I'm pretty sure the Weak EMH implies P = NP direction of that one is wrong because he's redefined the semantics of placing orders on the market in such a way that brokers need to solve an NP-complete problem to decide transaction prices, but it's subtle and I'm not sure and need to think about it more. Applying computational complexity to economics is an interesting idea whether or not this application of it is valid.
Speaking of subtly wrong but interesting approaches to P vs. NP, there's another paper on arxiv (which I couldn't dig up the link to easily and am short on time at the moment) that argues P = NP because soap films minimizing surface area solve NP-complete problems and you can simulate the underlying physics in polynomial time in the system size, but he never justifies the assumption that the corresponding physical system settles to equilibrium in polynomially many time steps, and I suspect on thermodynamic grounds that this is false; the energy gap between the lowest and next-lowest energy configurations decreases exponentially in system size.
no subject
Speaking of subtly wrong but interesting approaches to P vs. NP, there's another paper on arxiv (which I couldn't dig up the link to easily and am short on time at the moment) that argues P = NP because soap films minimizing surface area solve NP-complete problems and you can simulate the underlying physics in polynomial time in the system size, but he never justifies the assumption that the corresponding physical system settles to equilibrium in polynomially many time steps, and I suspect on thermodynamic grounds that this is false; the energy gap between the lowest and next-lowest energy configurations decreases exponentially in system size.