Is the short rate drift actually nonlinear?

Ait-Sahalia (1996) and Stanton (1997) use nonparametric estimators applied to short-term interest rate data to conclude that the drift function contai ns important nonlinearities. We study the finite-sample properties of their estimators by applying them to simulated sample paths of a square-root dif fusion. Although the drift function is linear, both estimators suggest nonl inearities of the type and magnitude reported in Ait-Sahalia (1996) and Sta nton (1997). Combined with the results of a weighted least squares estimato r, this evidence implies that nonlinearity of the short rate drift is not a robust stylized fact.