The GP's claim was that: > The rich are simply not smart enough to intelligently allocate their capital. [...] You don't see big investments in new technology,

The GP's claim was that: > The rich are simply not smart enough to intelligently allocate their capital. [...] You don't see big investments in new technology, science, infrastructure, or other complex and difficult areas because by and large the financiers do not understand these things. To me, this is saying: tech, science, and infrastructure and undercapitalized relative to their expected returns. It is very difficult for me to construct a model in which a) capital is foolishly providing too little capital to certain sectors relative to the returns it should get, and b) those sectors don't provide outsized returns to the capital they do get. Of course, such a model is possible! It's possible that returns are random but allocations are still irrational. But what does such a model look like? One possibility is that returns are totally unrelated to the underlying performance of an asset. For example: new technology is a more rational allocation of capital, and would provide greater return, but it doesn't, because return is totally disconnected from the performance of the technology. To me, that possibility seems quite unlikely. Yes, the anti-EMH position holds that prices frequently depart from their underlying value--but the idea that there is zero connection, as would be required by such a model, is not tenable. It's absolutely clear that real-world events have effects on asset prices. If, say, Oculus Rift becomes extremely popular with consumers, the capital that invested in it will receive large returns. Another possibility is that there are network effects that make it impossible to make the excess returns from these sectors unless the rest of the financial sector decides to invest in these sectors as well. But it's very difficult to see what those network effects might be, and it's even more difficult to imagine that they are present for all productive investments in these sectors. Unpredictability of asset returns is the primary testable prediction of the EMH. On the other hand, it is difficult to construct a reasonable anti-EMH position that incorporates the observed unpredictability of asset returns. Finally, if the claim that the market is ignoring great opportunities because of stupidity or irrationality does not imply that there are excess returns available on investments in those great opportunities, it's frankly difficult to see what the claim means. If I say "you should invest in my startup, because it's a more intelligent allocation of your capital than real estate--but no, I do not expect that you'll earn higher returns from my startup than you would from another investment," it's pretty difficult to understand what I actually mean by "more intelligent allocation."