil starts to run out?