I think that, sometime soon, people are going to stop making predictions.
Actually, tragically, that isn’t likely to happen. Human beings are seemingly irrepressible predicting machines, in that it is a way to sound intelligent without anyone being able to immediately call bullshit. By the time the future arrives, most people don’t care enough to back-test you. I was jolted into reconsidering this issue by the recent news from England, that they are going to go into austerity mode in order to reduce their national debt. Personally, this feels smart to me, in a “I’m no expert but isn’t balancing your books a good idea?” kind of way, but you can definitely find any number of economists violently committed to either side of the debate.
This is an old story, but it bear repeating: At the start of the year, a stockbroker sends a letter to 1,000 prospective clients, each containing his recommended “best pick” for the coming month. He sends a different ticker symbol to each of 10 groups of 100 prospects. In February, after 5 of the 10 stocks perform above average, he targets those 500 remaining prospective clients, further dividing them into 5 groups and repeating the trick. By June, assuming the same level of performance (ie, 5 of 10 picks perform above average), he has 30-odd prospective clients who think he’s a genius. Rinse and repeat, and you’re in business.
How different is that from Roubini and the rest of the economists out there? The bulls were right for the better part of a decade, until they were horrifically wrong, and the bears have famously predicted 7 of the last 2 recessions. Maybe it was always thus, or maybe it’s newly true, but I think at this point we can conclusively say this: the world is too complicated to predict. There are too many variables to fit in any one model. Just focus on what you know, do a good job, try to add value, say/do something novel, don’t listen to the experts, and you’ll be fine.