>But they trend toward efficiency over time The trouble is that, even if this is true, there is no theoretical upper bound on how long it can take. If "markets

>But they trend toward efficiency over time The trouble is that, even if this is true, there is no theoretical upper bound on how long it can take. If "markets trend to efficiency" means "don't worry, wait a few years and this inefficient market will be efficient again", then that's one thing. If it could take a couple of hundred years, that's quite another. As Keynes famously put it: > In the long run, we're all dead. Economists set themselves too easy, too useless a task, if in tempestuous seasons they can only tell us, that when the storm is long past, the ocean is flat again. There's an interesting analogy with non-Turing-complete programming languages that guarantee program termination. In the real world, a guarantee that a program will terminate eventually is often not particularly useful. What we really want is a guarantee that it will terminate within some reasonable period of time.