Cowen's argument shows a part of ABCT that is not sufficiently explicit and as long as it won't be explicitly microfounded it won't convince many economists.
A similar argument was proposed by Lachmann against Mises in 1943. Mises answered that error clusters were an empirical generalization, i.e., he considered the criticism to be irrelevant (cfr Garrison in "Time and money") because the alternative was too unlikely.
However, this implies that the theoretical link between monetary expansion (in terms or M or MV or whatever, it's immaterial) and error clusters is not theoretically well founded, but only a likely empirical assumption.
Now, I find it more convincing to neglect ratex because they are unlikely instead of neglecting an obvious fact because it has not been shown to be apriori necessary, but the problem exists anyway.
Let's try an answer based on four elements:
- Mises's conception of money as a coordination device ("The nonneutrality of money")
- Carilli and Dempster's prisoner's dilemma model.
- O'Driscoll's focus on coordination problems
- Garrison's analysis of the role of moral hazard in booms.
- Monetary policy has redistributive effects
- The competitive market process under these externalities does not lead to efficient outcomes
- Malinvesting creates profit opportunities, because the individually rational strategy is to participate on the victor's side in the redistributive game
- Not malinvesting is a superior outcome, but it is a prisoner's dilemma and it is unlikely that it can be put in practise because competitive pressures will remove conservative entrepreneurs from the market during the boom
- No one in the market knows the right solution, but prices signal and incentivise malinvestments
- Avoiding herding behavior is a tragedy of the commons because no single entrepreneur contributes to systemic fragility, but every entrepreneur can profit out of it by malinvesting, and suffers losses if he wants to avoid malinvesting (under socialized costs, everybody pays other people's costs, but only adding to malinvestment it is possible to gain profits).
- As long as herding doesn't occur on a mass scale, malinvesting is safe, especially if central banks are active in avoiding recessions.
- The only strategy to avoid losses is to escape the capital intensive sectors because they will fail during the recession (if central banks are not effective in socializing costs through countercylical means), and escape labor-intensive sectors because resources will be bid toward the other sectors during the boom.
- Autarky is the only real solution, but it is so inefficient that inefficient boom/Bust cycles are a superior alternative also when great depressions are likely. It is better to play a vicious game than not to play at all. Avoiding the externality and cost socialization effects of activist monetary policy would largely remove the viciousness of the game, resulting in a first best solution.