In this paper, we question whether it is the zero threshold or the target return that impacts the most the risk-taking behavior of individuals when experiencing negative interest rates (NIRs). In our experiment, we attribute either low or high target returns to participants and ask them to make independent investment decisions when the risk-free rate fluctuates around their target returns, and for some of them, turns out negative. We find that the prevailing reference point is the target return, whatever its level. This result still holds even when the risk-free rate becomes negative, suggesting that the target return is a better driver of risk-taking than the zero-lower bound.