Abstract
I employ bank-household relationship data from a Swiss bank to study, for the first time, the response of retail depositors to negative deposit rates. The tiered implementation allows me to employ both a Regression Kink Design and a Difference-in-Differences analysis to investigate the transmission of negative policy rates through the deposits channel. I find that affected households reduce their deposit holdings swiftly and substantially following the official announcement. They respond by investing with the bank some deposits exceeding the exemption threshold, by transferring some excess deposits to accounts held at other banks, and by withdrawing some cash. I do not observe any effect on consumption, which I proxy by debit and credit card spending. I also investigate if and how the characteristics of depositors differentially affect their responses.