[nasmorn]

That is funny because the proofs had the opposite effect on me. It dawned on me that the math of e.g walrasian price setting is surely correct but the story that this is supposedly how our economy works was much less believable after each supposedly logical assumption (free disposal, no money pump) could be traced back not to psychology or physical realities but conditions for the equations to remain solve able.

Looks to me like a cryptologist predicting what passwords people would choose based on the needed computing power to crack them. I wonder if 'password' made the list.

[rdlowrey]

I tend to agree with your position. The fundamentals of individual microeconomic decision-making are much more sound IMHO than their extrapolation to large-scale macroeconomic systems. At the macro scale I believe the system is far more complex than we can correctly model with anything but rough approximation. Otherwise, we'd already know exactly when/why/how future recessions and booms would occur. The High Frequency Trading question, though, exists entirely within the micro sphere where the math is rock-solid. As such, I've yet to encounter a logical refutation for its use.

[nasmorn]

This is one of my favorite papers explaining possible problems with arbitrage. http://www.math.mcmaster.ca/~grasselli/ShleiferVishny97.pdf It is not directly applicable to HFT though.