I consider the emerging dynamics of the *interplay* of all human decisions as part of "nature", and not as a human decision itself.  Speculative value is at the same time a guess of the outcome of that interplay, and an input to the dynamics itself.
My point is that this complex dynamics is, apart from some general principles, essentially not predictable.  It is a more global version of the efficient market hypothesis in a way: all the obvious is already taken into account, and the rest is essentially entropy to us, which we can only guess, and sometimes, be lucky, and make us believe that we understood something, while we only played at the lottery and won.