This preview shows page 1. Sign up to view the full content.
Unformatted text preview: od times and used up during the bad times. In addition, these norms can help coordinate the actions of individual agents and thus overcome the prisoner’s dilemmas‐type situations. Given that the externalities are known (or knowable); this militates in favor of rules over discretion. The mood swings paradigm also makes a strong case for anti‐cyclical prudential norms but for a different reason. Rather than systematically limiting the ship’s speed under clear weather, the main motive in this case is to lift up the yellow flag when, under foggy weather, “icebergs may possibly be lying ahead”. Hence, mood swings provide a rationale for a judgment‐based anti‐cyclical framework, much as the one in effect for monetary policy—
a framework where an independent body would have the discretion to calibrate the anti‐cyclical prudential instrument in light of evolving circumstances. 44 Consider finally the need for (and purpose of) a safety net (Table 3). To a large extent, this question relates to the scope for learning. In a system where learning is possible, it may be preferable to let agents face the hardships of financial crises and learn from experience.79 In the agency paradigm the system is not dominated by uncertainty and mood swings and, hence, should be broadly stationary (even if subjected to innovation). Therefore, agents should eventually learn. This might take a few crises and significant bruises (which in turn require that the ex‐post safety net not systematically validate the ex‐ante expectation of bailouts) but wisdom should eventually arise from the pain.80 Correspondingly, it would be better if the lender‐of‐last‐resort (LOLR) function did not exist. Bank runs are healthy manifestations of market discipline. Stopping runs unnecessarily protects banks that should fail and aggravates the misalignment of incentives for all other banks. Similarly, deposit insurance can only be justified by consumer protection but, given its adverse moral hazard implications, a pure agency supervisor would probably conclude that, on balance, the world would be a better place without it.81 By contrast, in the externalities paradigm, the nature of the problem makes learning irrelevant. As long as externalities are not internalized, participants only see their side of the story, no matter what. Moreover, there is no possible learning from exogenous and random acts of god or from self‐fulfilling runs in a multiple equilibrium world. Thus, to the extent that it is too expensive for society to prevent runs through large ex‐ante buffer requirements, an efficient LOLR becomes a socially superior solution and the cornerstone of the regulatory edifice. Also, as his forebears after the Great Depression, an externalities supervisor would conclude that deposit insurance is needed to induce the small uninformed depositors to join the banking system while preventing them from crying wolf and causing systemic havoc without justification. Again, however, having fire safety only a 911 call away hardly promotes incentives for keeping a fire extinguisher at home, another good exampl...
View Full Document