I exploit a single natural experiment to estimate the comparative causal effects of different financial stability policies on bank-level credit. In 1920, four Federal Reserve Banks hiked their interest rate indiscriminately to safeguard financial stability. Another four Reserve Banks employed targeted rate action aimed at over-leveraged banks instead. For identification, I draw on border regression discontinuities with the remaining Federal Reserve districts which did not change their stance. The uniform rate hike had counterproductive effects, whereas targeted policy caused credit to contract significantly. The relative strengths of the bank-lending and risk-shifting channels of monetary policy explain these differential outcomes.