Leverage limits as a form of capital regulation have a well-known, potential bug: If banks can’t lever returns as desired, they can boost returns on equity by shifting toward riskier, higher yielding assets. That reach for yield is the leverage rule “arbitrage.”
If banks intended to arbitrage the new rule by shifting toward riskier assets, when along that time line would one expect them to do so? Our original paper focused on the third quarter of 2014, when the SLR denominator was finalized, on the grounds that only then did banks know how binding the rule would be. While we found evidence of risk shifting starting then, our data ended before the compliance date in 2018. Did we miss part of the story?
The main takeaway in our revised paper is unchanged: Banks can be expected to arbitrage simple leverage rules by shifting toward riskier, higher yielding assets. Our revised paper shows that arbitrage was effected well before the compliance date for the new leverage rule, and not in the loan book, where one might most have expected it.