The dog and the frisbee.

This paper explores why the type of complex financial regulation developed over recent decades may be a suboptimal response to the increasing complexity of the financial system. Examples from other disciplines illustrate how decision-making in a complex environment can benefit from simple rules of thumb or 'heuristics'. We perform a set of empirical experiments to assess the relative performance of simple versus complex rules in a financial setting. We find that simple metrics, such as the leverage ratio and market-based measures of capital, outperform more complex risk-weighted measures and multiple-indicator models in their capacity to predict bank failure. A consistent message from these experiments is that complexity of models or portfolios can generate robustness problems. We outline five policy lessons from these findings, covering both the design of financial regulation itself and possible measures aimed at reducing complexity of the financial system more directly. [ABSTRACT FROM AUTHOR]